Mie-informed tandem neural network#

Here, we demonstrate how to train a design generator network capable to suggest core-shell particles with specific spectral response using PyMieDiff as differentiable forward-evaluator. The training pipeline follows the “Tandem” model:

target spectrum –> generator NN –> design –> Mie –> real spectrum

training loss is: MSE(target spec., real spec.)

author: O. Jackson, P. Wiecha, 06/2025

imports#

import time

import numpy as np
from tqdm import tqdm
import matplotlib.pyplot as plt
import torch
from torch import nn

import pymiediff as pmd

setup optimiation target#

We setup the main configuration here: torch device, parameter limits and wavelengths

# torch compute device
device = "cpu"

# general config
N_samples = 25000
n_max = 4  # maximum Mie order fixed for performance
eps_env = torch.tensor(1.0, device=device)

lim_r = torch.as_tensor([40, 100], device=device)
lim_n_re = torch.as_tensor([1.5, 4.0], device=device)
lim_n_im = torch.as_tensor([0.0, 0.1], device=device)

wl0 = torch.linspace(400, 800, 40, device=device)
k0 = 2 * torch.pi / wl0

generate reference spectra#

we generate a large number of reference Mie spectra for existing particles, that will be used as design targets during training.

Note: this step could also be done without any physics knowledge, for example with artificial spectra (e.g. Lorentzians), or a scattering maximization loss.

# datagen: generate existing spectra (won't use the geometries for training)
r_c = torch.rand((N_samples), device=device) * torch.diff(lim_r)[0] + lim_r[0]
d_s = torch.rand((N_samples), device=device) * torch.diff(lim_r)[0] + lim_r[0]
r_s = r_c + d_s
n_re = torch.rand((N_samples, 2), device=device) * torch.diff(lim_n_re)[0] + lim_n_re[0]
n_im = torch.rand((N_samples, 2), device=device) * torch.diff(lim_n_im)[0] + lim_n_im[0]
n = n_re + 1j * n_im

# low-level API: permittivity required as spectra (for vectorization)
eps_c = torch.ones_like(k0).unsqueeze(0) * n[:, 0].unsqueeze(1) ** 2
eps_s = torch.ones_like(k0).unsqueeze(0) * n[:, 1].unsqueeze(1) ** 2

all_particles = pmd.multishell.cross_sections(
    k0,
    r_c=r_c,
    eps_c=eps_c,
    r_s=r_s,
    eps_s=eps_s,
    eps_env=eps_env,
    n_max=n_max,
)

N_test = 128  # keep a few samples for testing
q_sca_target = all_particles["q_sca"][N_test:].to(dtype=torch.float32)
q_sca_target_test = all_particles["q_sca"][:N_test].to(dtype=torch.float32)

plt.plot(q_sca_target[30].detach().cpu().numpy())  # plot some test sample
ex 09 tandem
[<matplotlib.lines.Line2D object at 0x7f4abc60f680>]

Neural network classes / functions#

define the network model (simple MLP) and training loop

class FullyConnected(nn.Module):
    def __init__(self, hidden_dim=1024):
        super().__init__()
        self.fc_in = nn.Linear(len(k0), hidden_dim)
        self.relu1 = nn.ReLU()
        self.fc_1 = nn.Linear(hidden_dim, hidden_dim)
        self.relu2 = nn.ReLU()
        self.fc_2 = nn.Linear(hidden_dim, hidden_dim)
        self.relu3 = nn.ReLU()
        self.fc_out = nn.Linear(hidden_dim, 6)
        self.sigmoid = nn.Sigmoid()

    def forward(self, x):
        x = self.fc_in(x)
        x = self.relu1(x)
        x = self.fc_1(x)
        x = self.relu2(x)
        x = self.fc_2(x)
        x = self.relu3(x)
        x = self.fc_out(x)
        x = self.sigmoid(x)
        return x


def nn_pred_to_mie_geometry(pred):
    # implicit normalization: multiply by user-defined limits
    r_c = lim_r.max() * (pred[:, 0])
    r_s = lim_r.max() * (pred[:, 0] + pred[:, 1])
    n_c = lim_n_re.max() * pred[:, 2] + lim_n_im.max() * (1j * pred[:, 3])
    n_s = lim_n_re.max() * pred[:, 4] + lim_n_im.max() * (1j * pred[:, 5])

    eps_c = torch.ones_like(k0).unsqueeze(0) * n_c.unsqueeze(1) ** 2
    eps_s = torch.ones_like(k0).unsqueeze(0) * n_s.unsqueeze(1) ** 2

    return r_c, r_s, eps_c, eps_s


def train_loop(dataloader, model, loss_fn, optimizer):
    size = len(dataloader.dataset)
    # Set the model to training mode - important for batch normalization and dropout layers
    # Unnecessary in this situation but added for best practices
    model.train()
    prog_bar = tqdm(enumerate(dataloader), total=size // dataloader.batch_size)
    for i_batch, X in prog_bar:
        # model prediction: generate core-shell particles
        pred = model(X)

        # evaluate Mie
        r_c, r_s, eps_c, eps_s = nn_pred_to_mie_geometry(pred)
        res_mie = pmd.multishell.cross_sections(
            k0,
            r_c=r_c,
            eps_c=eps_c,
            r_s=r_s,
            eps_s=eps_s,
            eps_env=eps_env,
            n_max=n_max,
        )
        q_sca_mie = res_mie["q_sca"].to(dtype=torch.float32)

        # calc. loss
        loss = loss_fn(q_sca_mie, X)

        # Backpropagation
        loss.backward()
        optimizer.step()
        optimizer.zero_grad()

        # if i_batch % 100 == 0:
        loss, current = loss.item(), i_batch * dataloader.batch_size + len(X)
        prog_bar.set_description(f"loss: {loss:>7f}  [{current:>5d}/{size:>5d}]")

training the Mie-informed network#

here we use some simple, manually optimized training schedule.

model = FullyConnected().to(device)

confs = [
    dict(bs=32, lr=1e-4, n_ep=5),
    dict(bs=64, lr=1e-4, n_ep=5),
    dict(bs=128, lr=1e-4, n_ep=6),
    dict(bs=256, lr=1e-5, n_ep=6),
]

t_start = time.time()
for conf in confs:
    learning_rate = conf["lr"]
    batch_size = conf["bs"]
    epochs = conf["n_ep"]
    print("-------------------------------")
    print(f"LR={learning_rate}, batch_size={batch_size}")
    print("-------------------------------")

    loss_fn = nn.MSELoss()
    optimizer = torch.optim.AdamW(model.parameters(), lr=learning_rate)
    train_dataloader = torch.utils.data.DataLoader(q_sca_target, batch_size=batch_size)
    for t in range(epochs):
        print(f"Epoch {t+1}, time={time.time()-t_start:.2f}s")
        train_loop(train_dataloader, model, loss_fn, optimizer)
print("Done!")
-------------------------------
LR=0.0001, batch_size=32
-------------------------------
Epoch 1, time=0.00s

  0%|          | 0/777 [00:00<?, ?it/s]
loss: 5.448700  [   32/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 3.663493  [   64/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 2.298568  [   96/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 2.324271  [  128/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 2.463414  [  160/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.619177  [  192/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.218018  [  224/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.488135  [  256/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 2.377337  [  288/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.485607  [  320/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.640749  [  352/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.434125  [  384/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.255051  [  416/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.618245  [  448/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.495192  [  480/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.097729  [  512/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.251212  [  544/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.979863  [  576/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.157405  [  608/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.202960  [  640/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.945543  [  672/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.881338  [  704/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.065528  [  736/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.363751  [  768/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.969685  [  800/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.955333  [  832/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.110681  [  864/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.136441  [  896/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.162973  [  928/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.876811  [  960/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.838082  [  992/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 1.078258  [ 1024/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.007576  [ 1056/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.289056  [ 1088/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.061889  [ 1120/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.065302  [ 1152/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.060543  [ 1184/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.167970  [ 1216/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.059351  [ 1248/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.959957  [ 1280/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.192208  [ 1312/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.065513  [ 1344/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.986991  [ 1376/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.070646  [ 1408/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.863589  [ 1440/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.998641  [ 1472/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.017907  [ 1504/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.012782  [ 1536/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.033771  [ 1568/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.852011  [ 1600/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.943541  [ 1632/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.849178  [ 1664/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.857003  [ 1696/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.918904  [ 1728/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.031009  [ 1760/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.963788  [ 1792/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.260799  [ 1824/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.841250  [ 1856/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.023981  [ 1888/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.900256  [ 1920/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 1.064776  [ 1952/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.925141  [ 1984/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.841450  [ 2016/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.736603  [ 2048/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.906959  [ 2080/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.884796  [ 2112/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.989985  [ 2144/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.893747  [ 2176/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.792371  [ 2208/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.940944  [ 2240/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.844516  [ 2272/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.876358  [ 2304/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.815739  [ 2336/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.587187  [ 2368/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.734632  [ 2400/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 1.070787  [ 2432/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.638775  [ 2464/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.915217  [ 2496/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.818202  [ 2528/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.667059  [ 2560/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.759481  [ 2592/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.923577  [ 2624/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.878718  [ 2656/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.718933  [ 2688/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.912528  [ 2720/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.871151  [ 2752/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 1.044065  [ 2784/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.723555  [ 2816/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.885034  [ 2848/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 1.036914  [ 2880/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.809558  [ 2912/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.832565  [ 2944/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.663263  [ 2976/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.826658  [ 3008/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.684103  [ 3040/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.539309  [ 3072/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.743163  [ 3104/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.834029  [ 3136/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.698505  [ 3168/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.903880  [ 3200/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.759959  [ 3232/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.867580  [ 3264/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.972969  [ 3296/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.740204  [ 3328/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.900527  [ 3360/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.657807  [ 3392/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.980775  [ 3424/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.623247  [ 3456/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.954146  [ 3488/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.678354  [ 3520/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.645892  [ 3552/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.591773  [ 3584/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 1.095475  [ 3616/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.821005  [ 3648/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.606349  [ 3680/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.892637  [ 3712/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.728888  [ 3744/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.795566  [ 3776/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.726039  [ 3808/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.678046  [ 3840/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.770394  [ 3872/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.605684  [ 3904/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.729368  [ 3936/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.705876  [ 3968/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.688832  [ 4000/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.638765  [ 4032/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.574220  [ 4064/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.876115  [ 4096/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.541657  [ 4128/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.711450  [ 4160/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.682272  [ 4192/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.683032  [ 4224/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.854435  [ 4256/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.652961  [ 4288/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.591220  [ 4320/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.745088  [ 4352/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.637187  [ 4384/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.642716  [ 4416/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.712501  [ 4448/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.676388  [ 4480/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.701705  [ 4512/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.683462  [ 4544/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.733582  [ 4576/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.609111  [ 4608/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.689641  [ 4640/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.480562  [ 4672/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.614428  [ 4704/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.678851  [ 4736/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.732386  [ 4768/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.749226  [ 4800/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.703311  [ 4832/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.521778  [ 4864/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.507884  [ 4896/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.768646  [ 4928/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.714992  [ 4960/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.663723  [ 4992/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.661291  [ 5024/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.703369  [ 5056/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.675800  [ 5088/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.634283  [ 5120/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.684459  [ 5152/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.781871  [ 5184/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.643895  [ 5216/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.787199  [ 5248/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.619512  [ 5280/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.708954  [ 5312/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.667029  [ 5344/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.690075  [ 5376/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.658483  [ 5408/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.678483  [ 5440/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.614700  [ 5472/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.603363  [ 5504/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.608069  [ 5536/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.645513  [ 5568/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.616541  [ 5600/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.674389  [ 5632/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.767456  [ 5664/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.583954  [ 5696/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.512420  [ 5728/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.649347  [ 5760/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.616819  [ 5792/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.617867  [ 5824/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.796194  [ 5856/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.518474  [ 5888/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.714679  [ 5920/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.442995  [ 5952/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.517427  [ 5984/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.541128  [ 6016/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.495323  [ 6048/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.515211  [ 6080/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.439090  [ 6112/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.632916  [ 6144/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.727044  [ 6176/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.747546  [ 6208/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.553123  [ 6240/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.395849  [ 6272/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.517832  [ 6304/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.563865  [ 6336/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.546961  [ 6368/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.660370  [ 6400/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.538042  [ 6432/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.473806  [ 6464/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.533678  [ 6496/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.387664  [ 6528/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.404922  [ 6560/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.534566  [ 6592/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.533005  [ 6624/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.522008  [ 6656/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.564578  [ 6688/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.508735  [ 6720/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.759057  [ 6752/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.470274  [ 6784/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.744000  [ 6816/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.459984  [ 6848/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.463558  [ 6880/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.538192  [ 6912/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.417475  [ 6944/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.530301  [ 6976/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.557639  [ 7008/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.616453  [ 7040/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.498269  [ 7072/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.489556  [ 7104/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.652338  [ 7136/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.732578  [ 7168/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.645082  [ 7200/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.513922  [ 7232/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.539261  [ 7264/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.799324  [ 7296/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.472952  [ 7328/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.534694  [ 7360/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.542884  [ 7392/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.506920  [ 7424/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.481067  [ 7456/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.442637  [ 7488/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.523588  [ 7520/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.596199  [ 7552/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.602646  [ 7584/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.471975  [ 7616/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.513393  [ 7648/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.350695  [ 7680/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.747438  [ 7712/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.531238  [ 7744/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.466451  [ 7776/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.507040  [ 7808/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.239299  [ 7840/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.469188  [ 7872/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.545143  [ 7904/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.497254  [ 7936/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.401271  [ 7968/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.618092  [ 8000/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.504744  [ 8032/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.383856  [ 8064/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.462373  [ 8096/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.631609  [ 8128/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.542796  [ 8160/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.656749  [ 8192/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.426156  [ 8224/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.357494  [ 8256/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.365555  [ 8288/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.419323  [ 8320/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.470439  [ 8352/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.374216  [ 8384/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.483443  [ 8416/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.428302  [ 8448/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.488254  [ 8480/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.521381  [ 8512/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.421690  [ 8544/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.513558  [ 8576/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.510744  [ 8608/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.487041  [ 8640/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.692241  [ 8672/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.332279  [ 8704/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.712257  [ 8736/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.554774  [ 8768/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.439494  [ 8800/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.647328  [ 8832/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.437923  [ 8864/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.355299  [ 8896/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.468973  [ 8928/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.489199  [ 8960/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.514852  [ 8992/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.362683  [ 9024/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.454217  [ 9056/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.418170  [ 9088/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.563678  [ 9120/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.514727  [ 9152/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.648462  [ 9184/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.321158  [ 9216/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.478772  [ 9248/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.567859  [ 9280/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.420853  [ 9312/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.506302  [ 9344/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.602240  [ 9376/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.412735  [ 9408/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.451525  [ 9440/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.558400  [ 9472/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.430024  [ 9504/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.496836  [ 9536/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.647344  [ 9568/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.465062  [ 9600/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.505794  [ 9632/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.396252  [ 9664/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.408746  [ 9696/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.436139  [ 9728/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.550601  [ 9760/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.534968  [ 9792/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.320899  [ 9824/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.557011  [ 9856/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.414068  [ 9888/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.427900  [ 9920/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.443178  [ 9952/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.571267  [ 9984/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.467330  [10016/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.508847  [10048/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.387908  [10080/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.442852  [10112/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.519572  [10144/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.583122  [10176/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.600485  [10208/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.428535  [10240/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.544929  [10272/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.492122  [10304/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.476132  [10336/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.593576  [10368/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.614156  [10400/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.559559  [10432/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.484108  [10464/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.530840  [10496/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.528987  [10528/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.632759  [10560/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.633583  [10592/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.484657  [10624/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.417428  [10656/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.499047  [10688/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.474632  [10720/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.545221  [10752/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.318260  [10784/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.666019  [10816/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.468783  [10848/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.379275  [10880/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.460202  [10912/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.504083  [10944/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.326843  [10976/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.490097  [11008/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.431162  [11040/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.418099  [11072/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.441412  [11104/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.502852  [11136/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.676106  [11168/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.449186  [11200/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.380786  [11232/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.405778  [11264/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.555996  [11296/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.390253  [11328/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.539111  [11360/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.524484  [11392/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.488574  [11424/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.744495  [11456/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.444139  [11488/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.521785  [11520/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.518607  [11552/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.489261  [11584/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.471637  [11616/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.574206  [11648/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.496782  [11680/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.424928  [11712/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.586598  [11744/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.512546  [11776/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.318875  [11808/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.406139  [11840/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.485121  [11872/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.416627  [11904/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.508901  [11936/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.490011  [11968/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.519166  [12000/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.434755  [12032/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.404810  [12064/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.523661  [12096/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.514388  [12128/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.294106  [12160/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.464972  [12192/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.593300  [12224/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.506862  [12256/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.480553  [12288/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.485425  [12320/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.399036  [12352/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.330093  [12384/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.404147  [12416/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.529645  [12448/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.353614  [12480/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.403320  [12512/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.376682  [12544/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.471347  [12576/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.526796  [12608/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.328139  [12640/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.401838  [12672/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.684493  [12704/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.455921  [12736/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.359483  [12768/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.486836  [12800/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.462719  [12832/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.423925  [12864/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.451563  [12896/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.449925  [12928/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.433199  [12960/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.453318  [12992/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.335175  [13024/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.486410  [13056/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.524020  [13088/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.544302  [13120/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.438840  [13152/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.394432  [13184/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.430401  [13216/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.462845  [13248/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.487709  [13280/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.449452  [13312/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.324024  [13344/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.417986  [13376/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.479221  [13408/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.578277  [13440/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.337800  [13472/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.399097  [13504/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.343665  [13536/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.510120  [13568/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.402394  [13600/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.512672  [13632/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.309962  [13664/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.541067  [13696/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.352255  [13728/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.335172  [13760/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.368510  [13792/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.500297  [13824/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.495899  [13856/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.320860  [13888/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.540649  [13920/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.338154  [13952/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.471250  [13984/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.478890  [14016/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.295760  [14048/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.472142  [14080/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.286411  [14112/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.428685  [14144/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.463878  [14176/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.440785  [14208/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.425580  [14240/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.523268  [14272/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.526833  [14304/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.397676  [14336/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.474378  [14368/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.449664  [14400/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.303963  [14432/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.472512  [14464/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.330364  [14496/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.420289  [14528/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.368729  [14560/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.345973  [14592/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.414004  [14624/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.330256  [14656/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.383578  [14688/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.380113  [14720/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.588775  [14752/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.459903  [14784/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.443530  [14816/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.441572  [14848/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.408186  [14880/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.378248  [14912/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.489657  [14944/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.433293  [14976/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.534318  [15008/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.368415  [15040/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.401601  [15072/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.472205  [15104/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.522566  [15136/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.358398  [15168/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.503075  [15200/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.355528  [15232/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.398440  [15264/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.395839  [15296/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.446689  [15328/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.301980  [15360/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.372225  [15392/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.351524  [15424/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.351438  [15456/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.346903  [15488/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.348297  [15520/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.492771  [15552/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.422290  [15584/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.434082  [15616/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.414861  [15648/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.414433  [15680/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.601099  [15712/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.532972  [15744/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.395891  [15776/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.372291  [15808/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.641460  [15840/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.618281  [15872/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.376116  [15904/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.475130  [15936/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.612006  [15968/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.363322  [16000/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.365692  [16032/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.390096  [16064/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.609052  [16096/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.457431  [16128/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.369082  [16160/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.546495  [16192/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.644719  [16224/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.363397  [16256/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.301369  [16288/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.309107  [16320/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.397006  [16352/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.384712  [16384/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.460561  [16416/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.347333  [16448/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.354785  [16480/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.454660  [16512/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.536247  [16544/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.425474  [16576/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.285794  [16608/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.293526  [16640/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.467054  [16672/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.406811  [16704/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.419453  [16736/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.446881  [16768/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.520388  [16800/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.418686  [16832/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.399088  [16864/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.388615  [16896/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.419480  [16928/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.464948  [16960/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.354320  [16992/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.376799  [17024/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.465485  [17056/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.531368  [17088/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.312568  [17120/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.388329  [17152/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.448919  [17184/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.392582  [17216/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.393337  [17248/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.401279  [17280/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.420511  [17312/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.330511  [17344/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.517993  [17376/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.373700  [17408/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.464632  [17440/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.344447  [17472/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.446186  [17504/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.407651  [17536/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.323964  [17568/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.386219  [17600/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.364323  [17632/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.470013  [17664/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.359881  [17696/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.404507  [17728/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.359933  [17760/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.366230  [17792/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.328239  [17824/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.343382  [17856/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.354888  [17888/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.342384  [17920/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.440413  [17952/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.304789  [17984/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.270986  [18016/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.402048  [18048/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.300997  [18080/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.330074  [18112/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.385514  [18144/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.358278  [18176/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.313483  [18208/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.293657  [18240/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.404274  [18272/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.268384  [18304/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.434828  [18336/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.303429  [18368/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.294808  [18400/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.354382  [18432/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.346023  [18464/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.293719  [18496/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.389734  [18528/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.349967  [18560/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.258317  [18592/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.315560  [18624/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.398751  [18656/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.382400  [18688/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.266891  [18720/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.326146  [18752/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.447098  [18784/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.343265  [18816/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.453164  [18848/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.284076  [18880/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.400021  [18912/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.323491  [18944/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.414690  [18976/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.405972  [19008/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.266740  [19040/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.442540  [19072/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.424802  [19104/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.479264  [19136/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.333723  [19168/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.292154  [19200/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.285031  [19232/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.228260  [19264/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.310184  [19296/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.427393  [19328/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.299997  [19360/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.422263  [19392/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.319757  [19424/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.356080  [19456/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.306720  [19488/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.384665  [19520/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.322606  [19552/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.305714  [19584/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.356175  [19616/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.365771  [19648/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.435303  [19680/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.411198  [19712/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.509775  [19744/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.290471  [19776/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.354576  [19808/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.396693  [19840/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.398678  [19872/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.290421  [19904/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.392916  [19936/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.423040  [19968/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.343359  [20000/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.385984  [20032/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.386070  [20064/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.348620  [20096/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.312295  [20128/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.394371  [20160/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.547828  [20192/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.381137  [20224/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.333629  [20256/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.294285  [20288/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.467912  [20320/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.436856  [20352/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.390298  [20384/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.437056  [20416/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.288773  [20448/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.317635  [20480/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.469675  [20512/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.405985  [20544/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.309080  [20576/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.365673  [20608/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.492027  [20640/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.437511  [20672/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.486746  [20704/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.288749  [20736/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.381436  [20768/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.310648  [20800/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.376233  [20832/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.355911  [20864/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.510757  [20896/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.308812  [20928/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.254389  [20960/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.515697  [20992/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.444257  [21024/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.254018  [21056/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.331115  [21088/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.422268  [21120/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.406137  [21152/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.387927  [21184/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.401215  [21216/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.422137  [21248/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.467214  [21280/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.543023  [21312/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.351560  [21344/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.375082  [21376/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.329417  [21408/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.461973  [21440/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.294489  [21472/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.165654  [21504/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.416373  [21536/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.397015  [21568/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.596088  [21600/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.277502  [21632/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.281731  [21664/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.375466  [21696/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.374726  [21728/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.268567  [21760/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.274062  [21792/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.216668  [21824/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.422309  [21856/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.316591  [21888/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.413076  [21920/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.426437  [21952/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.378532  [21984/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.330323  [22016/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.339580  [22048/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.351520  [22080/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.387965  [22112/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.406794  [22144/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.315828  [22176/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.317977  [22208/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.437610  [22240/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.280629  [22272/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.370426  [22304/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.439755  [22336/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.450386  [22368/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.473990  [22400/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.353728  [22432/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.305225  [22464/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.339394  [22496/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.387535  [22528/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.223432  [22560/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.410597  [22592/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.283375  [22624/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.392131  [22656/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.336439  [22688/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.517732  [22720/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.272371  [22752/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.294427  [22784/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.400697  [22816/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.341480  [22848/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.343180  [22880/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.330610  [22912/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.424573  [22944/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.506812  [22976/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.289746  [23008/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.333248  [23040/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.287541  [23072/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.442625  [23104/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.317223  [23136/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.402213  [23168/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.425622  [23200/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.537813  [23232/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.362956  [23264/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.238879  [23296/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.303360  [23328/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.302297  [23360/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.364207  [23392/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.335333  [23424/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.354395  [23456/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.353510  [23488/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.407268  [23520/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.283559  [23552/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.296357  [23584/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.315954  [23616/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.270659  [23648/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.290232  [23680/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.383696  [23712/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.415338  [23744/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.247153  [23776/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.322163  [23808/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.397077  [23840/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.396980  [23872/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.388619  [23904/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.267641  [23936/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.399724  [23968/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.366524  [24000/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.382765  [24032/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.336132  [24064/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.255242  [24096/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.438657  [24128/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.343714  [24160/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.463360  [24192/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.298950  [24224/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.390201  [24256/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.276461  [24288/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.247414  [24320/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.454197  [24352/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.295771  [24384/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.487520  [24416/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.345810  [24448/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.256958  [24480/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.396443  [24512/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.241695  [24544/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.258353  [24576/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.402762  [24608/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.279682  [24640/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.263939  [24672/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.284468  [24704/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.304725  [24736/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.339177  [24768/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.318566  [24800/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.311885  [24832/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.309259  [24864/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.449649  [24872/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.449649  [24872/24872]: : 778it [00:23, 32.84it/s]
Epoch 2, time=23.69s

