eecabe35-2224-4988-9910-8e8434a0c281.txt

import os
import sys
with open(sys.argv[0]) as f:
    code = f.read() # read the code of this file ASAP, for logging
import uuid
import glob
import time
import contextlib
from dataclasses import dataclass

import numpy as np
import torch
from torch import nn
import torch.nn.functional as F
import torch.distributed as dist
import torch._inductor.config as config
from torch.nn.parallel import DistributedDataParallel as DDP
# Use of FlexAttention contributed by @KoszarskyB
from torch.nn.attention.flex_attention import flex_attention, create_block_mask
flex_attention = torch.compile(flex_attention, dynamic=False)
create_block_mask = torch.compile(create_block_mask, dynamic=False)

# -----------------------------------------------------------------------------
# Muon optimizer

def zeropower_via_svd(G, steps=None):
    U, S, V = G.svd()
    return U @ V.T

@torch.compile
def zeropower_via_newtonschulz5(G, steps=10, eps=1e-7):
    """
    Newton-Schulz iteration to compute the zeroth power / orthogonalization of G. We opt to use a
    quintic iteration whose coefficients are selected to maximize the slope at zero. For the purpose
    of minimizing steps, it turns out to be empirically effective to keep increasing the slope at
    zero even beyond the point where the iteration no longer converges all the way to one everywhere
    on the interval. This iteration therefore does not produce UV^T but rather something like US'V^T
    where S' is diagonal with S_{ii}' ~ Uniform(0.5, 1.5), which turns out not to hurt model
    performance at all relative to UV^T, where USV^T = G is the SVD.
    """
    assert len(G.shape) == 2
    a, b, c = (3.4445, -4.7750,  2.0315)
    X = G.bfloat16()
    X /= (X.norm() + eps) # ensure top singular value <= 1
    if G.size(0) > G.size(1):
        X = X.T
    for _ in range(steps):
        A = X @ X.T
        B = b * A + c * A @ A # adapted from suggestion by @jxbz, @leloykun, and @YouJiacheng
        X = a * X + B @ X
    if G.size(0) > G.size(1):
        X = X.T
    return X

zeropower_backends = dict(svd=zeropower_via_svd, newtonschulz5=zeropower_via_newtonschulz5)

class Muon(torch.optim.Optimizer):
    """
    Muon - MomentUm Orthogonalized by Newton-schulz

    Muon internally runs standard SGD-momentum, and then performs an orthogonalization post-
    processing step, in which each 2D parameter's update is replaced with the nearest orthogonal
    matrix. To efficiently orthogonalize each update, we use a Newton-Schulz iteration, which has
    the advantage that it can be stably run in bfloat16 on the GPU.

    Some warnings:
    - This optimizer assumes that all parameters passed in are 2D.
    - It should not be used for the embedding layer, the final fully connected layer, or any {0,1}-D
    parameters; those should all be optimized by a standard method (e.g., AdamW).
    - To use it with 4D convolutional filters, it works well to just flatten their last 3 dimensions.
    - We believe it is unlikely to work well for training with small batch size.
    - We believe it may not work well for finetuning pretrained models, but we haven't tested this.
    - We have not yet tried this optimizer for training scenarios larger than NanoGPT (124M).

    Arguments:
        lr: The learning rate used by the internal SGD.
        momentum: The momentum used by the internal SGD.
        nesterov: Whether to use Nesterov-style momentum in the internal SGD. (recommended)
        backend: The chosen backend for the orthogonalization step. (recommended: 'newtonschulz5')
        backend_steps: The number of iteration steps to use in the backend, if it is iterative.
    """
    def __init__(self, params, lr=0.02, momentum=0.95, nesterov=True,
                 backend='newtonschulz5', backend_steps=5):
        defaults = dict(lr=lr, momentum=momentum, nesterov=nesterov, backend=backend, backend_steps=backend_steps)
        super().__init__(params, defaults)

    def step(self):

        for group in self.param_groups:

            lr = group['lr']
            momentum = group['momentum']
            zeropower_backend = zeropower_backends[group['backend']]