  0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.321199  [   32/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.379899  [   64/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.285944  [   96/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.291913  [  128/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.354536  [  160/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.266818  [  192/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.262643  [  224/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.296268  [  256/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.277587  [  288/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.247391  [  320/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.420611  [  352/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.409803  [  384/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.213102  [  416/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.363554  [  448/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.395119  [  480/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.273261  [  512/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.276058  [  544/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.267622  [  576/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.256123  [  608/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.256231  [  640/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.318066  [  672/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.217886  [  704/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.356283  [  736/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.216446  [  768/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.255701  [  800/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.350390  [  832/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.476541  [  864/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.376237  [  896/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.361687  [  928/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.288321  [  960/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.312452  [  992/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.365211  [ 1024/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.241302  [ 1056/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.321956  [ 1088/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.299661  [ 1120/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.291968  [ 1152/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.326706  [ 1184/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.449162  [ 1216/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.284366  [ 1248/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.292315  [ 1280/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.478536  [ 1312/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.367570  [ 1344/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.357452  [ 1376/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.385586  [ 1408/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.266499  [ 1440/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.356019  [ 1472/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.429156  [ 1504/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.387255  [ 1536/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.348287  [ 1568/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.344500  [ 1600/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.345153  [ 1632/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.390501  [ 1664/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.333029  [ 1696/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.436320  [ 1728/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.377760  [ 1760/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.416300  [ 1792/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.435662  [ 1824/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.318103  [ 1856/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.375803  [ 1888/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.337353  [ 1920/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.420628  [ 1952/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.287259  [ 1984/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.251469  [ 2016/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.223608  [ 2048/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.367739  [ 2080/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.260508  [ 2112/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.417447  [ 2144/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.308830  [ 2176/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.239086  [ 2208/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.280524  [ 2240/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.323859  [ 2272/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.265341  [ 2304/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.394208  [ 2336/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.256918  [ 2368/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.323269  [ 2400/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.346541  [ 2432/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.238255  [ 2464/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.277269  [ 2496/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.378373  [ 2528/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.309451  [ 2560/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.271896  [ 2592/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.304492  [ 2624/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.324012  [ 2656/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.236462  [ 2688/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.324244  [ 2720/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.248035  [ 2752/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.471652  [ 2784/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.265707  [ 2816/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.415113  [ 2848/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.384879  [ 2880/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.249605  [ 2912/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.307309  [ 2944/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.252140  [ 2976/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.281557  [ 3008/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.317622  [ 3040/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.296100  [ 3072/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.342891  [ 3104/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.340519  [ 3136/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.222210  [ 3168/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.348850  [ 3200/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.340505  [ 3232/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.315857  [ 3264/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.422217  [ 3296/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.301146  [ 3328/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.330590  [ 3360/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.226649  [ 3392/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.407049  [ 3424/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.270466  [ 3456/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.430832  [ 3488/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.302546  [ 3520/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.335039  [ 3552/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.268721  [ 3584/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.347405  [ 3616/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.254941  [ 3648/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.292868  [ 3680/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.489535  [ 3712/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.307850  [ 3744/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.250625  [ 3776/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.300760  [ 3808/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.292565  [ 3840/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.320899  [ 3872/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.198272  [ 3904/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.373108  [ 3936/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.284578  [ 3968/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.309038  [ 4000/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.336448  [ 4032/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.276545  [ 4064/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.509292  [ 4096/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.282013  [ 4128/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.386379  [ 4160/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.301398  [ 4192/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.288118  [ 4224/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.427548  [ 4256/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.274848  [ 4288/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.244449  [ 4320/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.343408  [ 4352/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.259861  [ 4384/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.219111  [ 4416/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.392369  [ 4448/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.354246  [ 4480/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.352642  [ 4512/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.403620  [ 4544/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.356829  [ 4576/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.384111  [ 4608/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.401035  [ 4640/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.259892  [ 4672/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.393233  [ 4704/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.410945  [ 4736/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.404845  [ 4768/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.358039  [ 4800/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.391879  [ 4832/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.292996  [ 4864/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.298166  [ 4896/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.390777  [ 4928/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.411227  [ 4960/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.278542  [ 4992/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.397445  [ 5024/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.274664  [ 5056/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.370896  [ 5088/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.416818  [ 5120/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.303133  [ 5152/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.413748  [ 5184/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.307055  [ 5216/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.377750  [ 5248/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.331892  [ 5280/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.407197  [ 5312/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.328366  [ 5344/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.373006  [ 5376/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.395742  [ 5408/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.397866  [ 5440/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.279913  [ 5472/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.352862  [ 5504/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.319731  [ 5536/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.322788  [ 5568/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.310996  [ 5600/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.347918  [ 5632/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.453603  [ 5664/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.287532  [ 5696/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.216558  [ 5728/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.345663  [ 5760/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.346332  [ 5792/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.336728  [ 5824/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.485549  [ 5856/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.226733  [ 5888/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.425324  [ 5920/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.253209  [ 5952/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.246082  [ 5984/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.266604  [ 6016/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.272942  [ 6048/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.228812  [ 6080/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.264069  [ 6112/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.298879  [ 6144/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.313678  [ 6176/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.378576  [ 6208/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.221646  [ 6240/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.265242  [ 6272/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.306385  [ 6304/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.284636  [ 6336/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.253387  [ 6368/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.420786  [ 6400/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.234849  [ 6432/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.232044  [ 6464/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.253969  [ 6496/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.187895  [ 6528/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.201912  [ 6560/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.312035  [ 6592/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.287801  [ 6624/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.252280  [ 6656/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.306741  [ 6688/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.261750  [ 6720/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.344676  [ 6752/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.209838  [ 6784/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.424177  [ 6816/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.220688  [ 6848/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.247084  [ 6880/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.383085  [ 6912/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.267947  [ 6944/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.297383  [ 6976/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.357337  [ 7008/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.299353  [ 7040/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.309061  [ 7072/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.263218  [ 7104/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.426112  [ 7136/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.330752  [ 7168/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.460006  [ 7200/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.385451  [ 7232/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.354523  [ 7264/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.459632  [ 7296/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.245040  [ 7328/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.256484  [ 7360/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.311241  [ 7392/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.347213  [ 7424/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.266741  [ 7456/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.214014  [ 7488/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.259421  [ 7520/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.317469  [ 7552/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.277852  [ 7584/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.287398  [ 7616/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.294038  [ 7648/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.236374  [ 7680/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.443853  [ 7712/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.258800  [ 7744/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.275211  [ 7776/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.294761  [ 7808/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.185730  [ 7840/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.311859  [ 7872/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.316602  [ 7904/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.336708  [ 7936/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.239646  [ 7968/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.332025  [ 8000/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.302444  [ 8032/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.188348  [ 8064/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.279544  [ 8096/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.374581  [ 8128/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.332872  [ 8160/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.443320  [ 8192/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.283056  [ 8224/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.183099  [ 8256/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.192049  [ 8288/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.246930  [ 8320/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.269612  [ 8352/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.223978  [ 8384/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.283178  [ 8416/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.296413  [ 8448/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.270716  [ 8480/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.258485  [ 8512/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.278585  [ 8544/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.304110  [ 8576/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.355132  [ 8608/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.366720  [ 8640/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.430017  [ 8672/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.252884  [ 8704/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.502948  [ 8736/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.322494  [ 8768/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.232224  [ 8800/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.371032  [ 8832/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.386936  [ 8864/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.245494  [ 8896/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.225838  [ 8928/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.352749  [ 8960/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.365917  [ 8992/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.223097  [ 9024/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.297282  [ 9056/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.340726  [ 9088/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.318007  [ 9120/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.330012  [ 9152/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.395256  [ 9184/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.224911  [ 9216/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.345352  [ 9248/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.346348  [ 9280/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.242034  [ 9312/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.315248  [ 9344/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.425582  [ 9376/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.282405  [ 9408/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.260980  [ 9440/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.371535  [ 9472/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.250165  [ 9504/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.325483  [ 9536/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.388561  [ 9568/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.315658  [ 9600/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.307039  [ 9632/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.335091  [ 9664/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.281448  [ 9696/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.313617  [ 9728/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.344193  [ 9760/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.273591  [ 9792/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.215083  [ 9824/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.409554  [ 9856/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.249804  [ 9888/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.303871  [ 9920/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.277130  [ 9952/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.379795  [ 9984/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.291035  [10016/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.253729  [10048/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.189898  [10080/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.246539  [10112/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.317030  [10144/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.354075  [10176/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.268354  [10208/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.305089  [10240/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.371549  [10272/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.335768  [10304/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.261001  [10336/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.377635  [10368/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.367158  [10400/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.302886  [10432/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.271689  [10464/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.311708  [10496/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.338346  [10528/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.452931  [10560/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.387137  [10592/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.341691  [10624/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.240896  [10656/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.302932  [10688/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.378426  [10720/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.380468  [10752/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.307630  [10784/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.372420  [10816/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.339459  [10848/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.277079  [10880/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.358060  [10912/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.352704  [10944/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.213809  [10976/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.269091  [11008/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.278034  [11040/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.311215  [11072/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.292631  [11104/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.367979  [11136/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.424763  [11168/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.278472  [11200/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.324011  [11232/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.269191  [11264/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.312465  [11296/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.301648  [11328/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.314115  [11360/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.309074  [11392/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.224815  [11424/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.456540  [11456/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.248161  [11488/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.268902  [11520/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.340612  [11552/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.303440  [11584/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.337149  [11616/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.344116  [11648/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.304918  [11680/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.306018  [11712/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.321322  [11744/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.318233  [11776/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.253299  [11808/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.234002  [11840/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.267204  [11872/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.219119  [11904/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.324566  [11936/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.379999  [11968/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.313192  [12000/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.284313  [12032/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.282748  [12064/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.375748  [12096/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.309264  [12128/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.208285  [12160/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.366890  [12192/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.371512  [12224/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.337174  [12256/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.255238  [12288/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.348903  [12320/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.351482  [12352/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.207874  [12384/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.265185  [12416/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.360543  [12448/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.183235  [12480/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.274747  [12512/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.245377  [12544/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.279163  [12576/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.347588  [12608/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.208572  [12640/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.263656  [12672/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.390795  [12704/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.291882  [12736/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.275329  [12768/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.411973  [12800/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.293056  [12832/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.308307  [12864/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.350689  [12896/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.359066  [12928/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.369044  [12960/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.394291  [12992/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.230161  [13024/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.322012  [13056/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.308125  [13088/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.402821  [13120/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.279950  [13152/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.251036  [13184/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.268780  [13216/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.284926  [13248/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.298663  [13280/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.364959  [13312/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.208744  [13344/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.325692  [13376/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.345246  [13408/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.369805  [13440/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.254536  [13472/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.269645  [13504/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.196617  [13536/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.364898  [13568/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.287886  [13600/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.402837  [13632/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.222195  [13664/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.338569  [13696/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.211536  [13728/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.231559  [13760/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.243728  [13792/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.363930  [13824/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.390244  [13856/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.226631  [13888/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.362478  [13920/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.255153  [13952/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.326037  [13984/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.343184  [14016/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.212189  [14048/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.298394  [14080/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.208125  [14112/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.293454  [14144/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.326411  [14176/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.284590  [14208/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.319162  [14240/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.442390  [14272/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.346398  [14304/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.274514  [14336/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.360146  [14368/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.327817  [14400/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.215036  [14432/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.285003  [14464/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.201535  [14496/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.251022  [14528/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.217819  [14560/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.261825  [14592/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.285813  [14624/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.176865  [14656/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.255371  [14688/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.198487  [14720/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.364525  [14752/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.303321  [14784/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.345831  [14816/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.250819  [14848/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.275599  [14880/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.248324  [14912/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.372686  [14944/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.314805  [14976/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.353657  [15008/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.279210  [15040/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.285443  [15072/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.287028  [15104/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.340657  [15136/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.224500  [15168/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.380591  [15200/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.210911  [15232/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.262329  [15264/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.229486  [15296/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.322278  [15328/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.227025  [15360/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.271343  [15392/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.271663  [15424/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.250002  [15456/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.204637  [15488/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.244633  [15520/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.323348  [15552/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.269931  [15584/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.244008  [15616/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.237436  [15648/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.349481  [15680/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.312257  [15712/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.369815  [15744/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.272283  [15776/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.228592  [15808/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.336051  [15840/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.300735  [15872/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.191755  [15904/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.350221  [15936/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.368473  [15968/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.149961  [16000/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.250171  [16032/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.235441  [16064/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.312097  [16096/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.264830  [16128/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.253787  [16160/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.418474  [16192/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.363646  [16224/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.260897  [16256/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.267511  [16288/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.209467  [16320/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.275108  [16352/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.275938  [16384/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.345462  [16416/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.292513  [16448/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.239807  [16480/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.345984  [16512/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.454436  [16544/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.317695  [16576/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.192135  [16608/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.192755  [16640/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.343483  [16672/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.311445  [16704/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.352248  [16736/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.297888  [16768/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.375568  [16800/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.324133  [16832/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.255925  [16864/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.263375  [16896/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.358511  [16928/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.347365  [16960/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.257273  [16992/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.337670  [17024/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.390425  [17056/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.379479  [17088/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.223562  [17120/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.322202  [17152/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.337668  [17184/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.294312  [17216/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.309786  [17248/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.310658  [17280/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.321359  [17312/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.256207  [17344/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.353765  [17376/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.287754  [17408/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.375274  [17440/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.217034  [17472/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.349417  [17504/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.323076  [17536/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.249543  [17568/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.257552  [17600/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.267001  [17632/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.336182  [17664/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.262725  [17696/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.275710  [17728/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.269609  [17760/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.300158  [17792/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.253106  [17824/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.302913  [17856/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.263888  [17888/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.275860  [17920/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.312406  [17952/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.224046  [17984/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.220997  [18016/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.302218  [18048/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.219568  [18080/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.247359  [18112/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.236679  [18144/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.249246  [18176/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.216917  [18208/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.207742  [18240/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.281550  [18272/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.210628  [18304/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.344093  [18336/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.194303  [18368/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.246135  [18400/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.289864  [18432/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.294127  [18464/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.249726  [18496/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.346553  [18528/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.312542  [18560/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.201839  [18592/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.225444  [18624/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.306960  [18656/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.275703  [18688/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.225078  [18720/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.258122  [18752/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.371748  [18784/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.250263  [18816/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.416229  [18848/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.169098  [18880/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.280494  [18912/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.266658  [18944/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.312179  [18976/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.313062  [19008/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.194137  [19040/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.362763  [19072/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.289332  [19104/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.344841  [19136/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.244717  [19168/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.170956  [19200/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.230900  [19232/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.210360  [19264/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.289438  [19296/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.321527  [19328/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.239410  [19360/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.341821  [19392/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.217875  [19424/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.257780  [19456/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.227948  [19488/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.290427  [19520/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.229151  [19552/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.237106  [19584/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.274273  [19616/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.256031  [19648/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.406414  [19680/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.323335  [19712/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.358841  [19744/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.206242  [19776/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.247836  [19808/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.259283  [19840/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.284825  [19872/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.197479  [19904/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.259556  [19936/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.312435  [19968/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.269715  [20000/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.293027  [20032/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.300524  [20064/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.247605  [20096/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.271852  [20128/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.352737  [20160/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.447877  [20192/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.349540  [20224/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.258255  [20256/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.266602  [20288/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.326930  [20320/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.333121  [20352/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.294799  [20384/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.331058  [20416/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.148685  [20448/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.235770  [20480/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.350409  [20512/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.291484  [20544/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.200308  [20576/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.253162  [20608/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.367902  [20640/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.344959  [20672/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.421921  [20704/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.265235  [20736/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.274268  [20768/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.254199  [20800/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.301147  [20832/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.318753  [20864/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.402914  [20896/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.236012  [20928/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.188463  [20960/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.405613  [20992/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.389119  [21024/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.215825  [21056/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.196232  [21088/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.350836  [21120/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.291509  [21152/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.357673  [21184/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.279269  [21216/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.353334  [21248/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.385299  [21280/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.459371  [21312/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.246896  [21344/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.315384  [21376/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.249859  [21408/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.340423  [21440/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.224120  [21472/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.120141  [21504/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.332587  [21536/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.332770  [21568/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.447194  [21600/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.244770  [21632/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.220393  [21664/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.250542  [21696/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.239099  [21728/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.208117  [21760/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.226119  [21792/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.143420  [21824/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.231173  [21856/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.193206  [21888/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.303881  [21920/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.316289  [21952/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.285479  [21984/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.238465  [22016/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.265954  [22048/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.177373  [22080/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.277622  [22112/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.299459  [22144/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.229254  [22176/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.193629  [22208/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.316869  [22240/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.223704  [22272/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.233438  [22304/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.220683  [22336/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.222629  [22368/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.382025  [22400/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.305257  [22432/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.183329  [22464/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.209518  [22496/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.321792  [22528/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.189252  [22560/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.296477  [22592/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.227876  [22624/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.340450  [22656/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.219669  [22688/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.408563  [22720/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.237728  [22752/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.221468  [22784/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.262506  [22816/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.258258  [22848/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.254687  [22880/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.239647  [22912/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.312509  [22944/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.391957  [22976/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.216489  [23008/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.278222  [23040/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.225845  [23072/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.333917  [23104/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.257470  [23136/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.333538  [23168/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.294980  [23200/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.347626  [23232/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.295486  [23264/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.192630  [23296/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.164297  [23328/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.258098  [23360/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.301661  [23392/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.268342  [23424/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.306928  [23456/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.294816  [23488/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.362989  [23520/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.199165  [23552/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.207460  [23584/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.253465  [23616/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.225290  [23648/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.271044  [23680/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.300945  [23712/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.266610  [23744/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.203515  [23776/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.288213  [23808/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.326723  [23840/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.328794  [23872/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.260099  [23904/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.198993  [23936/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.307852  [23968/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.292196  [24000/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.279171  [24032/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.248678  [24064/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.206710  [24096/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.367059  [24128/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.307610  [24160/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.398764  [24192/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.238719  [24224/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.351577  [24256/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.213311  [24288/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.212418  [24320/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.379841  [24352/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.265550  [24384/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.332891  [24416/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.299463  [24448/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.212673  [24480/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.336086  [24512/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.222766  [24544/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.223325  [24576/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.339880  [24608/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.246372  [24640/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.219327  [24672/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.244044  [24704/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.231350  [24736/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.232230  [24768/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.270139  [24800/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.244743  [24832/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.229325  [24864/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.366332  [24872/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.366332  [24872/24872]: : 778it [00:27, 28.53it/s]
Epoch 3, time=50.96s

  0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.217217  [   32/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.279478  [   64/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.207960  [   96/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.255569  [  128/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.244372  [  160/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.208116  [  192/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.238783  [  224/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.233315  [  256/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.197558  [  288/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.211068  [  320/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.288196  [  352/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.260026  [  384/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.185930  [  416/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.290058  [  448/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.294698  [  480/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.196058  [  512/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.189994  [  544/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.207079  [  576/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.193521  [  608/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.209787  [  640/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.282214  [  672/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.178556  [  704/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.273472  [  736/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.166020  [  768/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.218943  [  800/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.286689  [  832/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.405545  [  864/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.315617  [  896/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.271154  [  928/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.212656  [  960/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.272834  [  992/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.276140  [ 1024/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.176756  [ 1056/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.266483  [ 1088/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.216419  [ 1120/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.253080  [ 1152/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.251968  [ 1184/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.340709  [ 1216/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.264018  [ 1248/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.222041  [ 1280/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.353549  [ 1312/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.291410  [ 1344/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.274849  [ 1376/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.340715  [ 1408/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.208404  [ 1440/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.310613  [ 1472/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.372317  [ 1504/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.282283  [ 1536/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.280050  [ 1568/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.278206  [ 1600/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.283481  [ 1632/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.280317  [ 1664/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.273735  [ 1696/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.357984  [ 1728/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.293056  [ 1760/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.274504  [ 1792/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.353979  [ 1824/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.197899  [ 1856/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.285417  [ 1888/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.287041  [ 1920/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.361133  [ 1952/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.201643  [ 1984/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.202519  [ 2016/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.175460  [ 2048/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.311814  [ 2080/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.205387  [ 2112/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.319435  [ 2144/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.233668  [ 2176/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.187519  [ 2208/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.240482  [ 2240/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.271024  [ 2272/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.223846  [ 2304/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.327637  [ 2336/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.229551  [ 2368/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.286006  [ 2400/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.270321  [ 2432/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.224321  [ 2464/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.226623  [ 2496/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.281958  [ 2528/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.233297  [ 2560/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.232132  [ 2592/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.223337  [ 2624/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.224128  [ 2656/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.181069  [ 2688/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.237540  [ 2720/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.185689  [ 2752/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.280895  [ 2784/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.231186  [ 2816/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.307837  [ 2848/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.312647  [ 2880/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.237076  [ 2912/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.221551  [ 2944/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.222300  [ 2976/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.207641  [ 3008/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.290275  [ 3040/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.268035  [ 3072/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.304539  [ 3104/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.268142  [ 3136/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.168219  [ 3168/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.297241  [ 3200/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.284635  [ 3232/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.256677  [ 3264/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.346842  [ 3296/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.249269  [ 3328/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.240948  [ 3360/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.188836  [ 3392/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.321852  [ 3424/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.226661  [ 3456/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.342825  [ 3488/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.233850  [ 3520/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.275925  [ 3552/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.189566  [ 3584/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.271427  [ 3616/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.213284  [ 3648/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.171007  [ 3680/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.364701  [ 3712/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.237328  [ 3744/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.199397  [ 3776/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.205422  [ 3808/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.231115  [ 3840/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.251095  [ 3872/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.150117  [ 3904/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.312661  [ 3936/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.252573  [ 3968/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.266538  [ 4000/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.279245  [ 4032/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.192284  [ 4064/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.395832  [ 4096/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.238691  [ 4128/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.309853  [ 4160/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.275895  [ 4192/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.251596  [ 4224/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.336646  [ 4256/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.225675  [ 4288/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.264029  [ 4320/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.252065  [ 4352/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.232878  [ 4384/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.203162  [ 4416/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.372377  [ 4448/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.302214  [ 4480/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.248636  [ 4512/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.317215  [ 4544/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.288828  [ 4576/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.311316  [ 4608/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.321624  [ 4640/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.158259  [ 4672/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.223299  [ 4704/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.357433  [ 4736/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.328316  [ 4768/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.277265  [ 4800/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.303514  [ 4832/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.248368  [ 4864/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.259180  [ 4896/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.301426  [ 4928/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.315670  [ 4960/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.241554  [ 4992/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.253126  [ 5024/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.262337  [ 5056/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.269334  [ 5088/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.341945  [ 5120/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.238960  [ 5152/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.316897  [ 5184/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.285747  [ 5216/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.329307  [ 5248/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.187294  [ 5280/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.280197  [ 5312/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.297475  [ 5344/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.346917  [ 5376/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.249355  [ 5408/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.319892  [ 5440/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.209500  [ 5472/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.206517  [ 5504/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.210503  [ 5536/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.269839  [ 5568/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.227447  [ 5600/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.279757  [ 5632/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.460946  [ 5664/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.249175  [ 5696/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.219532  [ 5728/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.266975  [ 5760/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.291616  [ 5792/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.260802  [ 5824/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.339845  [ 5856/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.187708  [ 5888/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.333118  [ 5920/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.181632  [ 5952/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.175176  [ 5984/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.328283  [ 6016/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.224539  [ 6048/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.219959  [ 6080/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.234120  [ 6112/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.242313  [ 6144/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.228970  [ 6176/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.323672  [ 6208/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.167939  [ 6240/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.227902  [ 6272/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.283777  [ 6304/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.231554  [ 6336/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.260169  [ 6368/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.365526  [ 6400/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.210341  [ 6432/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.195582  [ 6464/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.188053  [ 6496/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.162318  [ 6528/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.157438  [ 6560/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.283115  [ 6592/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.243992  [ 6624/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.205419  [ 6656/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.220544  [ 6688/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.233056  [ 6720/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.287199  [ 6752/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.155832  [ 6784/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.274682  [ 6816/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.177763  [ 6848/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.201839  [ 6880/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.252804  [ 6912/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.202577  [ 6944/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.241326  [ 6976/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.279226  [ 7008/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.223643  [ 7040/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.254260  [ 7072/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.204728  [ 7104/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.229879  [ 7136/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.272030  [ 7168/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.354893  [ 7200/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.241634  [ 7232/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.310004  [ 7264/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.365444  [ 7296/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.194711  [ 7328/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.243233  [ 7360/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.234320  [ 7392/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.318187  [ 7424/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.230692  [ 7456/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.181900  [ 7488/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.209468  [ 7520/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.284450  [ 7552/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.230945  [ 7584/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.250157  [ 7616/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.262189  [ 7648/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.202296  [ 7680/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.385501  [ 7712/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.217157  [ 7744/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.225178  [ 7776/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.257184  [ 7808/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.121772  [ 7840/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.255941  [ 7872/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.294237  [ 7904/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.272404  [ 7936/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.206641  [ 7968/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.276299  [ 8000/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.270516  [ 8032/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.158583  [ 8064/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.214086  [ 8096/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.316251  [ 8128/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.287998  [ 8160/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.347751  [ 8192/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.223101  [ 8224/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.165336  [ 8256/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.166557  [ 8288/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.218807  [ 8320/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.241859  [ 8352/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.202576  [ 8384/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.239749  [ 8416/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.259600  [ 8448/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.213577  [ 8480/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.270172  [ 8512/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.225701  [ 8544/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.265957  [ 8576/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.305357  [ 8608/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.261246  [ 8640/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.346524  [ 8672/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.206757  [ 8704/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.332838  [ 8736/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.272629  [ 8768/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.192309  [ 8800/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.305445  [ 8832/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.256817  [ 8864/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.189174  [ 8896/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.131986  [ 8928/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.279796  [ 8960/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.268655  [ 8992/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.159848  [ 9024/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.213447  [ 9056/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.279101  [ 9088/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.252565  [ 9120/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.256963  [ 9152/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.346902  [ 9184/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.178286  [ 9216/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.272930  [ 9248/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.266515  [ 9280/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.204486  [ 9312/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.276342  [ 9344/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.306348  [ 9376/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.235985  [ 9408/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.223183  [ 9440/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.300867  [ 9472/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.258781  [ 9504/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.267176  [ 9536/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.330512  [ 9568/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.274866  [ 9600/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.274083  [ 9632/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.219582  [ 9664/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.204780  [ 9696/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.224627  [ 9728/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.283444  [ 9760/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.240189  [ 9792/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.178916  [ 9824/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.344389  [ 9856/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.210078  [ 9888/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.236945  [ 9920/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.264983  [ 9952/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.316646  [ 9984/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.243499  [10016/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.223911  [10048/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.189221  [10080/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.187156  [10112/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.270075  [10144/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.280346  [10176/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.230796  [10208/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.217813  [10240/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.315038  [10272/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.262015  [10304/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.213075  [10336/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.285069  [10368/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.318437  [10400/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.230865  [10432/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.239983  [10464/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.271286  [10496/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.267849  [10528/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.307977  [10560/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.297882  [10592/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.226756  [10624/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.207888  [10656/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.265473  [10688/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.271580  [10720/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.353752  [10752/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.198442  [10784/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.309564  [10816/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.306746  [10848/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.245643  [10880/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.294726  [10912/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.328229  [10944/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.165155  [10976/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.200714  [11008/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.223424  [11040/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.268172  [11072/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.236604  [11104/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.344418  [11136/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.372586  [11168/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.230528  [11200/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.256772  [11232/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.214932  [11264/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.279508  [11296/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.212185  [11328/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.280292  [11360/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.210826  [11392/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.200752  [11424/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.363793  [11456/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.181631  [11488/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.286033  [11520/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.308805  [11552/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.301690  [11584/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.346129  [11616/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.318703  [11648/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.280264  [11680/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.302674  [11712/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.315594  [11744/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.306192  [11776/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.217952  [11808/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.188338  [11840/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.226643  [11872/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.191916  [11904/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.234237  [11936/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.280480  [11968/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.247726  [12000/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.277011  [12032/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.255945  [12064/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.346819  [12096/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.293535  [12128/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.222404  [12160/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.287873  [12192/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.361734  [12224/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.325383  [12256/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.244836  [12288/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.340722  [12320/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.333940  [12352/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.189043  [12384/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.189403  [12416/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.268110  [12448/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.157945  [12480/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.233637  [12512/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.249746  [12544/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.257621  [12576/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.302640  [12608/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.206577  [12640/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.263300  [12672/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.321910  [12704/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.214756  [12736/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.240661  [12768/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.328964  [12800/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.254413  [12832/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.270469  [12864/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.282373  [12896/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.299022  [12928/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.293311  [12960/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.334503  [12992/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.188861  [13024/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.287287  [13056/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.275252  [13088/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.321519  [13120/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.219817  [13152/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.201456  [13184/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.220183  [13216/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.275110  [13248/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.300382  [13280/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.257867  [13312/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.198789  [13344/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.277989  [13376/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.290638  [13408/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.257379  [13440/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.201804  [13472/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.258826  [13504/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.161752  [13536/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.271674  [13568/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.243113  [13600/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.272308  [13632/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.198122  [13664/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.270911  [13696/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.195047  [13728/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.236397  [13760/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.212902  [13792/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.353221  [13824/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.358247  [13856/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.228565  [13888/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.349526  [13920/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.244314  [13952/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.320172  [13984/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.305858  [14016/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.225668  [14048/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.329103  [14080/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.196103  [14112/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.294512  [14144/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.261823  [14176/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.257039  [14208/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.305557  [14240/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.369523  [14272/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.304524  [14304/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.238155  [14336/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.305676  [14368/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.246947  [14400/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.200055  [14432/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.294035  [14464/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.187236  [14496/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.208065  [14528/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.185492  [14560/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.212021  [14592/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.234006  [14624/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.150590  [14656/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.233254  [14688/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.143448  [14720/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.310081  [14752/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.217118  [14784/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.328610  [14816/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.201495  [14848/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.242747  [14880/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.232744  [14912/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.336565  [14944/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.282243  [14976/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.313706  [15008/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.245793  [15040/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.225169  [15072/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.248491  [15104/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.297536  [15136/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.174554  [15168/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.321189  [15200/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.180898  [15232/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.215027  [15264/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.221263  [15296/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.268156  [15328/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.205793  [15360/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.215591  [15392/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.227555  [15424/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.186923  [15456/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.178440  [15488/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.200221  [15520/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.243690  [15552/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.225863  [15584/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.189302  [15616/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.198128  [15648/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.261162  [15680/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.238314  [15712/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.286138  [15744/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.203684  [15776/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.182900  [15808/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.262764  [15840/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.242625  [15872/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.136417  [15904/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.257043  [15936/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.309269  [15968/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.122823  [16000/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.201173  [16032/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.210616  [16064/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.242404  [16096/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.242568  [16128/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.191792  [16160/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.295733  [16192/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.294962  [16224/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.215044  [16256/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.203112  [16288/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.134278  [16320/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.235145  [16352/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.289302  [16384/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.284419  [16416/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.205041  [16448/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.200487  [16480/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.336652  [16512/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.374289  [16544/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.306675  [16576/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.181577  [16608/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.188475  [16640/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.322272  [16672/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.279924  [16704/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.341093  [16736/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.221692  [16768/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.342973  [16800/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.275898  [16832/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.199213  [16864/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.244957  [16896/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.284773  [16928/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.251429  [16960/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.225988  [16992/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.219048  [17024/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.318447  [17056/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.334242  [17088/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.146755  [17120/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.245956  [17152/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.278807  [17184/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.219637  [17216/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.267251  [17248/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.261468  [17280/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.300245  [17312/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.249890  [17344/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.305893  [17376/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.229787  [17408/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.334636  [17440/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.186665  [17472/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.306625  [17504/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.239040  [17536/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.214994  [17568/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.227921  [17600/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.231827  [17632/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.264183  [17664/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.227342  [17696/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.208298  [17728/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.226221  [17760/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.272242  [17792/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.229451  [17824/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.252496  [17856/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.220105  [17888/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.230521  [17920/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.264746  [17952/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.196355  [17984/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.197810  [18016/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.277879  [18048/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.188898  [18080/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.213773  [18112/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.201297  [18144/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.216128  [18176/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.167489  [18208/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.190422  [18240/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.250462  [18272/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.193496  [18304/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.279840  [18336/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.155501  [18368/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.202879  [18400/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.224784  [18432/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.258950  [18464/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.212967  [18496/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.272225  [18528/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.271307  [18560/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.180997  [18592/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.198180  [18624/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.259033  [18656/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.232926  [18688/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.210964  [18720/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.236482  [18752/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.291695  [18784/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.226437  [18816/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.376797  [18848/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.156326  [18880/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.228633  [18912/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.213037  [18944/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.272025  [18976/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.273940  [19008/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.175669  [19040/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.304421  [19072/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.241903  [19104/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.284550  [19136/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.239129  [19168/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.144825  [19200/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.219018  [19232/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.172201  [19264/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.214830  [19296/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.269113  [19328/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.234983  [19360/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.299424  [19392/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.202063  [19424/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.224611  [19456/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.214904  [19488/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.262854  [19520/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.229833  [19552/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.180611  [19584/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.245565  [19616/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.236924  [19648/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.357999  [19680/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.282933  [19712/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.307286  [19744/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.179628  [19776/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.225526  [19808/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.242536  [19840/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.254489  [19872/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.183108  [19904/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.225867  [19936/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.273829  [19968/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.222898  [20000/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.259488  [20032/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.214894  [20064/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.219020  [20096/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.250947  [20128/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.316237  [20160/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.355523  [20192/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.260876  [20224/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.232572  [20256/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.260681  [20288/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.277454  [20320/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.299464  [20352/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.230689  [20384/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.261499  [20416/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.121024  [20448/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.257691  [20480/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.312495  [20512/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.257882  [20544/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.174901  [20576/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.192381  [20608/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.312460  [20640/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.316006  [20672/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.372685  [20704/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.210543  [20736/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.235276  [20768/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.215742  [20800/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.268410  [20832/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.221929  [20864/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.346498  [20896/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.230620  [20928/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.159566  [20960/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.336757  [20992/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.304891  [21024/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.196443  [21056/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.163346  [21088/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.295469  [21120/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.268015  [21152/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.235525  [21184/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.248425  [21216/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.290750  [21248/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.317020  [21280/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.387395  [21312/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.221987  [21344/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.240144  [21376/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.227180  [21408/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.296758  [21440/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.201151  [21472/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.109796  [21504/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.283575  [21536/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.294512  [21568/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.372268  [21600/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.258500  [21632/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.237039  [21664/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.200029  [21696/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.207072  [21728/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.185713  [21760/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.219362  [21792/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.121533  [21824/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.213510  [21856/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.167905  [21888/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.293972  [21920/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.274082  [21952/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.250431  [21984/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.213910  [22016/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.241514  [22048/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.150874  [22080/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.170749  [22112/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.274090  [22144/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.160765  [22176/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.214188  [22208/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.256511  [22240/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.202605  [22272/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.204931  [22304/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.180910  [22336/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.183406  [22368/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.329519  [22400/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.250888  [22432/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.168006  [22464/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.187447  [22496/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.283919  [22528/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.153522  [22560/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.271595  [22592/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.221994  [22624/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.273665  [22656/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.219425  [22688/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.318052  [22720/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.189578  [22752/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.208605  [22784/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.257596  [22816/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.232351  [22848/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.250936  [22880/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.222991  [22912/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.240123  [22944/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.360880  [22976/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.221535  [23008/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.216007  [23040/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.195894  [23072/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.287609  [23104/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.245174  [23136/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.269541  [23168/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.228435  [23200/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.301383  [23232/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.246501  [23264/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.162320  [23296/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.139279  [23328/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.237097  [23360/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.277608  [23392/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.230945  [23424/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.268461  [23456/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.304694  [23488/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.313949  [23520/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.208845  [23552/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.174831  [23584/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.243599  [23616/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.208595  [23648/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.216133  [23680/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.326358  [23712/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.243217  [23744/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.164325  [23776/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.233671  [23808/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.307490  [23840/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.292297  [23872/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.250900  [23904/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.189296  [23936/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.234184  [23968/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.237080  [24000/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.275472  [24032/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.230939  [24064/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.194064  [24096/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.345597  [24128/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.278288  [24160/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.373626  [24192/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.219944  [24224/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.317550  [24256/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.186736  [24288/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.188446  [24320/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.348277  [24352/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.262781  [24384/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.315569  [24416/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.276812  [24448/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.172018  [24480/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.317907  [24512/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.196447  [24544/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.176971  [24576/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.302376  [24608/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.221759  [24640/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.201197  [24672/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.204937  [24704/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.206223  [24736/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.218854  [24768/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.249930  [24800/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.226012  [24832/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.217750  [24864/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.303471  [24872/24872]:   0%|          | 0/777 [00:28<?, ?it/s]
loss: 0.303471  [24872/24872]: : 778it [00:28, 27.12it/s]
Epoch 4, time=79.65s

  0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.200849  [   32/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.233449  [   64/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.198935  [   96/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.221935  [  128/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.214121  [  160/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.177508  [  192/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.218048  [  224/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.218643  [  256/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.172192  [  288/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.183321  [  320/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.240527  [  352/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.233054  [  384/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.168839  [  416/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.254364  [  448/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.266548  [  480/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.171642  [  512/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.167308  [  544/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.192132  [  576/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.179325  [  608/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.199332  [  640/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.249925  [  672/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.169094  [  704/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.233067  [  736/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.155718  [  768/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.206263  [  800/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.264266  [  832/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.337279  [  864/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.307588  [  896/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.244130  [  928/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.196865  [  960/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.255677  [  992/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.250881  [ 1024/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.163027  [ 1056/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.213193  [ 1088/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.189012  [ 1120/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.218658  [ 1152/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.193555  [ 1184/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.306606  [ 1216/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.212946  [ 1248/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.199993  [ 1280/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.325255  [ 1312/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.261114  [ 1344/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.269159  [ 1376/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.327432  [ 1408/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.186904  [ 1440/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.304826  [ 1472/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.346829  [ 1504/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.253532  [ 1536/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.249544  [ 1568/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.262725  [ 1600/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.247464  [ 1632/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.265527  [ 1664/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.237542  [ 1696/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.330999  [ 1728/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.247886  [ 1760/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.253361  [ 1792/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.296792  [ 1824/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.175184  [ 1856/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.245827  [ 1888/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.263591  [ 1920/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.307229  [ 1952/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.163782  [ 1984/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.176261  [ 2016/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.154430  [ 2048/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.237659  [ 2080/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.196130  [ 2112/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.285196  [ 2144/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.201483  [ 2176/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.172958  [ 2208/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.216632  [ 2240/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.212499  [ 2272/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.202386  [ 2304/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.263998  [ 2336/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.190964  [ 2368/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.252233  [ 2400/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.248716  [ 2432/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.200365  [ 2464/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.184131  [ 2496/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.243053  [ 2528/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.210454  [ 2560/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.208927  [ 2592/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.209715  [ 2624/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.210578  [ 2656/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.155522  [ 2688/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.202578  [ 2720/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.151268  [ 2752/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.242737  [ 2784/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.199460  [ 2816/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.263520  [ 2848/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.301291  [ 2880/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.218192  [ 2912/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.171536  [ 2944/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.192586  [ 2976/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.185798  [ 3008/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.246724  [ 3040/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.233638  [ 3072/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.310065  [ 3104/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.236710  [ 3136/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.162627  [ 3168/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.320577  [ 3200/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.264747  [ 3232/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.261521  [ 3264/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.314336  [ 3296/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.235301  [ 3328/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.226738  [ 3360/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.171098  [ 3392/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.275127  [ 3424/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.198668  [ 3456/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.287317  [ 3488/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.207792  [ 3520/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.257838  [ 3552/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.183732  [ 3584/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.235130  [ 3616/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.186811  [ 3648/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.152737  [ 3680/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.304497  [ 3712/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.214954  [ 3744/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.175429  [ 3776/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.169895  [ 3808/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.199674  [ 3840/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.220289  [ 3872/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.130297  [ 3904/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.266513  [ 3936/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.235627  [ 3968/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.242149  [ 4000/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.231766  [ 4032/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.173053  [ 4064/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.376670  [ 4096/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.231808  [ 4128/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.294637  [ 4160/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.235585  [ 4192/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.247711  [ 4224/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.325684  [ 4256/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.203200  [ 4288/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.242336  [ 4320/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.243244  [ 4352/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.212803  [ 4384/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.204087  [ 4416/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.353566  [ 4448/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.217879  [ 4480/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.225098  [ 4512/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.296321  [ 4544/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.280376  [ 4576/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.256136  [ 4608/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.298693  [ 4640/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.143977  [ 4672/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.205069  [ 4704/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.312964  [ 4736/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.299695  [ 4768/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.258864  [ 4800/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.279167  [ 4832/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.215223  [ 4864/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.235616  [ 4896/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.263957  [ 4928/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.288545  [ 4960/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.211513  [ 4992/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.231592  [ 5024/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.211901  [ 5056/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.239634  [ 5088/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.301873  [ 5120/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.220818  [ 5152/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.301031  [ 5184/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.254698  [ 5216/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.269220  [ 5248/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.157466  [ 5280/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.245064  [ 5312/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.243295  [ 5344/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.297228  [ 5376/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.225790  [ 5408/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.270487  [ 5440/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.190642  [ 5472/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.184259  [ 5504/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.190405  [ 5536/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.248416  [ 5568/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.201362  [ 5600/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.260738  [ 5632/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.403368  [ 5664/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.224104  [ 5696/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.204059  [ 5728/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.232066  [ 5760/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.280268  [ 5792/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.217129  [ 5824/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.276096  [ 5856/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.166328  [ 5888/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.303189  [ 5920/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.165821  [ 5952/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.161480  [ 5984/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.266502  [ 6016/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.211018  [ 6048/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.207345  [ 6080/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.239037  [ 6112/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.242695  [ 6144/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.216604  [ 6176/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.358404  [ 6208/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.147307  [ 6240/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.175906  [ 6272/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.279814  [ 6304/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.236553  [ 6336/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.222789  [ 6368/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.335502  [ 6400/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.207719  [ 6432/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.175698  [ 6464/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.174930  [ 6496/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.131415  [ 6528/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.143960  [ 6560/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.272931  [ 6592/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.232490  [ 6624/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.179140  [ 6656/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.207513  [ 6688/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.214269  [ 6720/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.260297  [ 6752/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.143807  [ 6784/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.257552  [ 6816/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.157998  [ 6848/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.186820  [ 6880/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.220533  [ 6912/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.194671  [ 6944/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.219334  [ 6976/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.264736  [ 7008/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.214976  [ 7040/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.231547  [ 7072/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.200858  [ 7104/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.210958  [ 7136/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.250152  [ 7168/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.321171  [ 7200/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.227344  [ 7232/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.295234  [ 7264/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.304459  [ 7296/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.173745  [ 7328/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.184127  [ 7360/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.211650  [ 7392/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.285460  [ 7424/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.191489  [ 7456/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.169151  [ 7488/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.195620  [ 7520/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.231761  [ 7552/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.175317  [ 7584/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.220263  [ 7616/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.259536  [ 7648/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.165338  [ 7680/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.312012  [ 7712/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.192998  [ 7744/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.202089  [ 7776/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.239037  [ 7808/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.118441  [ 7840/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.243528  [ 7872/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.290257  [ 7904/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.257751  [ 7936/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.180324  [ 7968/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.230609  [ 8000/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.238693  [ 8032/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.141712  [ 8064/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.188336  [ 8096/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.294494  [ 8128/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.240408  [ 8160/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.309085  [ 8192/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.204201  [ 8224/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.143817  [ 8256/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.139166  [ 8288/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.201627  [ 8320/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.214129  [ 8352/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.193593  [ 8384/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.193488  [ 8416/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.237367  [ 8448/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.201686  [ 8480/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.204479  [ 8512/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.201941  [ 8544/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.264719  [ 8576/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.260643  [ 8608/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.246153  [ 8640/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.344323  [ 8672/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.192952  [ 8704/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.315292  [ 8736/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.267593  [ 8768/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.190957  [ 8800/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.284299  [ 8832/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.192761  [ 8864/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.184056  [ 8896/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.130181  [ 8928/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.251462  [ 8960/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.254642  [ 8992/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.148265  [ 9024/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.197345  [ 9056/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.257616  [ 9088/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.214246  [ 9120/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.233172  [ 9152/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.307125  [ 9184/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.175047  [ 9216/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.248434  [ 9248/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.248297  [ 9280/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.215183  [ 9312/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.253230  [ 9344/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.311256  [ 9376/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.235156  [ 9408/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.218151  [ 9440/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.279557  [ 9472/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.245985  [ 9504/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.247207  [ 9536/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.304969  [ 9568/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.255385  [ 9600/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.238567  [ 9632/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.196933  [ 9664/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.192361  [ 9696/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.199136  [ 9728/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.293953  [ 9760/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.241129  [ 9792/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.168855  [ 9824/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.317648  [ 9856/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.178820  [ 9888/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.213321  [ 9920/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.248240  [ 9952/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.296013  [ 9984/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.226437  [10016/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.237015  [10048/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.177744  [10080/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.172150  [10112/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.267966  [10144/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.284422  [10176/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.227166  [10208/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.183579  [10240/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.317411  [10272/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.239331  [10304/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.195899  [10336/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.272541  [10368/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.306374  [10400/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.218659  [10432/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.227516  [10464/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.253001  [10496/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.253723  [10528/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.286363  [10560/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.297871  [10592/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.216251  [10624/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.205100  [10656/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.298181  [10688/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.242638  [10720/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.305680  [10752/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.187584  [10784/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.286402  [10816/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.304972  [10848/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.242639  [10880/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.273026  [10912/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.279812  [10944/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.145360  [10976/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.206901  [11008/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.208253  [11040/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.259390  [11072/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.226106  [11104/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.319646  [11136/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.348603  [11168/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.209846  [11200/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.245734  [11232/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.195587  [11264/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.259756  [11296/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.194326  [11328/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.280074  [11360/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.187736  [11392/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.183295  [11424/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.314942  [11456/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.165113  [11488/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.233322  [11520/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.237170  [11552/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.242133  [11584/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.276167  [11616/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.259910  [11648/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.235400  [11680/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.263419  [11712/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.272807  [11744/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.243675  [11776/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.190652  [11808/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.165134  [11840/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.176303  [11872/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.182798  [11904/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.194041  [11936/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.251733  [11968/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.229936  [12000/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.256622  [12032/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.237886  [12064/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.291025  [12096/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.278463  [12128/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.221915  [12160/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.258656  [12192/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.329376  [12224/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.310772  [12256/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.207335  [12288/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.319495  [12320/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.318725  [12352/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.179644  [12384/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.177588  [12416/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.245344  [12448/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.148056  [12480/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.209900  [12512/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.221779  [12544/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.264147  [12576/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.268475  [12608/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.210873  [12640/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.271680  [12672/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.308060  [12704/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.216380  [12736/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.231299  [12768/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.322019  [12800/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.236444  [12832/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.241626  [12864/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.265215  [12896/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.270276  [12928/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.288589  [12960/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.309274  [12992/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.162968  [13024/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.273148  [13056/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.251017  [13088/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.304562  [13120/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.212407  [13152/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.181357  [13184/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.198764  [13216/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.260083  [13248/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.293724  [13280/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.261503  [13312/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.168330  [13344/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.262940  [13376/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.259653  [13408/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.250384  [13440/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.193021  [13472/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.235074  [13504/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.150342  [13536/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.261690  [13568/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.229130  [13600/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.249900  [13632/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.173497  [13664/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.242212  [13696/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.179032  [13728/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.190956  [13760/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.196307  [13792/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.320053  [13824/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.335022  [13856/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.210867  [13888/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.311454  [13920/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.210369  [13952/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.266430  [13984/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.289122  [14016/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.183576  [14048/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.257964  [14080/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.174375  [14112/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.212602  [14144/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.212240  [14176/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.202490  [14208/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.275440  [14240/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.281714  [14272/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.292310  [14304/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.219664  [14336/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.288207  [14368/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.221625  [14400/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.191055  [14432/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.244222  [14464/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.175056  [14496/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.201293  [14528/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.158978  [14560/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.172453  [14592/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.227156  [14624/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.179945  [14656/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.193348  [14688/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.122488  [14720/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.297512  [14752/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.191225  [14784/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.316810  [14816/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.184367  [14848/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.202981  [14880/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.249293  [14912/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.316760  [14944/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.206064  [14976/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.270340  [15008/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.224850  [15040/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.205502  [15072/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.237583  [15104/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.286942  [15136/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.142548  [15168/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.297434  [15200/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.165183  [15232/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.185624  [15264/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.169354  [15296/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.252572  [15328/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.205154  [15360/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.197631  [15392/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.204052  [15424/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.167089  [15456/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.169733  [15488/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.186189  [15520/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.225387  [15552/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.219882  [15584/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.172119  [15616/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.202020  [15648/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.250108  [15680/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.227551  [15712/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.271204  [15744/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.186675  [15776/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.173174  [15808/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.251165  [15840/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.207934  [15872/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.119277  [15904/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.244352  [15936/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.287970  [15968/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.116903  [16000/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.183254  [16032/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.207038  [16064/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.230639  [16096/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.255526  [16128/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.154765  [16160/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.259031  [16192/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.298019  [16224/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.196197  [16256/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.179945  [16288/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.136557  [16320/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.212674  [16352/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.246472  [16384/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.257301  [16416/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.201333  [16448/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.187879  [16480/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.315423  [16512/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.355292  [16544/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.263940  [16576/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.150449  [16608/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.168348  [16640/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.284601  [16672/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.268196  [16704/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.262444  [16736/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.166568  [16768/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.297747  [16800/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.282101  [16832/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.176778  [16864/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.209296  [16896/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.259723  [16928/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.220307  [16960/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.216364  [16992/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.199049  [17024/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.305127  [17056/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.328737  [17088/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.136719  [17120/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.246574  [17152/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.290150  [17184/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.193247  [17216/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.286676  [17248/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.267760  [17280/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.260231  [17312/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.276825  [17344/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.313541  [17376/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.207370  [17408/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.311862  [17440/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.191663  [17472/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.293422  [17504/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.229744  [17536/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.223932  [17568/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.218279  [17600/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.222187  [17632/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.234430  [17664/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.208925  [17696/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.198596  [17728/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.213245  [17760/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.258001  [17792/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.207931  [17824/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.217642  [17856/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.207328  [17888/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.218745  [17920/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.232308  [17952/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.199083  [17984/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.187030  [18016/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.246006  [18048/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.181718  [18080/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.194817  [18112/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.191937  [18144/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.201405  [18176/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.148444  [18208/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.180456  [18240/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.252064  [18272/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.170262  [18304/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.253573  [18336/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.143670  [18368/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.186022  [18400/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.213220  [18432/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.234278  [18464/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.194058  [18496/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.243284  [18528/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.254046  [18560/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.155603  [18592/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.175035  [18624/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.245556  [18656/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.219923  [18688/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.195114  [18720/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.213598  [18752/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.257273  [18784/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.201784  [18816/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.360535  [18848/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.144855  [18880/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.206042  [18912/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.196215  [18944/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.264228  [18976/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.256817  [19008/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.142733  [19040/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.265897  [19072/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.217257  [19104/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.248474  [19136/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.206720  [19168/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.133800  [19200/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.179118  [19232/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.154796  [19264/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.186998  [19296/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.250791  [19328/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.203453  [19360/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.303169  [19392/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.208138  [19424/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.228873  [19456/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.202030  [19488/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.246672  [19520/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.209071  [19552/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.183621  [19584/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.247348  [19616/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.223998  [19648/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.346067  [19680/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.257071  [19712/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.291692  [19744/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.162067  [19776/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.220881  [19808/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.244728  [19840/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.242650  [19872/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.161250  [19904/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.209816  [19936/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.264227  [19968/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.212889  [20000/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.232696  [20032/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.203887  [20064/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.216173  [20096/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.226087  [20128/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.253706  [20160/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.323651  [20192/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.267125  [20224/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.230043  [20256/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.251653  [20288/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.264735  [20320/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.253904  [20352/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.211899  [20384/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.242929  [20416/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.112845  [20448/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.237351  [20480/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.293909  [20512/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.247029  [20544/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.161890  [20576/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.171206  [20608/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.301222  [20640/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.297611  [20672/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.382526  [20704/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.209968  [20736/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.174451  [20768/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.199275  [20800/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.233054  [20832/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.227298  [20864/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.305199  [20896/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.211955  [20928/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.150668  [20960/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.309928  [20992/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.274605  [21024/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.223214  [21056/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.159890  [21088/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.264550  [21120/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.247572  [21152/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.217926  [21184/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.195752  [21216/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.296498  [21248/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.247898  [21280/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.335001  [21312/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.207438  [21344/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.225017  [21376/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.209470  [21408/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.283213  [21440/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.172524  [21472/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.103496  [21504/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.257505  [21536/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.259313  [21568/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.329437  [21600/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.215593  [21632/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.204675  [21664/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.190130  [21696/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.192135  [21728/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.170842  [21760/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.210895  [21792/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.119101  [21824/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.182206  [21856/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.153994  [21888/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.308031  [21920/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.236768  [21952/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.234208  [21984/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.206438  [22016/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.217386  [22048/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.151485  [22080/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.160563  [22112/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.260683  [22144/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.151033  [22176/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.170608  [22208/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.242470  [22240/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.196416  [22272/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.202604  [22304/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.167821  [22336/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.169290  [22368/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.275691  [22400/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.256701  [22432/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.152155  [22464/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.168883  [22496/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.280194  [22528/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.158824  [22560/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.229305  [22592/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.211666  [22624/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.266372  [22656/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.192812  [22688/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.340899  [22720/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.182696  [22752/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.172376  [22784/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.226605  [22816/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.204394  [22848/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.196335  [22880/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.200377  [22912/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.225984  [22944/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.316582  [22976/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.188807  [23008/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.198443  [23040/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.224524  [23072/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.255271  [23104/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.217479  [23136/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.253690  [23168/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.221085  [23200/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.272082  [23232/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.252164  [23264/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.160450  [23296/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.153467  [23328/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.227082  [23360/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.244535  [23392/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.253188  [23424/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.252532  [23456/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.274377  [23488/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.305713  [23520/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.185662  [23552/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.157605  [23584/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.210714  [23616/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.198797  [23648/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.215825  [23680/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.284015  [23712/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.234203  [23744/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.174847  [23776/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.216299  [23808/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.310165  [23840/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.292899  [23872/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.245820  [23904/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.157103  [23936/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.235726  [23968/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.244888  [24000/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.232290  [24032/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.215153  [24064/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.194867  [24096/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.318094  [24128/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.259636  [24160/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.364177  [24192/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.222746  [24224/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.268226  [24256/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.205687  [24288/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.175886  [24320/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.312228  [24352/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.250400  [24384/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.296747  [24416/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.239495  [24448/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.183905  [24480/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.288291  [24512/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.204730  [24544/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.174506  [24576/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.290803  [24608/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.206497  [24640/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.169853  [24672/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.195470  [24704/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.196613  [24736/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.204682  [24768/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.306961  [24800/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.217185  [24832/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.214905  [24864/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.274584  [24872/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.274584  [24872/24872]: : 778it [00:27, 27.87it/s]
Epoch 5, time=107.57s