            # generate weight updates in distributed fashion
            total_params = sum(p.numel() for p in group['params'])
            updates_flat = torch.zeros(total_params, device='cuda', dtype=torch.bfloat16)
            curr_idx = 0
            for i, p in enumerate(group['params']):
                # luckily this will perfectly distribute a transformer with multiple of 4 layers to 8 GPUs
                if i % int(os.environ['WORLD_SIZE']) == int(os.environ['RANK']):
                    g = p.grad
                    assert g is not None
                    state = self.state[p]
                    if 'momentum_buffer' not in state:
                        state['momentum_buffer'] = torch.zeros_like(g)
                    buf = state['momentum_buffer']
                    buf.mul_(momentum).add_(g)
                    g = g.add(buf, alpha=momentum) if group['nesterov'] else buf
                    g = zeropower_backend(g, steps=group['backend_steps'])
                    g *= max(1, g.size(0)/g.size(1))**0.5
                    updates_flat[curr_idx:curr_idx+p.numel()] = g.flatten()
                curr_idx += p.numel()

            # sync updates across devices. we are not memory-constrained so can do this simple deserialization
            dist.all_reduce(updates_flat, op=dist.ReduceOp.SUM)

            # deserialize and apply updates
            curr_idx = 0
            for p in group['params']:
                g = updates_flat[curr_idx:curr_idx+p.numel()].view_as(p.data).type_as(p.data)
                p.data.add_(g, alpha=-lr)
                curr_idx += p.numel()

# -----------------------------------------------------------------------------
# PyTorch nn.Module definitions for the GPT-2 model

def norm(x):
    return F.rms_norm(x, (x.size(-1),))

class CastedLinear(nn.Linear):

    def __init__(self, in_features, out_features):
        super().__init__(in_features, out_features, bias=False)

    def forward(self, x):
        return F.linear(x, self.weight.to(x.dtype))

class Rotary(torch.nn.Module):

    def __init__(self, dim, base=10000):
        super().__init__()
        self.register_buffer('inv_freq', (1 / base) ** (torch.arange(0, dim, 2) / dim))
        self.seq_len_cached = None
        self.cos_cached = None
        self.sin_cached = None

    def forward(self, x):
        seq_len = x.shape[1]
        if seq_len != self.seq_len_cached:
            t = torch.arange(seq_len, device=x.device)
            freqs = torch.outer(t, self.inv_freq)
            self.seq_len_cached = seq_len
            self.cos_cached = freqs.cos()
            self.sin_cached = freqs.sin()
        cos, sin = self.cos_cached[None, :, None, :], self.sin_cached[None, :, None, :]
        # apply_rotary_emb(x, cos, sin)
        x1, x2 = x.chunk(2, dim=3)
        y1 = x1 * cos + x2 * sin
        y2 = x1 * (-sin) + x2 * cos
        return torch.cat((y1, y2), 3).type_as(x)

class CausalSelfAttention(nn.Module):

    def __init__(self, dim, n_head):
        super().__init__()
        assert dim % n_head == 0
        self.n_head = n_head
        self.c_q = CastedLinear(dim, dim)
        self.c_k = CastedLinear(dim, dim)
        self.c_v = CastedLinear(dim, dim)
        # value residual lambda
        self.lamb = nn.Parameter(torch.tensor(0.5)) # @Grad62304977
        # rotary embeddings
        self.rotary = Rotary(dim // n_head) # dim // n_head = head_dim
        # output projection
        self.c_proj = CastedLinear(dim, dim)
        self.c_proj.weight.data.zero_() # zero init suggested by @Grad62304977

    def forward(self, x, vi, block_mask):
        B, T = x.size(0), x.size(1) # batch size, sequence length
        assert B == 1, "Must use batch size = 1 for FlexAttention"
        q = self.c_q(x).view(B, T, self.n_head, -1)
        k = self.c_k(x).view(B, T, self.n_head, -1)
        v = self.c_v(x).view(B, T, self.n_head, -1)
        v = (1 - self.lamb) * v + self.lamb * vi.view_as(v) # @Grad62304977
        q, k = norm(q), norm(k) # QK norm suggested by @Grad62304977
        q, k = self.rotary(q), self.rotary(k)
        y = flex_attention(q.transpose(1, 2), k.transpose(1, 2), v.transpose(1, 2), block_mask=block_mask)
        y = y.transpose(1, 2).contiguous().view_as(x) # re-assemble all head outputs side by side
        y = self.c_proj(y)
        return y

class MLP(nn.Module):

    def __init__(self, dim):
        super().__init__()
        self.c_fc   = CastedLinear(dim, 4 * dim)
        self.c_proj = CastedLinear(4 * dim, dim)
        self.c_proj.weight.data.zero_() # zero init suggested by @Grad62304977

    def forward(self, x):
        x = self.c_fc(x)
        x = F.relu(x).square() # https://arxiv.org/abs/2109.08668v2; ~1-2% better than GELU; suggested by @SKYLINEZ007 and @Grad62304977
        x = self.c_proj(x)
        return x

class Block(nn.Module):

    def __init__(self, config):
        super().__init__()
        self.attn = CausalSelfAttention(config.n_embd, config.n_head)
        self.mlp = MLP(config.n_embd)
        self.lambdas = nn.Parameter(torch.tensor([1., 0.]))

    def forward(self, x, vi, x0, block_mask):
        x = self.lambdas[0] * x + self.lambdas[1] * x0
        x = x + self.attn(norm(x), vi, block_mask)
        x = x + self.mlp(norm(x))
        return x

# -----------------------------------------------------------------------------
# The main GPT-2 model

@dataclass
class GPTConfig:
    vocab_size : int = 50304
    n_layer : int = 12
    n_head : int = 6 # head dim 128 suggested by @Grad62304977
    n_embd : int = 768

class GPT(nn.Module):

    def __init__(self, config):
        super().__init__()

        # U-net design by @brendanh0gan
        self.num_encoder_layers = config.n_layer // 2 # Half of the layers for encoder
        self.num_decoder_layers = config.n_layer - self.num_encoder_layers # Remaining for decoder
        # Add learnable skip connection weights for decoder layers
        self.skip_weights = nn.Parameter(torch.ones(self.num_decoder_layers))

        self.transformer = nn.ModuleDict(dict(
            wte = nn.Embedding(config.vocab_size, config.n_embd),
            # token value embeddings by @KoszarskyB - inspired by @Grad62304977's value residual learning
            vte = nn.Embedding(config.vocab_size, config.n_embd*12),
            h = nn.ModuleList([Block(config) for _ in range(config.n_layer)]),
        ))
        self.lm_head = CastedLinear(config.n_embd, config.vocab_size)
        self.lm_head.weight.data.zero_() # @Grad62304977

    def forward(self, idx, target, attn_blocksize):

        docs = (idx == 50256).cumsum(0)
        def document_causal_mask(b, h, q_idx, kv_idx):
          causal_mask = q_idx >= kv_idx
          document_mask = docs[q_idx] == docs[kv_idx]
          window_mask = q_idx - kv_idx < attn_blocksize
          return causal_mask & document_mask & window_mask

        S = len(idx)
        block_mask = create_block_mask(document_causal_mask, None, None, S, S, device="cuda", _compile=True)

        # forward the GPT model itself
        x = self.transformer.wte(idx[None]) # token embeddings of shape (b, t, n_embd)
        x = norm(x) # @Grad62304977
        x0 = x
        vi = self.transformer.vte(idx[None]).chunk(12, dim=-1)

        # Store outputs for U-Net skip connections
        skip_connections = []
        # Encoder pass - process only the first half of the blocks
        for i in range(self.num_encoder_layers):
            x = self.transformer.h[i](x, vi[i], x0, block_mask)
            skip_connections.append(x)
        # Decoder pass - process the remaining blocks with weighted skip connections
        for i in range(self.num_decoder_layers):
            x = x + self.skip_weights[i] * skip_connections.pop()
            x = self.transformer.h[self.num_encoder_layers + i](x, vi[self.num_encoder_layers+i], x0, block_mask)

        x = norm(x)
        logits = self.lm_head(x)
        logits = 30 * torch.tanh(logits / 30) # @Grad62304977
        logits = logits.float()
        loss = F.cross_entropy(logits.view(-1, logits.size(-1)), target.view(-1))
        return loss