  0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.180904  [   32/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.207688  [   64/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.181363  [   96/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.229516  [  128/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.190957  [  160/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.179499  [  192/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.180238  [  224/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.190214  [  256/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.186773  [  288/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.150284  [  320/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.224485  [  352/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.245334  [  384/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.139340  [  416/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.212345  [  448/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.242613  [  480/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.168356  [  512/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.159212  [  544/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.179059  [  576/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.169162  [  608/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.193467  [  640/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.237292  [  672/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.158023  [  704/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.240065  [  736/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.134358  [  768/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.208873  [  800/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.223995  [  832/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.278159  [  864/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.274275  [  896/24872]:   0%|          | 0/777 [00:00<?, ?it/s]
loss: 0.213584  [  928/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.197403  [  960/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.264146  [  992/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.227404  [ 1024/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.145637  [ 1056/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.191166  [ 1088/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.158037  [ 1120/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.202667  [ 1152/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.174303  [ 1184/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.291339  [ 1216/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.190729  [ 1248/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.190071  [ 1280/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.337380  [ 1312/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.244730  [ 1344/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.259400  [ 1376/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.271682  [ 1408/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.188039  [ 1440/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.308078  [ 1472/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.318354  [ 1504/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.242329  [ 1536/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.254492  [ 1568/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.228093  [ 1600/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.237392  [ 1632/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.237720  [ 1664/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.224904  [ 1696/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.304107  [ 1728/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.256774  [ 1760/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.252842  [ 1792/24872]:   0%|          | 0/777 [00:01<?, ?it/s]
loss: 0.274491  [ 1824/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.217631  [ 1856/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.216413  [ 1888/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.234360  [ 1920/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.271590  [ 1952/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.154930  [ 1984/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.178675  [ 2016/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.157059  [ 2048/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.197620  [ 2080/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.183334  [ 2112/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.286653  [ 2144/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.175484  [ 2176/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.155126  [ 2208/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.204651  [ 2240/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.158991  [ 2272/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.183352  [ 2304/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.245525  [ 2336/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.172677  [ 2368/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.222944  [ 2400/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.240867  [ 2432/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.185039  [ 2464/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.156647  [ 2496/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.225295  [ 2528/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.201820  [ 2560/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.222223  [ 2592/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.207511  [ 2624/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.179699  [ 2656/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.149020  [ 2688/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.187067  [ 2720/24872]:   0%|          | 0/777 [00:02<?, ?it/s]
loss: 0.144006  [ 2752/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.215568  [ 2784/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.182476  [ 2816/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.248916  [ 2848/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.270014  [ 2880/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.163672  [ 2912/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.148243  [ 2944/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.147743  [ 2976/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.164899  [ 3008/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.190925  [ 3040/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.192707  [ 3072/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.261086  [ 3104/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.217898  [ 3136/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.140763  [ 3168/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.238730  [ 3200/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.211227  [ 3232/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.227899  [ 3264/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.304858  [ 3296/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.242780  [ 3328/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.208946  [ 3360/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.178334  [ 3392/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.278613  [ 3424/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.176598  [ 3456/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.266891  [ 3488/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.183455  [ 3520/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.222091  [ 3552/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.163962  [ 3584/24872]:   0%|          | 0/777 [00:03<?, ?it/s]
loss: 0.219088  [ 3616/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.172630  [ 3648/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.149579  [ 3680/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.285019  [ 3712/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.197763  [ 3744/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.170418  [ 3776/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.156555  [ 3808/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.182488  [ 3840/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.197319  [ 3872/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.119146  [ 3904/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.232708  [ 3936/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.231778  [ 3968/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.237055  [ 4000/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.241292  [ 4032/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.171250  [ 4064/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.363329  [ 4096/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.205621  [ 4128/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.289749  [ 4160/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.201994  [ 4192/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.233883  [ 4224/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.282788  [ 4256/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.161176  [ 4288/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.159964  [ 4320/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.210615  [ 4352/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.162328  [ 4384/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.156033  [ 4416/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.268820  [ 4448/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.195629  [ 4480/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.197903  [ 4512/24872]:   0%|          | 0/777 [00:04<?, ?it/s]
loss: 0.249899  [ 4544/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.268661  [ 4576/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.210764  [ 4608/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.250234  [ 4640/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.132958  [ 4672/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.185887  [ 4704/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.295601  [ 4736/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.276767  [ 4768/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.240780  [ 4800/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.265566  [ 4832/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.197146  [ 4864/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.208845  [ 4896/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.247545  [ 4928/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.271101  [ 4960/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.183526  [ 4992/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.232430  [ 5024/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.191749  [ 5056/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.233957  [ 5088/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.288806  [ 5120/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.210466  [ 5152/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.298410  [ 5184/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.223284  [ 5216/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.267839  [ 5248/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.154249  [ 5280/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.228698  [ 5312/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.236523  [ 5344/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.265914  [ 5376/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.198058  [ 5408/24872]:   0%|          | 0/777 [00:05<?, ?it/s]
loss: 0.252791  [ 5440/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.190622  [ 5472/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.148789  [ 5504/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.181644  [ 5536/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.242265  [ 5568/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.174552  [ 5600/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.238280  [ 5632/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.377774  [ 5664/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.206450  [ 5696/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.193944  [ 5728/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.215532  [ 5760/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.255839  [ 5792/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.186068  [ 5824/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.265180  [ 5856/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.164463  [ 5888/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.286194  [ 5920/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.167009  [ 5952/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.159584  [ 5984/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.242942  [ 6016/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.187573  [ 6048/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.206837  [ 6080/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.208675  [ 6112/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.205047  [ 6144/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.232744  [ 6176/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.330888  [ 6208/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.146856  [ 6240/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.165278  [ 6272/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.265988  [ 6304/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.183369  [ 6336/24872]:   0%|          | 0/777 [00:06<?, ?it/s]
loss: 0.200734  [ 6368/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.318641  [ 6400/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.202323  [ 6432/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.150558  [ 6464/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.169035  [ 6496/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.134016  [ 6528/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.146895  [ 6560/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.263249  [ 6592/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.218672  [ 6624/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.163137  [ 6656/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.193378  [ 6688/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.205916  [ 6720/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.258657  [ 6752/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.134350  [ 6784/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.235656  [ 6816/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.149091  [ 6848/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.162742  [ 6880/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.201473  [ 6912/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.184980  [ 6944/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.198812  [ 6976/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.251470  [ 7008/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.199277  [ 7040/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.210298  [ 7072/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.192347  [ 7104/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.201935  [ 7136/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.225166  [ 7168/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.293181  [ 7200/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.214230  [ 7232/24872]:   0%|          | 0/777 [00:07<?, ?it/s]
loss: 0.275031  [ 7264/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.289379  [ 7296/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.162395  [ 7328/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.179030  [ 7360/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.203741  [ 7392/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.257360  [ 7424/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.182530  [ 7456/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.160728  [ 7488/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.184467  [ 7520/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.202479  [ 7552/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.161317  [ 7584/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.194006  [ 7616/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.247934  [ 7648/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.145752  [ 7680/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.271819  [ 7712/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.185019  [ 7744/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.190545  [ 7776/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.207576  [ 7808/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.118176  [ 7840/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.219083  [ 7872/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.263090  [ 7904/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.233096  [ 7936/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.161957  [ 7968/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.217176  [ 8000/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.222061  [ 8032/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.132942  [ 8064/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.181075  [ 8096/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.263591  [ 8128/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.212283  [ 8160/24872]:   0%|          | 0/777 [00:08<?, ?it/s]
loss: 0.271564  [ 8192/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.187197  [ 8224/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.123866  [ 8256/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.126403  [ 8288/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.182022  [ 8320/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.207193  [ 8352/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.178285  [ 8384/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.187715  [ 8416/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.198741  [ 8448/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.190608  [ 8480/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.188816  [ 8512/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.177550  [ 8544/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.227233  [ 8576/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.231949  [ 8608/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.243922  [ 8640/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.304125  [ 8672/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.173123  [ 8704/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.279148  [ 8736/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.277060  [ 8768/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.154640  [ 8800/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.264904  [ 8832/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.196345  [ 8864/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.158721  [ 8896/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.096645  [ 8928/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.249159  [ 8960/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.196996  [ 8992/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.135388  [ 9024/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.186050  [ 9056/24872]:   0%|          | 0/777 [00:09<?, ?it/s]
loss: 0.225079  [ 9088/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.185089  [ 9120/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.232617  [ 9152/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.273100  [ 9184/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.154521  [ 9216/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.221827  [ 9248/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.208126  [ 9280/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.176932  [ 9312/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.230548  [ 9344/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.307421  [ 9376/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.255779  [ 9408/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.217537  [ 9440/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.260366  [ 9472/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.180895  [ 9504/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.218145  [ 9536/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.266430  [ 9568/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.238148  [ 9600/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.208573  [ 9632/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.173933  [ 9664/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.167561  [ 9696/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.180224  [ 9728/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.240070  [ 9760/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.194868  [ 9792/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.163036  [ 9824/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.262393  [ 9856/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.165250  [ 9888/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.198520  [ 9920/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.214123  [ 9952/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.262796  [ 9984/24872]:   0%|          | 0/777 [00:10<?, ?it/s]
loss: 0.206409  [10016/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.195055  [10048/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.166791  [10080/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.164880  [10112/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.237523  [10144/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.247863  [10176/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.203903  [10208/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.164778  [10240/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.275721  [10272/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.201265  [10304/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.191983  [10336/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.266124  [10368/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.266352  [10400/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.194347  [10432/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.210505  [10464/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.205910  [10496/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.228160  [10528/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.271449  [10560/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.251838  [10592/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.195575  [10624/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.173830  [10656/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.223753  [10688/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.223034  [10720/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.259938  [10752/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.170285  [10784/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.256329  [10816/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.283039  [10848/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.233449  [10880/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.248098  [10912/24872]:   0%|          | 0/777 [00:11<?, ?it/s]
loss: 0.237676  [10944/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.136226  [10976/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.176094  [11008/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.188913  [11040/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.223921  [11072/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.220954  [11104/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.270398  [11136/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.312221  [11168/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.206372  [11200/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.221594  [11232/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.202386  [11264/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.222108  [11296/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.184445  [11328/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.253260  [11360/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.149745  [11392/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.173510  [11424/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.253495  [11456/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.147389  [11488/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.185048  [11520/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.203252  [11552/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.209089  [11584/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.203907  [11616/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.206117  [11648/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.225751  [11680/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.239809  [11712/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.238001  [11744/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.212835  [11776/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.191173  [11808/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.154268  [11840/24872]:   0%|          | 0/777 [00:12<?, ?it/s]
loss: 0.153403  [11872/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.162687  [11904/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.163629  [11936/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.218929  [11968/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.245176  [12000/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.234439  [12032/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.213679  [12064/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.249439  [12096/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.263020  [12128/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.168849  [12160/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.213397  [12192/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.286409  [12224/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.261579  [12256/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.182915  [12288/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.276938  [12320/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.250322  [12352/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.135038  [12384/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.148335  [12416/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.213264  [12448/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.147109  [12480/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.181616  [12512/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.183742  [12544/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.230732  [12576/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.222545  [12608/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.201299  [12640/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.240329  [12672/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.249704  [12704/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.208616  [12736/24872]:   0%|          | 0/777 [00:13<?, ?it/s]
loss: 0.217652  [12768/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.300338  [12800/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.206120  [12832/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.215136  [12864/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.266372  [12896/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.237743  [12928/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.233292  [12960/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.257896  [12992/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.155413  [13024/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.245664  [13056/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.198304  [13088/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.253375  [13120/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.262765  [13152/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.161910  [13184/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.176332  [13216/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.265365  [13248/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.246100  [13280/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.237440  [13312/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.180470  [13344/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.232206  [13376/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.220182  [13408/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.250975  [13440/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.160427  [13472/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.221275  [13504/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.149516  [13536/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.227185  [13568/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.191637  [13600/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.199148  [13632/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.171629  [13664/24872]:   0%|          | 0/777 [00:14<?, ?it/s]
loss: 0.214485  [13696/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.156713  [13728/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.181140  [13760/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.167135  [13792/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.275954  [13824/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.263977  [13856/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.196793  [13888/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.260196  [13920/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.204717  [13952/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.232668  [13984/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.219383  [14016/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.145128  [14048/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.180687  [14080/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.135839  [14112/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.150661  [14144/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.180473  [14176/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.195450  [14208/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.231211  [14240/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.255640  [14272/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.247629  [14304/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.182664  [14336/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.232919  [14368/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.197712  [14400/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.163607  [14432/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.251507  [14464/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.156373  [14496/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.187550  [14528/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.146892  [14560/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.147151  [14592/24872]:   0%|          | 0/777 [00:15<?, ?it/s]
loss: 0.222569  [14624/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.139863  [14656/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.190823  [14688/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.117480  [14720/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.232262  [14752/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.175868  [14784/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.272116  [14816/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.161266  [14848/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.165357  [14880/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.183546  [14912/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.276966  [14944/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.228330  [14976/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.283856  [15008/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.200989  [15040/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.187143  [15072/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.222800  [15104/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.284233  [15136/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.147108  [15168/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.247113  [15200/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.155088  [15232/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.183426  [15264/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.180524  [15296/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.231971  [15328/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.186428  [15360/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.201052  [15392/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.177107  [15424/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.166287  [15456/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.153625  [15488/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.132642  [15520/24872]:   0%|          | 0/777 [00:16<?, ?it/s]
loss: 0.202398  [15552/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.195422  [15584/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.143048  [15616/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.196718  [15648/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.229707  [15680/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.196851  [15712/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.268939  [15744/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.160893  [15776/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.164740  [15808/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.322417  [15840/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.167116  [15872/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.110998  [15904/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.227010  [15936/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.249756  [15968/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.098740  [16000/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.168496  [16032/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.196058  [16064/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.207596  [16096/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.228395  [16128/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.151504  [16160/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.195847  [16192/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.257921  [16224/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.179481  [16256/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.141226  [16288/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.102846  [16320/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.203502  [16352/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.218595  [16384/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.228193  [16416/24872]:   0%|          | 0/777 [00:17<?, ?it/s]
loss: 0.172980  [16448/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.160485  [16480/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.296215  [16512/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.313789  [16544/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.234415  [16576/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.119447  [16608/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.173225  [16640/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.265472  [16672/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.230175  [16704/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.254231  [16736/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.171771  [16768/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.269635  [16800/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.227454  [16832/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.176790  [16864/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.197522  [16896/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.259911  [16928/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.204796  [16960/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.217983  [16992/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.218940  [17024/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.256244  [17056/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.250899  [17088/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.132306  [17120/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.225789  [17152/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.276086  [17184/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.172863  [17216/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.237895  [17248/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.228169  [17280/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.228350  [17312/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.220376  [17344/24872]:   0%|          | 0/777 [00:18<?, ?it/s]
loss: 0.271298  [17376/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.199529  [17408/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.271120  [17440/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.153601  [17472/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.249898  [17504/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.216520  [17536/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.201256  [17568/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.215570  [17600/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.214268  [17632/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.205110  [17664/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.199824  [17696/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.197858  [17728/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.207500  [17760/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.266560  [17792/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.218829  [17824/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.198482  [17856/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.180975  [17888/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.185760  [17920/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.243436  [17952/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.198691  [17984/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.158667  [18016/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.231774  [18048/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.164189  [18080/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.194587  [18112/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.175122  [18144/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.171269  [18176/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.113047  [18208/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.180497  [18240/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.196530  [18272/24872]:   0%|          | 0/777 [00:19<?, ?it/s]
loss: 0.167704  [18304/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.183031  [18336/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.119043  [18368/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.175386  [18400/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.204549  [18432/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.226781  [18464/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.194950  [18496/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.224068  [18528/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.242369  [18560/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.137056  [18592/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.183030  [18624/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.220772  [18656/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.215527  [18688/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.190716  [18720/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.212119  [18752/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.253423  [18784/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.191693  [18816/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.284908  [18848/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.139894  [18880/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.181657  [18912/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.211419  [18944/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.250775  [18976/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.254166  [19008/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.127806  [19040/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.259500  [19072/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.219030  [19104/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.211158  [19136/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.176251  [19168/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.123476  [19200/24872]:   0%|          | 0/777 [00:20<?, ?it/s]
loss: 0.135804  [19232/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.153580  [19264/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.179249  [19296/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.207126  [19328/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.205995  [19360/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.257602  [19392/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.162843  [19424/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.189989  [19456/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.188877  [19488/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.203182  [19520/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.210447  [19552/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.148575  [19584/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.213725  [19616/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.233575  [19648/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.308482  [19680/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.234121  [19712/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.263830  [19744/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.131101  [19776/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.211418  [19808/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.210592  [19840/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.202946  [19872/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.172458  [19904/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.186940  [19936/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.246586  [19968/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.200809  [20000/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.244896  [20032/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.169257  [20064/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.175329  [20096/24872]:   0%|          | 0/777 [00:21<?, ?it/s]
loss: 0.211474  [20128/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.248538  [20160/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.292093  [20192/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.241652  [20224/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.250830  [20256/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.215238  [20288/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.264588  [20320/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.261494  [20352/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.232992  [20384/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.225435  [20416/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.142135  [20448/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.163149  [20480/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.291526  [20512/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.365161  [20544/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.153102  [20576/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.192260  [20608/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.284529  [20640/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.277883  [20672/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.515883  [20704/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.186873  [20736/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.148363  [20768/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.205283  [20800/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.231183  [20832/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.180285  [20864/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.266179  [20896/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.146067  [20928/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.135142  [20960/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.334926  [20992/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.276746  [21024/24872]:   0%|          | 0/777 [00:22<?, ?it/s]
loss: 0.174084  [21056/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.208714  [21088/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.297834  [21120/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.233324  [21152/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.273523  [21184/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.165888  [21216/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.288424  [21248/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.234533  [21280/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.326264  [21312/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.180867  [21344/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.210081  [21376/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.229376  [21408/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.288812  [21440/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.246992  [21472/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.134392  [21504/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.243366  [21536/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.266015  [21568/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.338612  [21600/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.231045  [21632/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.214230  [21664/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.202246  [21696/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.242658  [21728/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.204086  [21760/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.256420  [21792/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.138073  [21824/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.282265  [21856/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.175671  [21888/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.388747  [21920/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.325717  [21952/24872]:   0%|          | 0/777 [00:23<?, ?it/s]
loss: 0.240126  [21984/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.336226  [22016/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.296628  [22048/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.163268  [22080/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.164253  [22112/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.372569  [22144/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.192451  [22176/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.170442  [22208/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.290344  [22240/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.240579  [22272/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.237832  [22304/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.179066  [22336/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.193672  [22368/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.254587  [22400/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.241385  [22432/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.148456  [22464/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.185942  [22496/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.282907  [22528/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.183837  [22560/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.213958  [22592/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.206904  [22624/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.245530  [22656/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.219356  [22688/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.292874  [22720/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.188250  [22752/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.182968  [22784/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.185576  [22816/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.166109  [22848/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.193606  [22880/24872]:   0%|          | 0/777 [00:24<?, ?it/s]
loss: 0.173819  [22912/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.203100  [22944/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.290066  [22976/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.174250  [23008/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.159719  [23040/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.198109  [23072/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.210914  [23104/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.190086  [23136/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.241909  [23168/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.196436  [23200/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.254771  [23232/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.184224  [23264/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.131497  [23296/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.140331  [23328/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.205520  [23360/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.223796  [23392/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.207892  [23424/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.231232  [23456/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.244744  [23488/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.278279  [23520/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.183505  [23552/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.166500  [23584/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.188898  [23616/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.170367  [23648/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.216136  [23680/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.249910  [23712/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.189322  [23744/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.161385  [23776/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.173190  [23808/24872]:   0%|          | 0/777 [00:25<?, ?it/s]
loss: 0.319411  [23840/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.277224  [23872/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.205208  [23904/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.120674  [23936/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.160895  [23968/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.227092  [24000/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.227771  [24032/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.226905  [24064/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.177816  [24096/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.263867  [24128/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.282506  [24160/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.323407  [24192/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.171246  [24224/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.224648  [24256/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.151338  [24288/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.164658  [24320/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.276886  [24352/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.189525  [24384/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.282277  [24416/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.243363  [24448/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.147833  [24480/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.262915  [24512/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.153546  [24544/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.152037  [24576/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.298028  [24608/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.198104  [24640/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.153644  [24672/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.171082  [24704/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.166982  [24736/24872]:   0%|          | 0/777 [00:26<?, ?it/s]
loss: 0.192164  [24768/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.222834  [24800/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.199489  [24832/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.204046  [24864/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.233870  [24872/24872]:   0%|          | 0/777 [00:27<?, ?it/s]
loss: 0.233870  [24872/24872]: : 778it [00:27, 28.64it/s]
-------------------------------
LR=0.0001, batch_size=64
-------------------------------
Epoch 1, time=134.74s

  0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.163974  [   64/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.257246  [  128/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.259823  [  192/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.176968  [  256/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.189795  [  320/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.233499  [  384/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.182119  [  448/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.213433  [  512/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.174773  [  576/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.196562  [  640/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.195010  [  704/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.189389  [  768/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.207833  [  832/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.274366  [  896/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.176506  [  960/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.239550  [ 1024/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.167055  [ 1088/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.177015  [ 1152/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.213861  [ 1216/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.188025  [ 1280/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.272112  [ 1344/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.212831  [ 1408/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.198628  [ 1472/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.248276  [ 1536/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.230839  [ 1600/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.227453  [ 1664/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.262093  [ 1728/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.192747  [ 1792/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.194912  [ 1856/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.215248  [ 1920/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.181835  [ 1984/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.130507  [ 2048/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.172018  [ 2112/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.214118  [ 2176/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.148629  [ 2240/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.157986  [ 2304/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.199971  [ 2368/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.210601  [ 2432/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.143489  [ 2496/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.184004  [ 2560/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.171941  [ 2624/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.136443  [ 2688/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.155506  [ 2752/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.165309  [ 2816/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.243732  [ 2880/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.138476  [ 2944/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.129283  [ 3008/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.166618  [ 3072/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.197774  [ 3136/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.140825  [ 3200/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.170050  [ 3264/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.230245  [ 3328/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.156071  [ 3392/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.200456  [ 3456/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.180347  [ 3520/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.174048  [ 3584/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.162015  [ 3648/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.178730  [ 3712/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.171639  [ 3776/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.143926  [ 3840/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.133465  [ 3904/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.189764  [ 3968/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.190428  [ 4032/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.231189  [ 4096/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.179640  [ 4160/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.167394  [ 4224/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.182074  [ 4288/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.135752  [ 4352/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.125924  [ 4416/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.198898  [ 4480/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.168299  [ 4544/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.209996  [ 4608/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.166113  [ 4672/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.193911  [ 4736/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.214729  [ 4800/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.179790  [ 4864/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.192385  [ 4928/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.188993  [ 4992/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.175961  [ 5056/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.189281  [ 5120/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.202691  [ 5184/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.196206  [ 5248/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.153194  [ 5312/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.211974  [ 5376/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.170495  [ 5440/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.127777  [ 5504/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.183684  [ 5568/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.179464  [ 5632/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.268624  [ 5696/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.164991  [ 5760/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.189012  [ 5824/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.177820  [ 5888/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.170592  [ 5952/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.195837  [ 6016/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.164327  [ 6080/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.183660  [ 6144/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.236137  [ 6208/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.117684  [ 6272/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.184646  [ 6336/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.244357  [ 6400/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.156839  [ 6464/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.146035  [ 6528/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.168840  [ 6592/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.166981  [ 6656/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.189457  [ 6720/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.199887  [ 6784/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.183707  [ 6848/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.165843  [ 6912/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.164165  [ 6976/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.172306  [ 7040/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.159102  [ 7104/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.188582  [ 7168/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.213539  [ 7232/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.240598  [ 7296/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.151187  [ 7360/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.189439  [ 7424/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.138101  [ 7488/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.160404  [ 7552/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.163362  [ 7616/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.206207  [ 7680/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.197540  [ 7744/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.172607  [ 7808/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.139153  [ 7872/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.200833  [ 7936/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.156795  [ 8000/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.174894  [ 8064/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.156677  [ 8128/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.211465  [ 8192/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.122611  [ 8256/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.133831  [ 8320/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.161244  [ 8384/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.165689  [ 8448/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.183294  [ 8512/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.174074  [ 8576/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.207056  [ 8640/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.204286  [ 8704/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.244632  [ 8768/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.179061  [ 8832/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.146012  [ 8896/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.148669  [ 8960/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.143045  [ 9024/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.157764  [ 9088/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.178678  [ 9152/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.168495  [ 9216/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.176244  [ 9280/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.160935  [ 9344/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.248159  [ 9408/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.205250  [ 9472/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.173843  [ 9536/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.212357  [ 9600/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.159582  [ 9664/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.149648  [ 9728/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.166727  [ 9792/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.210195  [ 9856/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.159210  [ 9920/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.216142  [ 9984/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.181443  [10048/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.123005  [10112/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.199310  [10176/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.151942  [10240/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.230931  [10304/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.178491  [10368/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.206150  [10432/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.194271  [10496/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.206719  [10560/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.194017  [10624/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.185016  [10688/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.185805  [10752/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.196901  [10816/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.246084  [10880/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.199905  [10944/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.131019  [11008/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.176643  [11072/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.216315  [11136/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.227460  [11200/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.187065  [11264/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.221638  [11328/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.196777  [11392/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.186374  [11456/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.148881  [11520/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.171275  [11584/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.193002  [11648/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.197770  [11712/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.185358  [11776/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.167998  [11840/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.150641  [11904/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.173300  [11968/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.213471  [12032/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.211134  [12096/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.178827  [12160/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.253523  [12224/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.197465  [12288/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.274002  [12352/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.146556  [12416/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.141460  [12480/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.191141  [12544/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.197793  [12608/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.191178  [12672/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.185892  [12736/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.209935  [12800/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.189822  [12864/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.215229  [12928/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.211092  [12992/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.145932  [13056/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.203997  [13120/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.156976  [13184/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.169619  [13248/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.210156  [13312/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.165182  [13376/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.204276  [13440/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.185656  [13504/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.158638  [13568/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.166790  [13632/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.170457  [13696/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.153470  [13760/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.212378  [13824/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.191774  [13888/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.192179  [13952/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.194638  [14016/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.139277  [14080/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.124699  [14144/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.149887  [14208/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.187635  [14272/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.188334  [14336/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.188581  [14400/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.169435  [14464/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.136093  [14528/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.115233  [14592/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.167116  [14656/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.134093  [14720/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.177862  [14784/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.198879  [14848/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.166931  [14912/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.215515  [14976/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.169509  [15040/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.170711  [15104/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.173674  [15168/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.160554  [15232/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.139126  [15296/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.190185  [15360/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.156402  [15424/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.135210  [15488/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.132704  [15552/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.143313  [15616/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.186852  [15680/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.181184  [15744/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.143726  [15808/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.177171  [15872/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.150548  [15936/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.146508  [16000/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.169386  [16064/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.198544  [16128/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.185840  [16192/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.199115  [16256/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.117629  [16320/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.208590  [16384/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.189582  [16448/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.191220  [16512/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.240475  [16576/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.103476  [16640/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.212089  [16704/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.170255  [16768/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.207994  [16832/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.149823  [16896/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.193450  [16960/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.185337  [17024/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.212299  [17088/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.147776  [17152/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.177196  [17216/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.199926  [17280/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.197487  [17344/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.195960  [17408/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.176436  [17472/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.207220  [17536/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.164725  [17600/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.197185  [17664/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.155130  [17728/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.186507  [17792/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.179214  [17856/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.171486  [17920/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.197582  [17984/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.180246  [18048/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.145159  [18112/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.167404  [18176/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.139657  [18240/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.169878  [18304/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.155933  [18368/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.143660  [18432/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.206794  [18496/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.220586  [18560/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.143951  [18624/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.219553  [18688/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.178940  [18752/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.186932  [18816/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.203008  [18880/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.178419  [18944/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.233232  [19008/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.168906  [19072/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.173848  [19136/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.126752  [19200/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.121287  [19264/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.172454  [19328/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.200058  [19392/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.156799  [19456/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.169728  [19520/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.135435  [19584/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.186984  [19648/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.236874  [19712/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.152866  [19776/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.166606  [19840/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.150147  [19904/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.184950  [19968/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.181658  [20032/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.152662  [20096/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.170286  [20160/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.218268  [20224/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.172923  [20288/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.193867  [20352/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.178840  [20416/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.109063  [20480/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.238443  [20544/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.116461  [20608/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.213609  [20672/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.203122  [20736/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.132578  [20800/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.144710  [20864/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.156627  [20928/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.174329  [20992/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.199345  [21056/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.166312  [21120/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.171796  [21184/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.181210  [21248/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.240336  [21312/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.153467  [21376/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.208802  [21440/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.126258  [21504/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.171746  [21568/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.228908  [21632/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.133758  [21696/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.127128  [21760/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.141354  [21824/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.125400  [21888/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.199843  [21952/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.177074  [22016/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.161776  [22080/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.165798  [22144/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.133317  [22208/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.164012  [22272/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.149745  [22336/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.182282  [22400/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.154552  [22464/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.181522  [22528/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.163008  [22592/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.172354  [22656/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.205291  [22720/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.156027  [22784/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.163893  [22848/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.165116  [22912/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.211451  [22976/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.148264  [23040/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.187941  [23104/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.204182  [23168/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.207975  [23232/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.135525  [23296/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.152453  [23360/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.171143  [23424/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.191717  [23488/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.175076  [23552/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.134942  [23616/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.175202  [23680/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.204591  [23744/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.130779  [23808/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.260640  [23872/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.141998  [23936/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.167760  [24000/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.197130  [24064/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.195129  [24128/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.260377  [24192/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.174370  [24256/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.134019  [24320/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.222421  [24384/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.212518  [24448/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.189734  [24512/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.131215  [24576/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.216381  [24640/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.144298  [24704/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.161930  [24768/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.197655  [24832/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.182333  [24872/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.182333  [24872/24872]: : 389it [00:16, 24.30it/s]
Epoch 2, time=150.75s

  0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.148205  [   64/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.147549  [  128/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.143671  [  192/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.137973  [  256/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.119024  [  320/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.165894  [  384/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.142443  [  448/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.170628  [  512/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.142994  [  576/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.154400  [  640/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.153505  [  704/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.134229  [  768/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.151465  [  832/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.238669  [  896/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.151848  [  960/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.222022  [ 1024/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.143621  [ 1088/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.151971  [ 1152/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.187861  [ 1216/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.161710  [ 1280/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.234563  [ 1344/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.182325  [ 1408/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.184880  [ 1472/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.240622  [ 1536/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.176662  [ 1600/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.216729  [ 1664/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.211482  [ 1728/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.173693  [ 1792/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.190369  [ 1856/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.172644  [ 1920/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.162598  [ 1984/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.114328  [ 2048/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.148966  [ 2112/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.205933  [ 2176/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.130847  [ 2240/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.130928  [ 2304/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.174909  [ 2368/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.182770  [ 2432/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.143131  [ 2496/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.169863  [ 2560/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.156725  [ 2624/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.135996  [ 2688/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.146166  [ 2752/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.154716  [ 2816/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.252172  [ 2880/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.112621  [ 2944/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.113192  [ 3008/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.151073  [ 3072/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.171227  [ 3136/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.111335  [ 3200/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.142091  [ 3264/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.221150  [ 3328/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.144241  [ 3392/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.182294  [ 3456/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.159592  [ 3520/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.152571  [ 3584/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.137728  [ 3648/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.166372  [ 3712/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.145787  [ 3776/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.132388  [ 3840/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.123303  [ 3904/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.165464  [ 3968/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.175276  [ 4032/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.205470  [ 4096/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.154158  [ 4160/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.158518  [ 4224/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.165334  [ 4288/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.124033  [ 4352/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.112297  [ 4416/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.180186  [ 4480/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.145800  [ 4544/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.180810  [ 4608/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.159160  [ 4672/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.178429  [ 4736/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.195419  [ 4800/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.157146  [ 4864/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.186784  [ 4928/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.173286  [ 4992/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.172450  [ 5056/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.169671  [ 5120/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.199951  [ 5184/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.178245  [ 5248/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.147754  [ 5312/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.185317  [ 5376/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.151185  [ 5440/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.113389  [ 5504/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.173760  [ 5568/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.159564  [ 5632/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.246886  [ 5696/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.153048  [ 5760/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.150497  [ 5824/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.158711  [ 5888/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.154042  [ 5952/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.168919  [ 6016/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.136651  [ 6080/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.159393  [ 6144/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.200726  [ 6208/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.105802  [ 6272/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.163361  [ 6336/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.199218  [ 6400/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.135097  [ 6464/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.120302  [ 6528/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.159325  [ 6592/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.143414  [ 6656/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.165697  [ 6720/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.163363  [ 6784/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.169703  [ 6848/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.136885  [ 6912/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.142579  [ 6976/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.148447  [ 7040/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.144345  [ 7104/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.165011  [ 7168/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.197877  [ 7232/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.232243  [ 7296/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.133652  [ 7360/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.162900  [ 7424/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.119220  [ 7488/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.142978  [ 7552/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.152147  [ 7616/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.191576  [ 7680/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.167307  [ 7744/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.169744  [ 7808/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.123330  [ 7872/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.181457  [ 7936/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.151800  [ 8000/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.163805  [ 8064/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.142124  [ 8128/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.180674  [ 8192/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.112608  [ 8256/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.113905  [ 8320/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.172327  [ 8384/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.153804  [ 8448/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.165612  [ 8512/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.155351  [ 8576/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.194240  [ 8640/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.180082  [ 8704/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.198783  [ 8768/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.164437  [ 8832/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.137850  [ 8896/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.142711  [ 8960/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.128485  [ 9024/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.140635  [ 9088/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.169381  [ 9152/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.148100  [ 9216/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.163089  [ 9280/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.139539  [ 9344/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.217248  [ 9408/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.185509  [ 9472/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.158367  [ 9536/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.193637  [ 9600/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.144466  [ 9664/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.117436  [ 9728/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.143508  [ 9792/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.166336  [ 9856/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.148875  [ 9920/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.182361  [ 9984/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.139039  [10048/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.109107  [10112/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.166916  [10176/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.131885  [10240/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.197241  [10304/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.139958  [10368/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.169005  [10432/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.160322  [10496/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.171635  [10560/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.156296  [10624/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.155804  [10688/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.167060  [10752/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.188271  [10816/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.215437  [10880/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.176932  [10944/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.105643  [11008/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.143680  [11072/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.187749  [11136/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.193479  [11200/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.179585  [11264/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.173561  [11328/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.172955  [11392/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.167641  [11456/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.120996  [11520/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.147179  [11584/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.157292  [11648/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.163320  [11712/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.146920  [11776/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.149146  [11840/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.136772  [11904/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.159858  [11968/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.176983  [12032/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.185507  [12096/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.147024  [12160/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.196488  [12224/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.175187  [12288/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.224808  [12352/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.112756  [12416/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.121995  [12480/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.160593  [12544/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.160164  [12608/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.169294  [12672/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.151256  [12736/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.198191  [12800/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.170422  [12864/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.200391  [12928/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.195359  [12992/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.127267  [13056/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.185351  [13120/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.138818  [13184/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.171690  [13248/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.210394  [13312/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.147976  [13376/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.188179  [13440/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.151149  [13504/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.137448  [13568/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.161266  [13632/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.151027  [13696/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.148121  [13760/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.197032  [13824/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.177861  [13888/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.178943  [13952/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.184420  [14016/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.131090  [14080/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.103712  [14144/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.144221  [14208/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.172814  [14272/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.186171  [14336/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.180313  [14400/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.160255  [14464/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.134323  [14528/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.115079  [14592/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.164938  [14656/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.129262  [14720/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.181473  [14784/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.181761  [14848/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.170159  [14912/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.204460  [14976/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.166021  [15040/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.155481  [15104/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.168568  [15168/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.155128  [15232/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.136900  [15296/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.176170  [15360/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.158477  [15424/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.142903  [15488/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.114290  [15552/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.138645  [15616/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.188048  [15680/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.171938  [15744/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.146608  [15808/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.166787  [15872/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.141325  [15936/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.137075  [16000/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.136769  [16064/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.205144  [16128/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.152338  [16192/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.170671  [16256/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.100740  [16320/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.184827  [16384/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.175503  [16448/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.172959  [16512/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.242863  [16576/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.089912  [16640/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.194130  [16704/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.153464  [16768/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.210320  [16832/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.133437  [16896/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.195647  [16960/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.169576  [17024/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.192431  [17088/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.140129  [17152/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.160823  [17216/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.210337  [17280/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.160284  [17344/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.178685  [17408/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.164286  [17472/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.188305  [17536/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.151926  [17600/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.205053  [17664/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.146778  [17728/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.202085  [17792/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.184432  [17856/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.168788  [17920/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.203378  [17984/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.139528  [18048/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.160063  [18112/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.177986  [18176/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.109096  [18240/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.177187  [18304/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.124050  [18368/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.176962  [18432/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.196564  [18496/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.187228  [18560/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.161156  [18624/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.199479  [18688/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.164152  [18752/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.203034  [18816/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.187414  [18880/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.174344  [18944/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.213406  [19008/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.162781  [19072/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.179792  [19136/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.121972  [19200/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.107643  [19264/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.127224  [19328/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.173753  [19392/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.148430  [19456/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.141477  [19520/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.139829  [19584/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.176180  [19648/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.215628  [19712/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.132938  [19776/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.149261  [19840/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.140892  [19904/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.185002  [19968/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.164153  [20032/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.138707  [20096/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.189304  [20160/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.175765  [20224/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.175398  [20288/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.205439  [20352/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.172800  [20416/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.110888  [20480/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.203531  [20544/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.127636  [20608/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.210019  [20672/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.203533  [20736/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.134522  [20800/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.143822  [20864/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.157530  [20928/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.172751  [20992/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.164364  [21056/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.153413  [21120/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.162178  [21184/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.184965  [21248/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.232280  [21312/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.146166  [21376/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.205978  [21440/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.109945  [21504/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.175056  [21568/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.207933  [21632/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.157419  [21696/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.131177  [21760/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.138728  [21824/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.140806  [21888/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.165492  [21952/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.179766  [22016/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.152630  [22080/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.177880  [22144/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.119420  [22208/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.146266  [22272/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.146515  [22336/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.156674  [22400/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.128451  [22464/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.178457  [22528/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.163754  [22592/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.153432  [22656/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.176345  [22720/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.136798  [22784/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.147027  [22848/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.150947  [22912/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.187157  [22976/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.143708  [23040/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.164474  [23104/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.197796  [23168/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.179156  [23232/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.143703  [23296/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.144790  [23360/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.172538  [23424/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.183296  [23488/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.166911  [23552/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.135098  [23616/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.161579  [23680/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.191525  [23744/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.133473  [23808/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.226319  [23872/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.137334  [23936/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.134843  [24000/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.187817  [24064/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.175366  [24128/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.233268  [24192/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.147202  [24256/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.118186  [24320/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.189091  [24384/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.234918  [24448/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.183749  [24512/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.115553  [24576/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.189257  [24640/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.138025  [24704/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.140384  [24768/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.180664  [24832/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.173768  [24872/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.173768  [24872/24872]: : 389it [00:16, 24.25it/s]
Epoch 3, time=166.79s