# -----------------------------------------------------------------------------
# Our own simple Distributed Data Loader

def _peek_data_shard(filename):
    # only reads the header, returns header data
    with open(filename, "rb") as f:
        # first read the header, which is 256 int32 integers (4 bytes each)
        header = np.frombuffer(f.read(256*4), dtype=np.int32)
    if header[0] != 20240520:
        print("ERROR: magic number mismatch in the data .bin file!")
        print("---> HINT: Are you passing in a correct file with --input_bin?")
        print("---> HINT: Dataset encoding changed recently, re-run data prepro or refer again to README")
        print("---> HINT: For example re-run: `python dev/data/tinyshakespeare.py`, then re-try")
        exit(1)
    assert header[1] == 1, "unsupported version"
    ntok = header[2] # number of tokens (claimed)
    return ntok # for now just return the number of tokens

def _load_data_shard(filename):
    with open(filename, "rb") as f:
        # first read the header, which is 256 int32 integers (4 bytes each)
        header = np.frombuffer(f.read(256*4), dtype=np.int32)
        assert header[0] == 20240520, "magic number mismatch in the data .bin file"
        assert header[1] == 1, "unsupported version"
        ntok = header[2] # number of tokens (claimed)
        # the rest of it are tokens, stored as uint16
        tokens = np.frombuffer(f.read(), dtype=np.uint16)
    assert len(tokens) == ntok, "number of tokens read does not match header?"
    return tokens

class DistributedDataLoader:
    def __init__(self, filename_pattern, T, process_rank, num_processes):
        self.process_rank = process_rank
        self.num_processes = num_processes
        self.T = T

        # glob files that match the pattern
        self.files = sorted(glob.glob(filename_pattern))
        assert len(self.files) > 0, f"did not find any files that match the pattern {filename_pattern}"

        # load and validate all data shards, count number of tokens in total
        ntok_total = 0
        for fname in self.files:
            shard_ntok = _peek_data_shard(fname)
            assert shard_ntok >= num_processes * T + 1
            ntok_total += int(shard_ntok)
        self.ntok_total = ntok_total

        self.reset()

    def reset(self):
        self.current_shard = -1
        self.advance()

    def advance(self): # advance to next data shard
        self.current_shard = (self.current_shard + 1) % len(self.files)
        self.current_position = self.process_rank * self.T
        self.tokens = _load_data_shard(self.files[self.current_shard])

    def next_batch(self):
        batch_size = self.T * self.num_processes
        buf = self.tokens[self.current_position:self.current_position+self.T+1]
        buf = torch.tensor(buf.astype(np.int32), dtype=torch.long)
        x = buf[:-1] # inputs
        y = buf[1:] # targets
        # advance current position and load next shard if necessary
        self.current_position += batch_size
        if self.current_position + batch_size >= len(self.tokens):
            self.advance()
        return x.cuda(), y.cuda()

# -----------------------------------------------------------------------------
# int main

@dataclass
class Hyperparameters:
    # data hyperparams
    input_bin : str = 'data/fineweb10B/fineweb_train_*.bin' # input .bin to train on
    input_val_bin : str = 'data/fineweb10B/fineweb_val_*.bin' # input .bin to eval validation loss on
    # optimization hyperparams
    batch_size : int = 8 # batch size, in sequences, across all devices
    sequence_length : int = 64*1024 # sequence length, in tokens
    num_iterations : int = 1530 # number of iterations to run
    warmup_iters : int = 0
    cooldown_iters : int = 600 # number of iterations of linear warmup/cooldown for triangular or trapezoidal schedule
    weight_decay : float = 0
    # evaluation and logging hyperparams
    val_loss_every : int = 125 # every how many steps to evaluate val loss? 0 for only at the end
    val_tokens : int = 10485760 # how many tokens of validation data? it's important to keep this fixed for consistent comparisons
    save_every : int = 0 # every how many steps to save the checkpoint? 0 for only at the end
args = Hyperparameters()

# set up DDP (distributed data parallel). torchrun sets this env variable
assert torch.cuda.is_available()
dist.init_process_group(backend='nccl')
ddp_rank = int(os.environ['RANK'])
ddp_local_rank = int(os.environ['LOCAL_RANK'])
ddp_world_size = int(os.environ['WORLD_SIZE'])
device = f'cuda:{ddp_local_rank}'
torch.cuda.set_device(device)
print(f"using device: {device}")
master_process = (ddp_rank == 0) # this process will do logging, checkpointing etc.