  0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.157437  [   64/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.151315  [  128/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.131418  [  192/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.126369  [  256/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.112664  [  320/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.158617  [  384/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.133989  [  448/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.155912  [  512/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.126392  [  576/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.135485  [  640/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.145456  [  704/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.139895  [  768/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.135772  [  832/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.211516  [  896/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.135653  [  960/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.205228  [ 1024/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.136988  [ 1088/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.138242  [ 1152/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.170655  [ 1216/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.136708  [ 1280/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.206170  [ 1344/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.185839  [ 1408/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.181803  [ 1472/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.201506  [ 1536/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.164360  [ 1600/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.210121  [ 1664/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.192168  [ 1728/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.152073  [ 1792/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.175003  [ 1856/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.152771  [ 1920/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.141195  [ 1984/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.102730  [ 2048/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.141125  [ 2112/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.181215  [ 2176/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.120833  [ 2240/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.120718  [ 2304/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.170403  [ 2368/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.176146  [ 2432/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.142080  [ 2496/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.144629  [ 2560/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.146869  [ 2624/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.120068  [ 2688/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.159706  [ 2752/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.151146  [ 2816/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.203021  [ 2880/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.121444  [ 2944/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.107471  [ 3008/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.143155  [ 3072/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.166811  [ 3136/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.099405  [ 3200/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.144065  [ 3264/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.213712  [ 3328/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.147959  [ 3392/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.194168  [ 3456/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.144719  [ 3520/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.148758  [ 3584/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.122857  [ 3648/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.161175  [ 3712/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.135467  [ 3776/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.145083  [ 3840/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.117526  [ 3904/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.159129  [ 3968/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.160933  [ 4032/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.190724  [ 4096/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.153685  [ 4160/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.143651  [ 4224/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.153172  [ 4288/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.137182  [ 4352/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.099228  [ 4416/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.196055  [ 4480/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.143356  [ 4544/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.182360  [ 4608/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.152982  [ 4672/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.161665  [ 4736/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.190140  [ 4800/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.147825  [ 4864/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.178226  [ 4928/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.171070  [ 4992/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.158633  [ 5056/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.150163  [ 5120/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.196165  [ 5184/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.156459  [ 5248/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.127806  [ 5312/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.172204  [ 5376/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.131112  [ 5440/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.112560  [ 5504/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.159167  [ 5568/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.161366  [ 5632/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.207175  [ 5696/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.137839  [ 5760/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.145989  [ 5824/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.143429  [ 5888/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.147361  [ 5952/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.165203  [ 6016/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.132454  [ 6080/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.139915  [ 6144/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.174863  [ 6208/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.098715  [ 6272/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.178977  [ 6336/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.176437  [ 6400/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.110572  [ 6464/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.126072  [ 6528/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.124523  [ 6592/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.142445  [ 6656/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.160250  [ 6720/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.158677  [ 6784/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.164950  [ 6848/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.125829  [ 6912/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.133651  [ 6976/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.128673  [ 7040/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.130789  [ 7104/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.147564  [ 7168/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.179388  [ 7232/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.180485  [ 7296/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.120058  [ 7360/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.137041  [ 7424/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.110256  [ 7488/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.138850  [ 7552/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.147639  [ 7616/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.184589  [ 7680/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.152927  [ 7744/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.163927  [ 7808/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.127856  [ 7872/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.168364  [ 7936/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.134538  [ 8000/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.141109  [ 8064/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.130391  [ 8128/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.165362  [ 8192/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.111584  [ 8256/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.106352  [ 8320/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.127725  [ 8384/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.145637  [ 8448/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.166425  [ 8512/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.156102  [ 8576/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.184106  [ 8640/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.191511  [ 8704/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.179756  [ 8768/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.158896  [ 8832/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.123886  [ 8896/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.141615  [ 8960/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.127795  [ 9024/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.126467  [ 9088/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.154731  [ 9152/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.150024  [ 9216/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.154476  [ 9280/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.139637  [ 9344/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.164739  [ 9408/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.188460  [ 9472/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.151969  [ 9536/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.184954  [ 9600/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.135856  [ 9664/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.112846  [ 9728/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.137258  [ 9792/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.148993  [ 9856/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.162189  [ 9920/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.166413  [ 9984/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.137561  [10048/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.093024  [10112/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.149124  [10176/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.129824  [10240/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.192205  [10304/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.148106  [10368/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.178278  [10432/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.134608  [10496/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.158084  [10560/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.164118  [10624/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.153504  [10688/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.187254  [10752/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.177967  [10816/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.219404  [10880/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.185658  [10944/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.116479  [11008/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.157241  [11072/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.154957  [11136/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.210485  [11200/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.158165  [11264/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.182701  [11328/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.140176  [11392/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.190118  [11456/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.145370  [11520/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.131088  [11584/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.170658  [11648/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.171106  [11712/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.140707  [11776/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.169579  [11840/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.128694  [11904/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.148978  [11968/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.169203  [12032/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.181396  [12096/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.151627  [12160/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.193352  [12224/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.153663  [12288/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.225104  [12352/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.134263  [12416/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.111636  [12480/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.138821  [12544/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.150191  [12608/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.139600  [12672/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.156280  [12736/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.185463  [12800/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.184279  [12864/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.184061  [12928/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.180948  [12992/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.119300  [13056/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.173683  [13120/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.142110  [13184/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.176673  [13248/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.194621  [13312/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.142598  [13376/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.179064  [13440/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.133673  [13504/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.125949  [13568/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.157312  [13632/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.129951  [13696/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.150439  [13760/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.180508  [13824/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.164833  [13888/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.186946  [13952/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.162632  [14016/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.122088  [14080/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.110163  [14144/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.125613  [14208/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.157887  [14272/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.199389  [14336/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.152608  [14400/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.129662  [14464/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.141381  [14528/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.087754  [14592/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.149080  [14656/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.108759  [14720/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.156973  [14784/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.162871  [14848/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.119929  [14912/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.152475  [14976/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.141032  [15040/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.141424  [15104/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.141345  [15168/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.148078  [15232/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.117530  [15296/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.158366  [15360/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.144624  [15424/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.147797  [15488/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.110400  [15552/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.133927  [15616/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.176858  [15680/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.160504  [15744/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.149008  [15808/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.144776  [15872/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.124406  [15936/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.143336  [16000/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.118035  [16064/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.151070  [16128/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.143517  [16192/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.183098  [16256/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.094451  [16320/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.167803  [16384/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.157530  [16448/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.149082  [16512/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.206064  [16576/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.081732  [16640/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.187262  [16704/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.152989  [16768/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.169059  [16832/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.138809  [16896/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.171042  [16960/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.166086  [17024/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.185308  [17088/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.121155  [17152/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.130727  [17216/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.187059  [17280/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.140041  [17344/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.183709  [17408/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.150838  [17472/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.166420  [17536/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.142628  [17600/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.196754  [17664/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.135955  [17728/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.192530  [17792/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.167675  [17856/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.150727  [17920/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.193019  [17984/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.149123  [18048/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.137709  [18112/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.176165  [18176/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.108581  [18240/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.167310  [18304/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.174599  [18368/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.156428  [18432/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.223096  [18496/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.200146  [18560/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.152580  [18624/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.225429  [18688/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.194192  [18752/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.184763  [18816/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.207640  [18880/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.163027  [18944/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.235877  [19008/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.137688  [19072/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.150804  [19136/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.115147  [19200/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.088666  [19264/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.146446  [19328/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.200743  [19392/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.147625  [19456/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.133334  [19520/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.131065  [19584/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.196100  [19648/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.211119  [19712/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.149634  [19776/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.150295  [19840/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.124011  [19904/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.201989  [19968/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.148951  [20032/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.139576  [20096/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.147085  [20160/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.193211  [20224/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.161740  [20288/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.183348  [20352/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.171296  [20416/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.087494  [20480/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.182245  [20544/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.098986  [20608/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.178710  [20672/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.186824  [20736/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.118886  [20800/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.133362  [20864/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.130819  [20928/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.153898  [20992/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.148074  [21056/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.135640  [21120/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.141013  [21184/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.160658  [21248/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.216485  [21312/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.137107  [21376/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.202801  [21440/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.101975  [21504/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.154324  [21568/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.200720  [21632/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.128707  [21696/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.109306  [21760/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.116836  [21824/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.103396  [21888/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.159570  [21952/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.151636  [22016/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.134924  [22080/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.149263  [22144/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.113570  [22208/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.135455  [22272/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.150734  [22336/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.133227  [22400/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.122896  [22464/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.166954  [22528/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.148247  [22592/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.146045  [22656/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.152786  [22720/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.134066  [22784/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.142693  [22848/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.139152  [22912/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.172904  [22976/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.136354  [23040/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.155585  [23104/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.193605  [23168/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.175712  [23232/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.137974  [23296/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.139191  [23360/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.160460  [23424/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.195282  [23488/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.154740  [23552/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.118236  [23616/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.146667  [23680/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.177542  [23744/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.130309  [23808/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.208387  [23872/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.134189  [23936/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.127962  [24000/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.196019  [24064/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.172580  [24128/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.235490  [24192/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.176997  [24256/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.122911  [24320/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.182669  [24384/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.245355  [24448/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.190818  [24512/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.121491  [24576/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.221855  [24640/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.119200  [24704/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.145582  [24768/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.185769  [24832/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.174007  [24872/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.174007  [24872/24872]: : 389it [00:17, 21.76it/s]
Epoch 4, time=184.67s

  0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.155637  [   64/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.140122  [  128/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.125975  [  192/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.143831  [  256/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.107935  [  320/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.165160  [  384/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.134560  [  448/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.131583  [  512/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.122002  [  576/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.128829  [  640/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.132072  [  704/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.135621  [  768/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.126871  [  832/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.201774  [  896/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.131304  [  960/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.197302  [ 1024/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.130760  [ 1088/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.127902  [ 1152/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.159275  [ 1216/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.113751  [ 1280/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.198755  [ 1344/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.176311  [ 1408/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.168547  [ 1472/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.189809  [ 1536/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.162098  [ 1600/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.200925  [ 1664/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.201221  [ 1728/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.141883  [ 1792/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.168105  [ 1856/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.145139  [ 1920/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.142116  [ 1984/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.105936  [ 2048/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.134631  [ 2112/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.161775  [ 2176/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.110372  [ 2240/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.115247  [ 2304/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.171554  [ 2368/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.189059  [ 2432/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.109587  [ 2496/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.150246  [ 2560/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.135795  [ 2624/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.132814  [ 2688/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.142659  [ 2752/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.151589  [ 2816/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.168448  [ 2880/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.111520  [ 2944/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.101196  [ 3008/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.136006  [ 3072/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.135598  [ 3136/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.110975  [ 3200/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.134599  [ 3264/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.195316  [ 3328/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.139829  [ 3392/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.181266  [ 3456/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.133753  [ 3520/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.141219  [ 3584/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.135866  [ 3648/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.159343  [ 3712/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.137870  [ 3776/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.117318  [ 3840/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.105876  [ 3904/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.144684  [ 3968/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.158880  [ 4032/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.177514  [ 4096/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.144384  [ 4160/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.139875  [ 4224/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.140442  [ 4288/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.106434  [ 4352/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.111650  [ 4416/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.166569  [ 4480/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.129486  [ 4544/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.166781  [ 4608/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.151298  [ 4672/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.151588  [ 4736/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.175421  [ 4800/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.137107  [ 4864/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.169017  [ 4928/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.159604  [ 4992/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.156980  [ 5056/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.146896  [ 5120/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.182767  [ 5184/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.148011  [ 5248/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.116487  [ 5312/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.168006  [ 5376/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.125580  [ 5440/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.117705  [ 5504/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.152429  [ 5568/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.146082  [ 5632/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.209886  [ 5696/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.122915  [ 5760/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.144845  [ 5824/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.136486  [ 5888/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.142065  [ 5952/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.165952  [ 6016/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.126524  [ 6080/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.142005  [ 6144/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.166640  [ 6208/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.108650  [ 6272/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.157638  [ 6336/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.176599  [ 6400/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.106967  [ 6464/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.114866  [ 6528/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.123442  [ 6592/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.121253  [ 6656/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.151751  [ 6720/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.157335  [ 6784/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.150124  [ 6848/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.130503  [ 6912/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.120867  [ 6976/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.112989  [ 7040/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.134135  [ 7104/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.138803  [ 7168/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.166931  [ 7232/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.177568  [ 7296/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.113878  [ 7360/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.146992  [ 7424/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.107193  [ 7488/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.122710  [ 7552/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.141655  [ 7616/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.156601  [ 7680/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.143989  [ 7744/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.127517  [ 7808/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.108303  [ 7872/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.153216  [ 7936/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.123069  [ 8000/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.142136  [ 8064/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.115180  [ 8128/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.171496  [ 8192/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.109144  [ 8256/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.101770  [ 8320/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.132288  [ 8384/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.171633  [ 8448/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.149066  [ 8512/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.158445  [ 8576/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.177778  [ 8640/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.165456  [ 8704/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.171570  [ 8768/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.164758  [ 8832/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.121782  [ 8896/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.136450  [ 8960/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.128973  [ 9024/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.141755  [ 9088/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.166344  [ 9152/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.152409  [ 9216/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.127221  [ 9280/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.145653  [ 9344/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.202146  [ 9408/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.194696  [ 9472/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.144421  [ 9536/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.165709  [ 9600/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.146338  [ 9664/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.101614  [ 9728/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.150766  [ 9792/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.164387  [ 9856/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.131642  [ 9920/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.181272  [ 9984/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.134380  [10048/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.105556  [10112/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.134001  [10176/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.119898  [10240/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.183700  [10304/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.133208  [10368/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.148545  [10432/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.146365  [10496/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.152077  [10560/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.152340  [10624/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.147999  [10688/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.150384  [10752/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.164696  [10816/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.195709  [10880/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.163321  [10944/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.101828  [11008/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.152401  [11072/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.147935  [11136/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.175568  [11200/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.142703  [11264/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.148091  [11328/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.147278  [11392/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.163600  [11456/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.118293  [11520/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.130703  [11584/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.145375  [11648/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.167144  [11712/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.131622  [11776/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.126713  [11840/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.125977  [11904/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.152451  [11968/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.146131  [12032/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.190708  [12096/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.136433  [12160/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.189737  [12224/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.154248  [12288/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.226864  [12352/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.102019  [12416/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.124025  [12480/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.139108  [12544/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.152826  [12608/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.167110  [12672/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.153618  [12736/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.184542  [12800/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.155208  [12864/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.177879  [12928/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.180514  [12992/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.111448  [13056/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.175552  [13120/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.134740  [13184/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.160929  [13248/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.172234  [13312/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.130528  [13376/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.155515  [13440/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.151657  [13504/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.114013  [13568/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.146982  [13632/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.120776  [13696/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.136695  [13760/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.196090  [13824/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.142033  [13888/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.166926  [13952/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.148696  [14016/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.117833  [14080/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.103657  [14144/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.123465  [14208/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.151148  [14272/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.157842  [14336/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.151158  [14400/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.118374  [14464/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.122741  [14528/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.089862  [14592/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.149099  [14656/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.111736  [14720/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.143245  [14784/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.173977  [14848/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.117478  [14912/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.149247  [14976/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.131957  [15040/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.133684  [15104/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.125108  [15168/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.130556  [15232/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.106473  [15296/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.151608  [15360/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.136719  [15424/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.130210  [15488/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.094211  [15552/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.120927  [15616/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.145773  [15680/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.145483  [15744/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.141330  [15808/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.145327  [15872/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.121863  [15936/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.138668  [16000/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.125839  [16064/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.159397  [16128/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.136484  [16192/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.168103  [16256/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.094586  [16320/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.169424  [16384/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.148001  [16448/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.153192  [16512/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.225240  [16576/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.088956  [16640/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.194323  [16704/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.143920  [16768/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.206955  [16832/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.150341  [16896/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.155340  [16960/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.165327  [17024/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.189103  [17088/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.119409  [17152/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.139745  [17216/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.174951  [17280/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.154159  [17344/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.188388  [17408/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.151693  [17472/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.154591  [17536/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.178284  [17600/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.160226  [17664/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.145609  [17728/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.186962  [17792/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.172925  [17856/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.167906  [17920/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.157500  [17984/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.118522  [18048/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.128383  [18112/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.143710  [18176/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.099325  [18240/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.115182  [18304/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.096337  [18368/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.150405  [18432/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.150698  [18496/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.190377  [18560/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.137185  [18624/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.195666  [18688/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.138422  [18752/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.211653  [18816/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.163271  [18880/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.130939  [18944/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.195132  [19008/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.137406  [19072/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.167496  [19136/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.113562  [19200/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.074854  [19264/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.100842  [19328/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.153697  [19392/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.133949  [19456/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.131201  [19520/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.115525  [19584/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.157278  [19648/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.203113  [19712/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.109147  [19776/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.138470  [19840/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.121800  [19904/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.191509  [19968/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.138801  [20032/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.132080  [20096/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.168350  [20160/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.160658  [20224/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.160967  [20288/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.169291  [20352/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.153533  [20416/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.087398  [20480/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.179546  [20544/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.093874  [20608/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.171952  [20672/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.194364  [20736/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.118316  [20800/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.110406  [20864/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.135683  [20928/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.146411  [20992/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.131462  [21056/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.148195  [21120/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.137730  [21184/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.171224  [21248/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.213809  [21312/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.122607  [21376/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.199682  [21440/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.110252  [21504/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.144445  [21568/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.220577  [21632/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.128833  [21696/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.105527  [21760/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.120544  [21824/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.095057  [21888/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.151342  [21952/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.136250  [22016/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.132595  [22080/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.128513  [22144/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.105778  [22208/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.123551  [22272/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.116556  [22336/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.119161  [22400/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.094728  [22464/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.150442  [22528/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.135902  [22592/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.119455  [22656/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.149084  [22720/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.115235  [22784/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.119269  [22848/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.133045  [22912/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.154204  [22976/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.125210  [23040/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.139986  [23104/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.148507  [23168/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.161338  [23232/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.120843  [23296/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.091434  [23360/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.135454  [23424/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.155257  [23488/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.148385  [23552/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.106904  [23616/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.135369  [23680/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.152613  [23744/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.095535  [23808/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.188500  [23872/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.118537  [23936/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.124621  [24000/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.160482  [24064/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.150368  [24128/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.207223  [24192/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.141434  [24256/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.099662  [24320/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.176168  [24384/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.194973  [24448/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.168462  [24512/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.102994  [24576/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.175376  [24640/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.111007  [24704/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.124367  [24768/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.154628  [24832/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.139772  [24872/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.139772  [24872/24872]: : 389it [00:18, 20.87it/s]
Epoch 5, time=203.30s

  0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.112533  [   64/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.111596  [  128/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.108599  [  192/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.118483  [  256/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.118154  [  320/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.127059  [  384/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.112232  [  448/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.116576  [  512/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.109726  [  576/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.129927  [  640/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.128707  [  704/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.098110  [  768/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.108242  [  832/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.182869  [  896/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.105253  [  960/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.152898  [ 1024/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.115509  [ 1088/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.116279  [ 1152/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.142778  [ 1216/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.098184  [ 1280/24872]:   0%|          | 0/388 [00:00<?, ?it/s]
loss: 0.186332  [ 1344/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.159831  [ 1408/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.160857  [ 1472/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.174570  [ 1536/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.120685  [ 1600/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.227157  [ 1664/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.147759  [ 1728/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.142220  [ 1792/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.149282  [ 1856/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.141704  [ 1920/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.129973  [ 1984/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.106769  [ 2048/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.124453  [ 2112/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.139771  [ 2176/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.101323  [ 2240/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.113678  [ 2304/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.151629  [ 2368/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.181646  [ 2432/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.107025  [ 2496/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.127489  [ 2560/24872]:   0%|          | 0/388 [00:01<?, ?it/s]
loss: 0.148888  [ 2624/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.114044  [ 2688/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.150621  [ 2752/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.142083  [ 2816/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.173965  [ 2880/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.100613  [ 2944/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.114410  [ 3008/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.136637  [ 3072/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.142763  [ 3136/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.114105  [ 3200/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.134872  [ 3264/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.183565  [ 3328/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.143805  [ 3392/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.206704  [ 3456/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.147562  [ 3520/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.155807  [ 3584/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.115293  [ 3648/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.155954  [ 3712/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.143011  [ 3776/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.117734  [ 3840/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.104965  [ 3904/24872]:   0%|          | 0/388 [00:02<?, ?it/s]
loss: 0.121399  [ 3968/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.149727  [ 4032/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.183973  [ 4096/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.140066  [ 4160/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.132816  [ 4224/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.160876  [ 4288/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.114773  [ 4352/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.088317  [ 4416/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.182417  [ 4480/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.130801  [ 4544/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.173821  [ 4608/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.140934  [ 4672/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.139750  [ 4736/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.156260  [ 4800/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.126107  [ 4864/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.158001  [ 4928/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.137074  [ 4992/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.135028  [ 5056/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.124570  [ 5120/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.156260  [ 5184/24872]:   0%|          | 0/388 [00:03<?, ?it/s]
loss: 0.135016  [ 5248/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.106737  [ 5312/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.158591  [ 5376/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.105952  [ 5440/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.113478  [ 5504/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.148739  [ 5568/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.128372  [ 5632/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.185707  [ 5696/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.111799  [ 5760/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.128344  [ 5824/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.128125  [ 5888/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.134860  [ 5952/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.157672  [ 6016/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.118283  [ 6080/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.114193  [ 6144/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.164274  [ 6208/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.095402  [ 6272/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.154486  [ 6336/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.155502  [ 6400/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.099599  [ 6464/24872]:   0%|          | 0/388 [00:04<?, ?it/s]
loss: 0.111733  [ 6528/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.124113  [ 6592/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.113903  [ 6656/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.149832  [ 6720/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.150654  [ 6784/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.154849  [ 6848/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.125807  [ 6912/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.112368  [ 6976/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.107163  [ 7040/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.123720  [ 7104/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.124826  [ 7168/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.156096  [ 7232/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.150217  [ 7296/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.114900  [ 7360/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.131451  [ 7424/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.101127  [ 7488/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.125440  [ 7552/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.132377  [ 7616/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.156125  [ 7680/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.151598  [ 7744/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.131930  [ 7808/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.109165  [ 7872/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.141440  [ 7936/24872]:   0%|          | 0/388 [00:05<?, ?it/s]
loss: 0.118623  [ 8000/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.119777  [ 8064/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.102433  [ 8128/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.131087  [ 8192/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.093883  [ 8256/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.087994  [ 8320/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.107975  [ 8384/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.154259  [ 8448/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.146704  [ 8512/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.150003  [ 8576/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.155587  [ 8640/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.143617  [ 8704/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.142936  [ 8768/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.137794  [ 8832/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.107160  [ 8896/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.121283  [ 8960/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.119441  [ 9024/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.128187  [ 9088/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.143836  [ 9152/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.132530  [ 9216/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.105593  [ 9280/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.129779  [ 9344/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.144285  [ 9408/24872]:   0%|          | 0/388 [00:06<?, ?it/s]
loss: 0.167694  [ 9472/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.121207  [ 9536/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.153926  [ 9600/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.124972  [ 9664/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.089583  [ 9728/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.114127  [ 9792/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.127357  [ 9856/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.123034  [ 9920/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.143604  [ 9984/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.120141  [10048/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.091104  [10112/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.122116  [10176/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.111213  [10240/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.176052  [10304/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.134976  [10368/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.155472  [10432/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.118677  [10496/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.132583  [10560/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.143115  [10624/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.125715  [10688/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.143881  [10752/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.164043  [10816/24872]:   0%|          | 0/388 [00:07<?, ?it/s]
loss: 0.165148  [10880/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.148508  [10944/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.086853  [11008/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.140814  [11072/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.129402  [11136/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.155130  [11200/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.143143  [11264/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.137190  [11328/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.117573  [11392/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.165281  [11456/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.117649  [11520/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.113756  [11584/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.138189  [11648/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.149675  [11712/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.123649  [11776/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.119843  [11840/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.106293  [11904/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.142982  [11968/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.135965  [12032/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.173185  [12096/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.126394  [12160/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.149173  [12224/24872]:   0%|          | 0/388 [00:08<?, ?it/s]
loss: 0.143016  [12288/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.192608  [12352/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.097271  [12416/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.108395  [12480/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.113863  [12544/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.145844  [12608/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.129614  [12672/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.128768  [12736/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.169645  [12800/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.144231  [12864/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.169268  [12928/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.166336  [12992/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.110344  [13056/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.168170  [13120/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.124515  [13184/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.142500  [13248/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.151160  [13312/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.136873  [13376/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.147388  [13440/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.135540  [13504/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.108921  [13568/24872]:   0%|          | 0/388 [00:09<?, ?it/s]
loss: 0.144995  [13632/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.115583  [13696/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.130441  [13760/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.197025  [13824/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.135374  [13888/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.161389  [13952/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.155631  [14016/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.110699  [14080/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.104376  [14144/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.117924  [14208/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.137046  [14272/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.150240  [14336/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.152548  [14400/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.117835  [14464/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.137683  [14528/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.095893  [14592/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.143253  [14656/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.103223  [14720/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.151893  [14784/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.166411  [14848/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.122331  [14912/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.132716  [14976/24872]:   0%|          | 0/388 [00:10<?, ?it/s]
loss: 0.142714  [15040/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.129305  [15104/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.118941  [15168/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.126339  [15232/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.104525  [15296/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.150997  [15360/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.131993  [15424/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.121054  [15488/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.090181  [15552/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.113943  [15616/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.136591  [15680/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.137712  [15744/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.133546  [15808/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.124313  [15872/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.116525  [15936/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.133251  [16000/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.101127  [16064/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.157513  [16128/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.124616  [16192/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.168478  [16256/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.084192  [16320/24872]:   0%|          | 0/388 [00:11<?, ?it/s]
loss: 0.149863  [16384/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.151438  [16448/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.138103  [16512/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.182115  [16576/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.077537  [16640/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.159776  [16704/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.120042  [16768/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.158169  [16832/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.120316  [16896/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.141493  [16960/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.142146  [17024/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.169619  [17088/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.108243  [17152/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.114853  [17216/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.161376  [17280/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.130493  [17344/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.166186  [17408/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.137031  [17472/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.133655  [17536/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.159961  [17600/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.151918  [17664/24872]:   0%|          | 0/388 [00:12<?, ?it/s]
loss: 0.130719  [17728/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.162614  [17792/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.140171  [17856/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.175393  [17920/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.137814  [17984/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.114333  [18048/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.105712  [18112/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.106293  [18176/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.101988  [18240/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.106507  [18304/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.097805  [18368/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.151945  [18432/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.130552  [18496/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.156169  [18560/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.103696  [18624/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.155698  [18688/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.166595  [18752/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.149345  [18816/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.169329  [18880/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.121087  [18944/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.180981  [19008/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.132263  [19072/24872]:   0%|          | 0/388 [00:13<?, ?it/s]
loss: 0.137702  [19136/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.102939  [19200/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.071451  [19264/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.088545  [19328/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.141895  [19392/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.123679  [19456/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.125357  [19520/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.116889  [19584/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.149582  [19648/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.164718  [19712/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.132270  [19776/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.126102  [19840/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.118104  [19904/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.178164  [19968/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.124553  [20032/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.112976  [20096/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.143152  [20160/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.159675  [20224/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.144087  [20288/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.160978  [20352/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.145012  [20416/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.073172  [20480/24872]:   0%|          | 0/388 [00:14<?, ?it/s]
loss: 0.165199  [20544/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.081847  [20608/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.153665  [20672/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.156155  [20736/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.108825  [20800/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.106334  [20864/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.127280  [20928/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.129718  [20992/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.120496  [21056/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.125955  [21120/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.104833  [21184/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.151864  [21248/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.186467  [21312/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.113047  [21376/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.173831  [21440/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.089087  [21504/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.122842  [21568/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.168275  [21632/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.100474  [21696/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.093036  [21760/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.105716  [21824/24872]:   0%|          | 0/388 [00:15<?, ?it/s]
loss: 0.096315  [21888/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.143878  [21952/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.130243  [22016/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.123594  [22080/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.153876  [22144/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.109550  [22208/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.114910  [22272/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.115780  [22336/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.121093  [22400/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.097517  [22464/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.151244  [22528/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.135667  [22592/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.122397  [22656/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.150631  [22720/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.114776  [22784/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.124554  [22848/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.128749  [22912/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.158015  [22976/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.128684  [23040/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.141769  [23104/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.156516  [23168/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.149180  [23232/24872]:   0%|          | 0/388 [00:16<?, ?it/s]
loss: 0.122908  [23296/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.088543  [23360/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.128203  [23424/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.139995  [23488/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.141786  [23552/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.105066  [23616/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.130878  [23680/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.141802  [23744/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.093533  [23808/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.180557  [23872/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.113826  [23936/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.122585  [24000/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.151134  [24064/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.142562  [24128/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.202555  [24192/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.119424  [24256/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.097207  [24320/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.175097  [24384/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.175692  [24448/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.164534  [24512/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.090774  [24576/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.158462  [24640/24872]:   0%|          | 0/388 [00:17<?, ?it/s]
loss: 0.109507  [24704/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.108750  [24768/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.150023  [24832/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.137341  [24872/24872]:   0%|          | 0/388 [00:18<?, ?it/s]
loss: 0.137341  [24872/24872]: : 389it [00:18, 21.41it/s]
-------------------------------
LR=0.0001, batch_size=128
-------------------------------
Epoch 1, time=221.47s

  0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.102065  [  128/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.331610  [  256/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.199787  [  384/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.224357  [  512/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.164470  [  640/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.182419  [  768/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.213347  [  896/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.148940  [ 1024/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.149916  [ 1152/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.172707  [ 1280/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.204565  [ 1408/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.194021  [ 1536/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.163938  [ 1664/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.147339  [ 1792/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.141879  [ 1920/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.113090  [ 2048/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.138989  [ 2176/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.119712  [ 2304/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.150507  [ 2432/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.125325  [ 2560/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.123442  [ 2688/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.130600  [ 2816/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.132535  [ 2944/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.111733  [ 3072/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.099899  [ 3200/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.141626  [ 3328/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.159329  [ 3456/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.127703  [ 3584/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.119793  [ 3712/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.114686  [ 3840/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.096610  [ 3968/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.152884  [ 4096/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.126232  [ 4224/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.117396  [ 4352/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.123500  [ 4480/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.141051  [ 4608/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.146849  [ 4736/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.132412  [ 4864/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.136491  [ 4992/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.117772  [ 5120/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.134712  [ 5248/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.119512  [ 5376/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.090899  [ 5504/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.134919  [ 5632/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.131641  [ 5760/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.117105  [ 5888/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.106407  [ 6016/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.099250  [ 6144/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.105807  [ 6272/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.133904  [ 6400/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.086515  [ 6528/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.094013  [ 6656/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.127346  [ 6784/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.118383  [ 6912/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.096028  [ 7040/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.109692  [ 7168/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.144461  [ 7296/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.102176  [ 7424/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.096333  [ 7552/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.121181  [ 7680/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.112423  [ 7808/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.109708  [ 7936/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.112327  [ 8064/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.107154  [ 8192/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.080710  [ 8320/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.101484  [ 8448/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.132584  [ 8576/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.133966  [ 8704/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.125912  [ 8832/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.100994  [ 8960/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.108409  [ 9088/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.121872  [ 9216/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.101703  [ 9344/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.142903  [ 9472/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.130429  [ 9600/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.102574  [ 9728/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.115264  [ 9856/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.125984  [ 9984/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.094486  [10112/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.106058  [10240/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.145457  [10368/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.125492  [10496/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.122494  [10624/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.123823  [10752/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.151387  [10880/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.099478  [11008/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.117076  [11136/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.132623  [11264/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.115971  [11392/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.119461  [11520/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.117131  [11648/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.118389  [11776/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.114574  [11904/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.125415  [12032/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.141122  [12160/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.131142  [12288/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.137528  [12416/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.103174  [12544/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.112893  [12672/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.146115  [12800/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.155616  [12928/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.131617  [13056/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.134885  [13184/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.144271  [13312/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.122975  [13440/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.103106  [13568/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.115363  [13696/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.150815  [13824/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.150470  [13952/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.126454  [14080/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.100143  [14208/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.124654  [14336/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.114177  [14464/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.100465  [14592/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.102631  [14720/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.150439  [14848/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.110586  [14976/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.118816  [15104/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.118161  [15232/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.118365  [15360/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.118247  [15488/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.095313  [15616/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.134312  [15744/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128097  [15872/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.117941  [16000/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.121129  [16128/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.131006  [16256/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.112360  [16384/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.127935  [16512/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.120010  [16640/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.124547  [16768/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.106467  [16896/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.129461  [17024/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.129701  [17152/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.124858  [17280/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.120493  [17408/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.109631  [17536/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.122331  [17664/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.124408  [17792/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.119991  [17920/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.108666  [18048/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.092313  [18176/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.092528  [18304/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.106628  [18432/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.124358  [18560/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.118565  [18688/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.125183  [18816/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.120254  [18944/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.139219  [19072/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.101850  [19200/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.076799  [19328/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.122080  [19456/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.101299  [19584/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.144195  [19712/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.114404  [19840/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.129341  [19968/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.106297  [20096/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.128250  [20224/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.137869  [20352/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.100417  [20480/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.112787  [20608/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.142650  [20736/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.098094  [20864/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.108391  [20992/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.114875  [21120/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.113896  [21248/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.136390  [21376/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.127653  [21504/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.139634  [21632/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.088108  [21760/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.103748  [21888/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.129031  [22016/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.128951  [22144/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107883  [22272/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.099052  [22400/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.117687  [22528/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107025  [22656/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.116072  [22784/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.110716  [22912/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.124053  [23040/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.124104  [23168/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.122303  [23296/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.098751  [23424/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.134937  [23552/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.110529  [23680/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107109  [23808/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.147237  [23936/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.121447  [24064/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.161721  [24192/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.100271  [24320/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.155398  [24448/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.117763  [24576/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.123582  [24704/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.117096  [24832/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.116204  [24872/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.116204  [24872/24872]: : 195it [00:14, 13.38it/s]
Epoch 2, time=236.04s

  0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.113300  [  128/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.097131  [  256/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.099611  [  384/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.092595  [  512/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.101235  [  640/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.105183  [  768/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.126430  [  896/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.116845  [ 1024/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.107345  [ 1152/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.103474  [ 1280/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.148661  [ 1408/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.156676  [ 1536/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.133689  [ 1664/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.115599  [ 1792/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.124256  [ 1920/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.097871  [ 2048/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.117950  [ 2176/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.092909  [ 2304/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.147307  [ 2432/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.104196  [ 2560/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.105335  [ 2688/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.113534  [ 2816/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.118223  [ 2944/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.101305  [ 3072/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.085804  [ 3200/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.123679  [ 3328/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.148186  [ 3456/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.104898  [ 3584/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.107659  [ 3712/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.101829  [ 3840/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.099618  [ 3968/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.142953  [ 4096/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.117163  [ 4224/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.108348  [ 4352/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.115202  [ 4480/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.124994  [ 4608/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.124826  [ 4736/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.130559  [ 4864/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.136568  [ 4992/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.111108  [ 5120/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.129317  [ 5248/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.110368  [ 5376/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.091799  [ 5504/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.125525  [ 5632/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.124609  [ 5760/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.106282  [ 5888/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.114897  [ 6016/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.100961  [ 6144/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.105681  [ 6272/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.129332  [ 6400/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.082406  [ 6528/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.089396  [ 6656/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.123879  [ 6784/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.112891  [ 6912/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.095290  [ 7040/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.105645  [ 7168/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.133025  [ 7296/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.099815  [ 7424/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.093868  [ 7552/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.114891  [ 7680/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.111346  [ 7808/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.106749  [ 7936/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.112948  [ 8064/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.100141  [ 8192/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.077315  [ 8320/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.101636  [ 8448/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.136188  [ 8576/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.133090  [ 8704/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.113838  [ 8832/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.098151  [ 8960/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.111809  [ 9088/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.124773  [ 9216/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.103635  [ 9344/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.146737  [ 9472/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.130459  [ 9600/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.123371  [ 9728/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.114356  [ 9856/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.131760  [ 9984/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.093386  [10112/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.105001  [10240/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.148212  [10368/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.124060  [10496/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.121727  [10624/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.121427  [10752/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.153418  [10880/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.104752  [11008/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.115454  [11136/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.128891  [11264/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.107206  [11392/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.119727  [11520/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.114464  [11648/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.113332  [11776/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.108413  [11904/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.126817  [12032/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.129868  [12160/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.119627  [12288/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.134523  [12416/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.094236  [12544/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.108811  [12672/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.125831  [12800/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.132938  [12928/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.127857  [13056/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.134013  [13184/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.139488  [13312/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.117048  [13440/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.098808  [13568/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.111533  [13696/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.137683  [13824/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.132075  [13952/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.111726  [14080/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.092996  [14208/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.125852  [14336/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.110571  [14464/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.087702  [14592/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.108814  [14720/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.134664  [14848/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.112178  [14976/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.129296  [15104/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.111601  [15232/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.124103  [15360/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.116294  [15488/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.096540  [15616/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128902  [15744/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.130748  [15872/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.116604  [16000/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.122257  [16128/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.130657  [16256/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.112675  [16384/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.120498  [16512/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.115378  [16640/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128621  [16768/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.105589  [16896/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.124760  [17024/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.124564  [17152/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.122376  [17280/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.115986  [17408/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.106751  [17536/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.120048  [17664/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.124425  [17792/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.115474  [17920/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.109319  [18048/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.088298  [18176/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.088986  [18304/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.100045  [18432/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.113332  [18560/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.113601  [18688/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.121172  [18816/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.116263  [18944/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.132754  [19072/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.094777  [19200/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.069059  [19328/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.114822  [19456/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.100302  [19584/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.138399  [19712/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.105936  [19840/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.132045  [19968/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.102956  [20096/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.126329  [20224/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.133671  [20352/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.097707  [20480/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.112973  [20608/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.133218  [20736/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.093556  [20864/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.107182  [20992/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.111352  [21120/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.108340  [21248/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.136696  [21376/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.121169  [21504/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.125920  [21632/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.085901  [21760/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.099464  [21888/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.127047  [22016/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.119078  [22144/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.098591  [22272/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.097534  [22400/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107731  [22528/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.108190  [22656/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.109038  [22784/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.109390  [22912/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.122641  [23040/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.121321  [23168/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.116998  [23296/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.095769  [23424/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.126060  [23552/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.106698  [23680/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.102107  [23808/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.136256  [23936/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.120582  [24064/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.153293  [24192/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.096142  [24320/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.148625  [24448/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.113061  [24576/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.120934  [24704/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.111921  [24832/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.110355  [24872/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.110355  [24872/24872]: : 195it [00:14, 13.11it/s]
Epoch 3, time=250.92s

  0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.096954  [  128/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.097683  [  256/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.088089  [  384/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.090789  [  512/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.092663  [  640/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.098677  [  768/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.118046  [  896/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.108044  [ 1024/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.104584  [ 1152/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.091824  [ 1280/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.138349  [ 1408/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.150077  [ 1536/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.125488  [ 1664/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.099935  [ 1792/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.116337  [ 1920/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.088149  [ 2048/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.107045  [ 2176/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.086189  [ 2304/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.134689  [ 2432/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.097793  [ 2560/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.097264  [ 2688/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.106150  [ 2816/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.113722  [ 2944/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.098498  [ 3072/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.079425  [ 3200/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.117383  [ 3328/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.141105  [ 3456/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.099665  [ 3584/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.101788  [ 3712/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.099607  [ 3840/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.086038  [ 3968/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.130832  [ 4096/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.110861  [ 4224/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.100601  [ 4352/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.108675  [ 4480/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.120181  [ 4608/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.130298  [ 4736/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.127507  [ 4864/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.128794  [ 4992/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.107807  [ 5120/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.120907  [ 5248/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.103085  [ 5376/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.085756  [ 5504/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.120299  [ 5632/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.120598  [ 5760/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.107209  [ 5888/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.106201  [ 6016/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.102384  [ 6144/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.105438  [ 6272/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.123218  [ 6400/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.079734  [ 6528/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.085241  [ 6656/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.120605  [ 6784/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.110558  [ 6912/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.087797  [ 7040/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.102607  [ 7168/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.127310  [ 7296/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.096252  [ 7424/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.090855  [ 7552/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.112359  [ 7680/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.109705  [ 7808/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.103731  [ 7936/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.102587  [ 8064/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.095477  [ 8192/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.073549  [ 8320/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.095778  [ 8448/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.129901  [ 8576/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.118475  [ 8704/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.105934  [ 8832/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.093563  [ 8960/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.109036  [ 9088/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.115007  [ 9216/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.107032  [ 9344/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.132382  [ 9472/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.123087  [ 9600/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.097870  [ 9728/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.103403  [ 9856/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.121053  [ 9984/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.091537  [10112/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.094137  [10240/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.136060  [10368/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.122306  [10496/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.115561  [10624/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.113689  [10752/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.143878  [10880/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.104914  [11008/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.121381  [11136/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.121250  [11264/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.102094  [11392/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.110667  [11520/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.114743  [11648/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.112390  [11776/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.111826  [11904/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.115566  [12032/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.137346  [12160/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.118588  [12288/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.131477  [12416/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.094410  [12544/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.106617  [12672/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.118228  [12800/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.129073  [12928/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.127779  [13056/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.125362  [13184/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.137749  [13312/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.124674  [13440/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.090625  [13568/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.108676  [13696/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.134558  [13824/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.120446  [13952/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.105750  [14080/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.099327  [14208/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.141880  [14336/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.120728  [14464/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.095018  [14592/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.111279  [14720/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.143282  [14848/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.113757  [14976/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.112724  [15104/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.118307  [15232/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.113055  [15360/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.125545  [15488/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.099214  [15616/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.143000  [15744/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.137094  [15872/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.119216  [16000/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.109737  [16128/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.133208  [16256/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.124988  [16384/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.124992  [16512/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.123697  [16640/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.134832  [16768/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.119484  [16896/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.117999  [17024/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.140140  [17152/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.135267  [17280/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.140554  [17408/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.113962  [17536/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.129595  [17664/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.123681  [17792/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.120864  [17920/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.113784  [18048/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.106794  [18176/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.090865  [18304/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.098164  [18432/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.116752  [18560/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.123765  [18688/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.120764  [18816/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.114788  [18944/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.128041  [19072/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.096651  [19200/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.062914  [19328/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.113016  [19456/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.083551  [19584/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.145876  [19712/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.101768  [19840/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.134391  [19968/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.102175  [20096/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.138737  [20224/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.132121  [20352/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.095730  [20480/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.110845  [20608/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.130918  [20736/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.092338  [20864/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.109891  [20992/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.110863  [21120/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.125519  [21248/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.141726  [21376/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.122058  [21504/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.122348  [21632/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.092399  [21760/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.087095  [21888/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.128538  [22016/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.111202  [22144/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.096908  [22272/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.100415  [22400/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.101731  [22528/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.117081  [22656/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107138  [22784/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.108404  [22912/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.115160  [23040/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.120700  [23168/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.114867  [23296/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.094403  [23424/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.120703  [23552/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.102733  [23680/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.104479  [23808/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.137061  [23936/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.120804  [24064/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.153125  [24192/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.089586  [24320/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.149014  [24448/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.113225  [24576/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.124275  [24704/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.110244  [24832/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.123045  [24872/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.123045  [24872/24872]: : 195it [00:14, 13.12it/s]
Epoch 4, time=265.78s

  0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.096756  [  128/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.107352  [  256/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.085886  [  384/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.098765  [  512/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.093755  [  640/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.098198  [  768/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.121494  [  896/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.106525  [ 1024/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.110500  [ 1152/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.092605  [ 1280/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.157458  [ 1408/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.161342  [ 1536/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.140835  [ 1664/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.116468  [ 1792/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.118900  [ 1920/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.098761  [ 2048/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.112006  [ 2176/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.095536  [ 2304/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.125492  [ 2432/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.111010  [ 2560/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.098162  [ 2688/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.113324  [ 2816/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.119928  [ 2944/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.094204  [ 3072/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.099497  [ 3200/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.120153  [ 3328/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.146201  [ 3456/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.111419  [ 3584/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.118364  [ 3712/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.096193  [ 3840/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.103202  [ 3968/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.135065  [ 4096/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.127775  [ 4224/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.107352  [ 4352/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.108523  [ 4480/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.134602  [ 4608/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.121252  [ 4736/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.133659  [ 4864/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.126516  [ 4992/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.111729  [ 5120/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.123297  [ 5248/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.107491  [ 5376/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.088962  [ 5504/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.122149  [ 5632/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.120337  [ 5760/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.109571  [ 5888/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.108286  [ 6016/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.097112  [ 6144/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.102503  [ 6272/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.124918  [ 6400/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.078472  [ 6528/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.104014  [ 6656/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.119140  [ 6784/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.112840  [ 6912/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.085920  [ 7040/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.105314  [ 7168/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.123002  [ 7296/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.100296  [ 7424/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.093379  [ 7552/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.114263  [ 7680/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.107422  [ 7808/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.104760  [ 7936/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.100546  [ 8064/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.101958  [ 8192/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.085917  [ 8320/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.101682  [ 8448/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.127925  [ 8576/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.130935  [ 8704/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.104926  [ 8832/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.094743  [ 8960/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.105284  [ 9088/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.117332  [ 9216/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.103628  [ 9344/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.130088  [ 9472/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.124899  [ 9600/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.094564  [ 9728/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.110045  [ 9856/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.116985  [ 9984/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.092604  [10112/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.100393  [10240/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.139412  [10368/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.115528  [10496/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.113346  [10624/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.116092  [10752/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.144237  [10880/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.092267  [11008/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.107656  [11136/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.114136  [11264/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.097750  [11392/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.107567  [11520/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.109075  [11648/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.108172  [11776/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.101464  [11904/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.110130  [12032/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.120344  [12160/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.112124  [12288/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.124132  [12416/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.087986  [12544/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.099473  [12672/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.123566  [12800/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.123954  [12928/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.116944  [13056/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.117036  [13184/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.118147  [13312/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.111856  [13440/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.087041  [13568/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.101489  [13696/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.127636  [13824/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.114461  [13952/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.102878  [14080/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.088700  [14208/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.111816  [14336/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.104692  [14464/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.081103  [14592/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.097856  [14720/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.127599  [14848/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.096588  [14976/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.109895  [15104/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.098442  [15232/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.107553  [15360/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.115409  [15488/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.086839  [15616/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.122117  [15744/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.126912  [15872/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.110179  [16000/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.110977  [16128/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128239  [16256/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.117120  [16384/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.119047  [16512/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.134737  [16640/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.126077  [16768/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.118562  [16896/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.131354  [17024/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.126420  [17152/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.133826  [17280/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.117852  [17408/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.136055  [17536/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.112302  [17664/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.153007  [17792/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.133326  [17920/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.118413  [18048/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.123420  [18176/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.093302  [18304/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.128306  [18432/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.117494  [18560/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.110815  [18688/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.135433  [18816/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.116335  [18944/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.159105  [19072/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.090212  [19200/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.075181  [19328/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.119402  [19456/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.093096  [19584/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.151512  [19712/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.111975  [19840/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.140260  [19968/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.113898  [20096/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.125173  [20224/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.142947  [20352/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.095486  [20480/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.105982  [20608/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.125777  [20736/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.092211  [20864/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.106414  [20992/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.098733  [21120/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.111788  [21248/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.132474  [21376/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.115123  [21504/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.126813  [21632/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.092663  [21760/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.092715  [21888/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.129808  [22016/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.114347  [22144/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.099234  [22272/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.100864  [22400/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.106385  [22528/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.095406  [22656/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.113040  [22784/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.101414  [22912/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.124637  [23040/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.117448  [23168/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.120483  [23296/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.095758  [23424/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.123434  [23552/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.106879  [23680/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.099173  [23808/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.148823  [23936/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.116257  [24064/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.159532  [24192/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.092923  [24320/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.141847  [24448/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.117991  [24576/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.120844  [24704/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.117363  [24832/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.116124  [24872/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.116124  [24872/24872]: : 195it [00:14, 13.18it/s]
Epoch 5, time=280.58s

  0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.115844  [  128/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.100378  [  256/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.090694  [  384/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.099573  [  512/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.095083  [  640/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.107255  [  768/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.117548  [  896/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.108624  [ 1024/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.113164  [ 1152/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.089542  [ 1280/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.145806  [ 1408/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.149228  [ 1536/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.126997  [ 1664/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.104183  [ 1792/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.110546  [ 1920/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.085790  [ 2048/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.105668  [ 2176/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.088455  [ 2304/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.131154  [ 2432/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.104512  [ 2560/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.097099  [ 2688/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.104400  [ 2816/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.120866  [ 2944/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.087587  [ 3072/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.089136  [ 3200/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.119083  [ 3328/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.135781  [ 3456/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.103153  [ 3584/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.096046  [ 3712/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.093680  [ 3840/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.086882  [ 3968/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.122565  [ 4096/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.109490  [ 4224/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.098024  [ 4352/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.100903  [ 4480/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.110448  [ 4608/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.123838  [ 4736/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.111503  [ 4864/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.126345  [ 4992/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.101281  [ 5120/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.113082  [ 5248/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.101944  [ 5376/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.089466  [ 5504/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.116147  [ 5632/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.110142  [ 5760/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.108607  [ 5888/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.101732  [ 6016/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.096755  [ 6144/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.096582  [ 6272/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.118442  [ 6400/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.081793  [ 6528/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.081250  [ 6656/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.114391  [ 6784/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.105111  [ 6912/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.083011  [ 7040/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.103242  [ 7168/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.116583  [ 7296/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.093601  [ 7424/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.084752  [ 7552/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.109228  [ 7680/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.112835  [ 7808/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.097705  [ 7936/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.095044  [ 8064/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.114193  [ 8192/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.079954  [ 8320/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.108997  [ 8448/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.123775  [ 8576/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.131375  [ 8704/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.109882  [ 8832/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.108795  [ 8960/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.098487  [ 9088/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.118839  [ 9216/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.101684  [ 9344/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.139723  [ 9472/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.129994  [ 9600/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.102560  [ 9728/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.120184  [ 9856/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.111819  [ 9984/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.121248  [10112/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.097305  [10240/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.147231  [10368/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.116693  [10496/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.107043  [10624/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.117636  [10752/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.150229  [10880/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.106874  [11008/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.116672  [11136/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.120469  [11264/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.110693  [11392/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.105599  [11520/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.111599  [11648/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.105290  [11776/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.098895  [11904/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.112207  [12032/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.117486  [12160/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.110028  [12288/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.117838  [12416/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.087965  [12544/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.098587  [12672/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.113452  [12800/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.121801  [12928/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.110589  [13056/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.117060  [13184/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.125216  [13312/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.103456  [13440/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.085806  [13568/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.103650  [13696/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.133025  [13824/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.114824  [13952/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.104389  [14080/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.089311  [14208/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.114853  [14336/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.094787  [14464/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.077901  [14592/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.103719  [14720/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.126768  [14848/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.096822  [14976/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.105128  [15104/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.097575  [15232/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.100966  [15360/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.118623  [15488/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.085405  [15616/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.117987  [15744/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128160  [15872/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.103017  [16000/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.096314  [16128/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.126040  [16256/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.102918  [16384/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.111198  [16512/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.115375  [16640/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128673  [16768/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.096220  [16896/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.123914  [17024/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.107783  [17152/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.104656  [17280/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.115260  [17408/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.098696  [17536/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.117549  [17664/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.123185  [17792/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.104037  [17920/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.110019  [18048/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.089508  [18176/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.090695  [18304/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.095512  [18432/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.105345  [18560/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.106973  [18688/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.109787  [18816/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.121937  [18944/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.123698  [19072/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.097964  [19200/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.064143  [19328/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.110318  [19456/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.080656  [19584/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.143128  [19712/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.119599  [19840/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.126765  [19968/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.106622  [20096/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.121480  [20224/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.130495  [20352/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.094696  [20480/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.098179  [20608/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.120498  [20736/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.091066  [20864/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.099274  [20992/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.097236  [21120/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.113824  [21248/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.128941  [21376/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.108916  [21504/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.117775  [21632/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.082597  [21760/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.089668  [21888/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.121289  [22016/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.105502  [22144/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.096929  [22272/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.089778  [22400/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.104492  [22528/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.091171  [22656/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107200  [22784/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.098490  [22912/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.111727  [23040/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.120591  [23168/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107107  [23296/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.086979  [23424/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.115487  [23552/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.094233  [23680/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.100796  [23808/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.137111  [23936/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.113021  [24064/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.140540  [24192/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.084677  [24320/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.129860  [24448/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.107352  [24576/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.112017  [24704/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.109899  [24832/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.108950  [24872/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.108950  [24872/24872]: : 195it [00:14, 13.12it/s]
Epoch 6, time=295.44s

  0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.094967  [  128/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.092852  [  256/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.083793  [  384/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.088414  [  512/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.089219  [  640/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.092601  [  768/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.107566  [  896/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.098889  [ 1024/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.100956  [ 1152/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.088949  [ 1280/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.137061  [ 1408/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.141020  [ 1536/24872]:   0%|          | 0/194 [00:00<?, ?it/s]
loss: 0.109087  [ 1664/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.090863  [ 1792/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.107958  [ 1920/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.077258  [ 2048/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.103928  [ 2176/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.073435  [ 2304/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.115209  [ 2432/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.090897  [ 2560/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.089936  [ 2688/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.098759  [ 2816/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.104025  [ 2944/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.088001  [ 3072/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.068985  [ 3200/24872]:   0%|          | 0/194 [00:01<?, ?it/s]
loss: 0.110873  [ 3328/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.121597  [ 3456/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.094024  [ 3584/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.092091  [ 3712/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.084962  [ 3840/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.082405  [ 3968/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.113291  [ 4096/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.106960  [ 4224/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.095598  [ 4352/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.096253  [ 4480/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.110601  [ 4608/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.123506  [ 4736/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.109206  [ 4864/24872]:   0%|          | 0/194 [00:02<?, ?it/s]
loss: 0.115798  [ 4992/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.102137  [ 5120/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.110853  [ 5248/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.100926  [ 5376/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.078881  [ 5504/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.110166  [ 5632/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.102667  [ 5760/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.098693  [ 5888/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.087861  [ 6016/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.091287  [ 6144/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.085276  [ 6272/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.112775  [ 6400/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.070861  [ 6528/24872]:   0%|          | 0/194 [00:03<?, ?it/s]
loss: 0.078004  [ 6656/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.102484  [ 6784/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.104414  [ 6912/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.080589  [ 7040/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.098359  [ 7168/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.111638  [ 7296/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.090694  [ 7424/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.083486  [ 7552/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.107702  [ 7680/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.110021  [ 7808/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.092900  [ 7936/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.089678  [ 8064/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.088828  [ 8192/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.073757  [ 8320/24872]:   0%|          | 0/194 [00:04<?, ?it/s]
loss: 0.088428  [ 8448/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.123307  [ 8576/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.116308  [ 8704/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.105208  [ 8832/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.091716  [ 8960/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.105637  [ 9088/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.122445  [ 9216/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.088306  [ 9344/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.147845  [ 9472/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.110601  [ 9600/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.115853  [ 9728/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.118245  [ 9856/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.117655  [ 9984/24872]:   0%|          | 0/194 [00:05<?, ?it/s]
loss: 0.124049  [10112/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.088982  [10240/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.153339  [10368/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.099188  [10496/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.116878  [10624/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.104954  [10752/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.138818  [10880/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.105551  [11008/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.108314  [11136/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.117129  [11264/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.101283  [11392/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.100922  [11520/24872]:   0%|          | 0/194 [00:06<?, ?it/s]
loss: 0.108126  [11648/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.111383  [11776/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.090624  [11904/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.111136  [12032/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.110289  [12160/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.108065  [12288/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.113273  [12416/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.092639  [12544/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.098789  [12672/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.109282  [12800/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.130613  [12928/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.106804  [13056/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.122256  [13184/24872]:   0%|          | 0/194 [00:07<?, ?it/s]
loss: 0.110609  [13312/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.102183  [13440/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.083308  [13568/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.098691  [13696/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.121029  [13824/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.112879  [13952/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.091448  [14080/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.096151  [14208/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.101617  [14336/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.114470  [14464/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.069494  [14592/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.098557  [14720/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.115370  [14848/24872]:   0%|          | 0/194 [00:08<?, ?it/s]
loss: 0.092038  [14976/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.106428  [15104/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.101074  [15232/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.107940  [15360/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.114180  [15488/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.084886  [15616/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.128017  [15744/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.109341  [15872/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.103263  [16000/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.097838  [16128/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.126072  [16256/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.093179  [16384/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.108076  [16512/24872]:   0%|          | 0/194 [00:09<?, ?it/s]
loss: 0.104418  [16640/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.109279  [16768/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.092774  [16896/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.113791  [17024/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.099427  [17152/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.103176  [17280/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.119504  [17408/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.094728  [17536/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.107139  [17664/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.116493  [17792/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.102550  [17920/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.104084  [18048/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.083705  [18176/24872]:   0%|          | 0/194 [00:10<?, ?it/s]
loss: 0.080213  [18304/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.087474  [18432/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.095421  [18560/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.099598  [18688/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.100993  [18816/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.105371  [18944/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.112542  [19072/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.085394  [19200/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.055284  [19328/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.100460  [19456/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.077749  [19584/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.142974  [19712/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.095143  [19840/24872]:   0%|          | 0/194 [00:11<?, ?it/s]
loss: 0.126884  [19968/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.097993  [20096/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.119902  [20224/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.121859  [20352/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.091260  [20480/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.095418  [20608/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.121835  [20736/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.092546  [20864/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.099164  [20992/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.103987  [21120/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.107209  [21248/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.129023  [21376/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.107403  [21504/24872]:   0%|          | 0/194 [00:12<?, ?it/s]
loss: 0.112041  [21632/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.079336  [21760/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.100868  [21888/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.122640  [22016/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.107803  [22144/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.095178  [22272/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.086386  [22400/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.103029  [22528/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.099973  [22656/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.105806  [22784/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.098933  [22912/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.112793  [23040/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.113056  [23168/24872]:   0%|          | 0/194 [00:13<?, ?it/s]
loss: 0.108124  [23296/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.086306  [23424/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.118229  [23552/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.096134  [23680/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.097471  [23808/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.144805  [23936/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.108549  [24064/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.134784  [24192/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.083706  [24320/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.120986  [24448/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.107619  [24576/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.111443  [24704/24872]:   0%|          | 0/194 [00:14<?, ?it/s]
loss: 0.108050  [24832/24872]:   0%|          | 0/194 [00:15<?, ?it/s]
loss: 0.096211  [24872/24872]:   0%|          | 0/194 [00:15<?, ?it/s]
loss: 0.096211  [24872/24872]: : 195it [00:15, 12.93it/s]
-------------------------------
LR=1e-05, batch_size=256
-------------------------------
Epoch 1, time=310.52s

  0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.089531  [  256/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.080566  [  512/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.087837  [  768/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.098700  [ 1024/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.086621  [ 1280/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.132619  [ 1536/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.093953  [ 1792/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.090036  [ 2048/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.083601  [ 2304/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.098985  [ 2560/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.089251  [ 2816/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.089454  [ 3072/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.081735  [ 3328/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.101179  [ 3584/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.083114  [ 3840/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.090757  [ 4096/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.095073  [ 4352/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.101214  [ 4608/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.104909  [ 4864/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.103619  [ 5120/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.098726  [ 5376/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.085319  [ 5632/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.097417  [ 5888/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.084072  [ 6144/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.094260  [ 6400/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.068555  [ 6656/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.094457  [ 6912/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.076570  [ 7168/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.095289  [ 7424/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.094266  [ 7680/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.091407  [ 7936/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.082417  [ 8192/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.071220  [ 8448/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.107388  [ 8704/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.085514  [ 8960/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.095110  [ 9216/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.099748  [ 9472/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.088590  [ 9728/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.098378  [ 9984/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.079351  [10240/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.103312  [10496/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.096230  [10752/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.104445  [11008/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.095297  [11264/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.088344  [11520/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.091265  [11776/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.089725  [12032/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.094400  [12288/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.087975  [12544/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.089455  [12800/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.103856  [13056/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.099465  [13312/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.082870  [13568/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.095147  [13824/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.091638  [14080/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.084969  [14336/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.072866  [14592/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.095324  [14848/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.086139  [15104/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.087362  [15360/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.087141  [15616/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.108400  [15872/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.091377  [16128/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.100842  [16384/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.094304  [16640/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.092906  [16896/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.096426  [17152/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.097028  [17408/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.091846  [17664/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.097510  [17920/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.083967  [18176/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.072332  [18432/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.090885  [18688/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.096331  [18944/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.096293  [19200/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.073133  [19456/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.101423  [19712/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.103651  [19968/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.099003  [20224/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.094583  [20480/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.104035  [20736/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.082254  [20992/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.088935  [21248/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.111791  [21504/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.084730  [21760/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.088914  [22016/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.087958  [22272/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.082574  [22528/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.089447  [22784/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.097325  [23040/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.100521  [23296/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.089764  [23552/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.089145  [23808/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.112996  [24064/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.099458  [24320/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.102711  [24576/24872]:   0%|          | 0/97 [00:12<?, ?it/s]
loss: 0.096511  [24832/24872]:   0%|          | 0/97 [00:12<?, ?it/s]
loss: 0.090339  [24872/24872]:   0%|          | 0/97 [00:12<?, ?it/s]
loss: 0.090339  [24872/24872]: : 98it [00:12,  8.04it/s]
Epoch 2, time=322.72s

  0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.080289  [  256/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.075157  [  512/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.081295  [  768/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.093834  [ 1024/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.082590  [ 1280/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.124369  [ 1536/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.089112  [ 1792/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.085681  [ 2048/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.079811  [ 2304/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.096992  [ 2560/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.086928  [ 2816/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.087736  [ 3072/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.079691  [ 3328/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.100332  [ 3584/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.079591  [ 3840/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.087027  [ 4096/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.092707  [ 4352/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.098841  [ 4608/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.102053  [ 4864/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.101797  [ 5120/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.095074  [ 5376/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.083650  [ 5632/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.095563  [ 5888/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.082914  [ 6144/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.091946  [ 6400/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.066604  [ 6656/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.092647  [ 6912/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.075515  [ 7168/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.092690  [ 7424/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.091810  [ 7680/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.088756  [ 7936/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.078893  [ 8192/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.069752  [ 8448/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.105047  [ 8704/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.083537  [ 8960/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.092386  [ 9216/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.096536  [ 9472/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.087097  [ 9728/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.097027  [ 9984/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.077617  [10240/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.102766  [10496/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.093030  [10752/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.101340  [11008/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.094216  [11264/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.086612  [11520/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.089876  [11776/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.087471  [12032/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.091888  [12288/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.086811  [12544/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.086981  [12800/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.101705  [13056/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.097653  [13312/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.081341  [13568/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.093794  [13824/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.089500  [14080/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.082983  [14336/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.071204  [14592/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.093491  [14848/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.085107  [15104/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.086458  [15360/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.086136  [15616/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.105209  [15872/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.089930  [16128/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.099382  [16384/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.092220  [16640/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.091427  [16896/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.094799  [17152/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.095650  [17408/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.090177  [17664/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.096456  [17920/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.082110  [18176/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.070924  [18432/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.089456  [18688/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.094371  [18944/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.093514  [19200/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.071641  [19456/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.100476  [19712/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.101802  [19968/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.097002  [20224/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.093705  [20480/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.099813  [20736/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.080827  [20992/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.087365  [21248/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.111225  [21504/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.083654  [21760/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.088192  [22016/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.087151  [22272/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.081216  [22528/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.088710  [22784/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.096713  [23040/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.100248  [23296/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.089415  [23552/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.088779  [23808/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.112072  [24064/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.098627  [24320/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.101509  [24576/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.095638  [24832/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.089338  [24872/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.089338  [24872/24872]: : 98it [00:11,  8.32it/s]
Epoch 3, time=334.50s

  0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.079654  [  256/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.074132  [  512/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.080430  [  768/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.092950  [ 1024/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.081459  [ 1280/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.122081  [ 1536/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.087814  [ 1792/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.084755  [ 2048/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.078478  [ 2304/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.095789  [ 2560/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.085809  [ 2816/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.086632  [ 3072/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.078527  [ 3328/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.098992  [ 3584/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.078453  [ 3840/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.086187  [ 4096/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.091771  [ 4352/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.098055  [ 4608/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.100601  [ 4864/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.100977  [ 5120/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.094121  [ 5376/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.082597  [ 5632/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.094610  [ 5888/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.082401  [ 6144/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.090766  [ 6400/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.065793  [ 6656/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.091871  [ 6912/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.074930  [ 7168/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.091055  [ 7424/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.090710  [ 7680/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.088654  [ 7936/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.077853  [ 8192/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.068994  [ 8448/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.103736  [ 8704/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.082687  [ 8960/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.091376  [ 9216/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.094826  [ 9472/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.086516  [ 9728/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.095379  [ 9984/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.076911  [10240/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.102096  [10496/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.091891  [10752/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.100147  [11008/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.093447  [11264/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085836  [11520/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.088894  [11776/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.086644  [12032/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.090584  [12288/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.086233  [12544/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085542  [12800/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.100648  [13056/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.096761  [13312/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.079998  [13568/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.092775  [13824/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.088011  [14080/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.082052  [14336/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.070282  [14592/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.092825  [14848/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.084246  [15104/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.085916  [15360/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.085491  [15616/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.103636  [15872/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.089016  [16128/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.098404  [16384/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.091072  [16640/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.090612  [16896/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.093710  [17152/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.094771  [17408/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.089355  [17664/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.095645  [17920/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.080992  [18176/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.070169  [18432/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.088404  [18688/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.093167  [18944/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.090870  [19200/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.070892  [19456/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.099711  [19712/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.100405  [19968/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.095829  [20224/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.093181  [20480/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.097923  [20736/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.079594  [20992/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.086494  [21248/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.110798  [21504/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.082819  [21760/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.087578  [22016/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.086356  [22272/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.080052  [22528/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.087925  [22784/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.096119  [23040/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.099561  [23296/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.088850  [23552/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.088274  [23808/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.111136  [24064/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.097855  [24320/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.100635  [24576/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.094934  [24832/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.088525  [24872/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.088525  [24872/24872]: : 98it [00:11,  8.27it/s]
Epoch 4, time=346.35s

  0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.079140  [  256/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.073481  [  512/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.079753  [  768/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.092215  [ 1024/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.080737  [ 1280/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.120562  [ 1536/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.086745  [ 1792/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.084033  [ 2048/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.077624  [ 2304/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.094881  [ 2560/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.084992  [ 2816/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.085897  [ 3072/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.077661  [ 3328/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.098081  [ 3584/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.077657  [ 3840/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.085523  [ 4096/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.091087  [ 4352/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.097174  [ 4608/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.099332  [ 4864/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.100281  [ 5120/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.093397  [ 5376/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.081712  [ 5632/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.093509  [ 5888/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.081952  [ 6144/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.089764  [ 6400/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.064928  [ 6656/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.091209  [ 6912/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.074356  [ 7168/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.089855  [ 7424/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.089820  [ 7680/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.088225  [ 7936/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.077113  [ 8192/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.068483  [ 8448/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.102507  [ 8704/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.081888  [ 8960/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.090718  [ 9216/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.093473  [ 9472/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.085943  [ 9728/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.094021  [ 9984/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.076334  [10240/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.101331  [10496/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.090941  [10752/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.099322  [11008/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.092836  [11264/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085138  [11520/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.088066  [11776/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085904  [12032/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.089323  [12288/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085732  [12544/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.084259  [12800/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.099550  [13056/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.096084  [13312/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.078646  [13568/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.091869  [13824/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.086721  [14080/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.081251  [14336/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.069430  [14592/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.092256  [14848/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.083455  [15104/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.085413  [15360/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.085008  [15616/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.102491  [15872/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.088195  [16128/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.097655  [16384/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.090038  [16640/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.089806  [16896/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.092781  [17152/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.094247  [17408/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.088622  [17664/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.094957  [17920/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.080148  [18176/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.069572  [18432/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.087473  [18688/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.092206  [18944/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.088905  [19200/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.070290  [19456/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.099024  [19712/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.099456  [19968/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.094832  [20224/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.092716  [20480/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.096651  [20736/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.078737  [20992/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.085891  [21248/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.110470  [21504/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.082122  [21760/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.086850  [22016/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.085483  [22272/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.079352  [22528/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.087372  [22784/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.095543  [23040/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.098858  [23296/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.088332  [23552/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087827  [23808/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.110131  [24064/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.097124  [24320/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.099924  [24576/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.094286  [24832/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087859  [24872/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087859  [24872/24872]: : 98it [00:11,  8.33it/s]
Epoch 5, time=358.12s

  0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.078780  [  256/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.072808  [  512/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.079060  [  768/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.091505  [ 1024/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.080177  [ 1280/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.119312  [ 1536/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.085882  [ 1792/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.083382  [ 2048/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.076963  [ 2304/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.094100  [ 2560/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.084196  [ 2816/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.085251  [ 3072/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.076900  [ 3328/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.097281  [ 3584/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.076956  [ 3840/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.084978  [ 4096/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.090568  [ 4352/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.096347  [ 4608/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.098295  [ 4864/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.099645  [ 5120/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.092773  [ 5376/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.080933  [ 5632/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.092638  [ 5888/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.081524  [ 6144/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.088738  [ 6400/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.064154  [ 6656/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.090618  [ 6912/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.073823  [ 7168/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.088822  [ 7424/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.089050  [ 7680/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.087572  [ 7936/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.076483  [ 8192/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.067972  [ 8448/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.101483  [ 8704/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.081209  [ 8960/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.090100  [ 9216/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.092403  [ 9472/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.085348  [ 9728/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.092866  [ 9984/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.075831  [10240/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.100600  [10496/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.090083  [10752/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.098713  [11008/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.092347  [11264/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.084614  [11520/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.087298  [11776/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085253  [12032/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.088228  [12288/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.085262  [12544/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.083218  [12800/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.098637  [13056/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.095458  [13312/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.077553  [13568/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.091097  [13824/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.085672  [14080/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.080653  [14336/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.068578  [14592/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.091701  [14848/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.082767  [15104/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.084840  [15360/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.084506  [15616/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.101560  [15872/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.087510  [16128/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.096923  [16384/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.089071  [16640/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.089065  [16896/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.091920  [17152/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.093746  [17408/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.087872  [17664/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.094340  [17920/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.079415  [18176/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.069052  [18432/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.086641  [18688/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.091478  [18944/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.087617  [19200/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.069728  [19456/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.098403  [19712/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.098766  [19968/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.093935  [20224/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.092200  [20480/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.095678  [20736/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.078199  [20992/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.085232  [21248/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.110137  [21504/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.081369  [21760/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.086100  [22016/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.084651  [22272/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.078832  [22528/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.086770  [22784/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.094885  [23040/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.098241  [23296/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087817  [23552/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087365  [23808/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.109118  [24064/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.096498  [24320/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.099261  [24576/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.093759  [24832/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087381  [24872/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.087381  [24872/24872]: : 98it [00:11,  8.31it/s]
Epoch 6, time=369.92s

  0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.078452  [  256/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.072155  [  512/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.078461  [  768/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.090898  [ 1024/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.079598  [ 1280/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.118251  [ 1536/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.085082  [ 1792/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.082746  [ 2048/24872]:   0%|          | 0/97 [00:00<?, ?it/s]
loss: 0.076417  [ 2304/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.093533  [ 2560/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.083478  [ 2816/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.084608  [ 3072/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.076236  [ 3328/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.096626  [ 3584/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.076328  [ 3840/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.084499  [ 4096/24872]:   0%|          | 0/97 [00:01<?, ?it/s]
loss: 0.090119  [ 4352/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.095599  [ 4608/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.097335  [ 4864/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.099037  [ 5120/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.092240  [ 5376/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.080225  [ 5632/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.091825  [ 5888/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.081125  [ 6144/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.087822  [ 6400/24872]:   0%|          | 0/97 [00:02<?, ?it/s]
loss: 0.063461  [ 6656/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.090096  [ 6912/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.073393  [ 7168/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.087990  [ 7424/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.088287  [ 7680/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.086887  [ 7936/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.075972  [ 8192/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.067468  [ 8448/24872]:   0%|          | 0/97 [00:03<?, ?it/s]
loss: 0.100483  [ 8704/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.080593  [ 8960/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.089511  [ 9216/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.091402  [ 9472/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.084761  [ 9728/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.092085  [ 9984/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.075344  [10240/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.099838  [10496/24872]:   0%|          | 0/97 [00:04<?, ?it/s]
loss: 0.089234  [10752/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.098111  [11008/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.091917  [11264/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.084080  [11520/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.086605  [11776/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.084672  [12032/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.087138  [12288/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.084844  [12544/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.082191  [12800/24872]:   0%|          | 0/97 [00:05<?, ?it/s]
loss: 0.097774  [13056/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.094895  [13312/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.076668  [13568/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.090499  [13824/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.084766  [14080/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.080130  [14336/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.067746  [14592/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.091207  [14848/24872]:   0%|          | 0/97 [00:06<?, ?it/s]
loss: 0.082158  [15104/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.084164  [15360/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.084055  [15616/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.100673  [15872/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.086934  [16128/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.096299  [16384/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.088251  [16640/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.088436  [16896/24872]:   0%|          | 0/97 [00:07<?, ?it/s]
loss: 0.091153  [17152/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.093236  [17408/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.087231  [17664/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.093798  [17920/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.078804  [18176/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.068519  [18432/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.085910  [18688/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.090783  [18944/24872]:   0%|          | 0/97 [00:08<?, ?it/s]
loss: 0.086663  [19200/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.069221  [19456/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.097780  [19712/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.098182  [19968/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.093180  [20224/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.091663  [20480/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.094394  [20736/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.077677  [20992/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.084683  [21248/24872]:   0%|          | 0/97 [00:09<?, ?it/s]
loss: 0.109861  [21504/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.080627  [21760/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.085309  [22016/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.084034  [22272/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.078475  [22528/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.086366  [22784/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.094485  [23040/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.097704  [23296/24872]:   0%|          | 0/97 [00:10<?, ?it/s]
loss: 0.087260  [23552/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.086962  [23808/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.108112  [24064/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.095904  [24320/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.098742  [24576/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.093334  [24832/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.086912  [24872/24872]:   0%|          | 0/97 [00:11<?, ?it/s]
loss: 0.086912  [24872/24872]: : 98it [00:11,  8.37it/s]
Done!

test the network#

Do some qualitative tests: Let the trained network predict some particle geometries and compare their Mie spectra with the traget spectrum.

# pick a few of the training samples for testing.
# Note: Ideally tests should be done on separate samples!
sca_test = q_sca_target_test
pred = model(sca_test)

# evaluate Mie
r_c_test, r_s_test, eps_c_test, eps_s_test = nn_pred_to_mie_geometry(pred)
res_mie = pmd.multishell.cross_sections(
    k0,
    r_c=r_c_test,
    eps_c=eps_c_test,
    r_s=r_s_test,
    eps_s=eps_s_test,
    eps_env=eps_env,
    n_max=n_max,
)

# plot
i_plot = np.random.randint(len(sca_test), size=4)
plt.figure(figsize=(12, 10))
for i_n, i in enumerate(i_plot):
    plt.subplot(2, 2, i_n + 1)
    plt.plot(
        wl0.detach().cpu().numpy(),
        sca_test[i].detach().cpu().numpy(),
        label="reference",
    )
    plt.plot(
        wl0.detach().cpu().numpy(),
        res_mie["q_sca"][i].detach().cpu().numpy(),
        label="predicted particle",
    )
    plt.legend()
    plt.xlabel("wavelength (nm)")
    plt.ylabel("scat. efficiency")
plt.show()
ex 09 tandem

Total running time of the script: (6 minutes 27.181 seconds)

Estimated memory usage: 2013 MB

Gallery generated by Sphinx-Gallery