# begin logging
logfile = None
if master_process:
    run_id = str(uuid.uuid4())
    logdir = 'logs/%s/' % run_id
    os.makedirs(logdir, exist_ok=True)
    logfile = 'logs/%s.txt' % run_id
    # create the log file
    with open(logfile, "w") as f:
        # begin the log by printing this file (the Python code)
        f.write(code)
        f.write('='*100 + '\n')
def print0(s, logonly=False):
    if master_process:
        with open(logfile, "a") as f:
            if not logonly:
                print(s)
            f.write(s+'\n')
# log information about the hardware/software environment this is running on
# and print the full `nvidia-smi` to file
print0(f"Running pytorch {torch.version.__version__} compiled for CUDA {torch.version.cuda}\nnvidia-smi:")
import subprocess
result = subprocess.run(['nvidia-smi'], stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
print0(f'{result.stdout}', logonly=True)
print0('='*100, logonly=True)

# convenience variables
T = args.sequence_length
# calculate the number of steps to take in the val loop.
assert args.val_tokens % (T * ddp_world_size) == 0
val_steps = args.val_tokens // (T * ddp_world_size)
# calculate the steps of gradient accumulation required to attain the desired global batch size.
assert args.batch_size % (ddp_world_size) == 0
train_accumulation_steps = args.batch_size // ddp_world_size

# load tokens
train_loader = DistributedDataLoader(args.input_bin, T, ddp_rank, ddp_world_size)
val_loader = DistributedDataLoader(args.input_val_bin, T, ddp_rank, ddp_world_size)
print0(f"Training DataLoader: total number of tokens: {train_loader.ntok_total} across {len(train_loader.files)} files")
print0(f"Validation DataLoader: total number of tokens: {val_loader.ntok_total} across {len(val_loader.files)} files")
print0('='*100, logonly=True)
x, y = train_loader.next_batch()

# there are only 50257 unique GPT-2 tokens; we extend to nearest multiple of 128 for efficiency. suggested to me by @Grad62304977.
# this originates from Karpathy's experiments.
num_vocab = 50304
model = GPT(GPTConfig(vocab_size=num_vocab, n_layer=12, n_head=6, n_embd=768))
model = model.cuda().bfloat16()
for m in model.modules():
    if isinstance(m, CastedLinear):
        m.float()
if hasattr(config, "coordinate_descent_tuning"):
    config.coordinate_descent_tuning = True # suggested by @Chillee
model = torch.compile(model)
# here we wrap model into DDP container
model = DDP(model, device_ids=[ddp_local_rank])
raw_model = model.module # always contains the "raw" unwrapped model

# init the optimizer(s)
optimizer1 = torch.optim.Adam([raw_model.transformer.wte.weight, raw_model.transformer.vte.weight], lr=0.6, betas=(0.8, 0.95), fused=True)
optimizer2 = torch.optim.Adam([raw_model.lm_head.weight], lr=0.008, betas=(0.8, 0.95), fused=True)
params = list(raw_model.transformer.h.parameters())
matrix_params = [p for p in params if p.ndim == 2]
scalar_params = [p for p in params if p.ndim < 2] + [raw_model.skip_weights]
optimizer3 = Muon(matrix_params, lr=0.05, momentum=0.95)
optimizer4 = torch.optim.Adam(scalar_params, lr=0.04, betas=(0.8, 0.95), fused=True) # note that this learning rate is neither sensitive nor tuned
optimizers = [optimizer1, optimizer2, optimizer3, optimizer4]
# learning rate decay scheduler (linear warmup and cooldown)
def get_lr(it):
    assert it <= args.num_iterations
    # 1) linear warmup for warmup_iters steps
    if it < args.warmup_iters:
        return (it+1) / args.warmup_iters
    # 2) constant lr for a while
    elif it < args.num_iterations - args.cooldown_iters:
        return 1.0
    # 3) linear cooldown
    else:
        decay_ratio = (args.num_iterations - it) / args.cooldown_iters
        return decay_ratio
schedulers = [torch.optim.lr_scheduler.LambdaLR(opt, get_lr) for opt in optimizers]

# Start training loop
training_time_ms = 0
# start the clock
torch.cuda.synchronize()
t0 = time.time()
# begin training
for step in range(args.num_iterations + 1):
    last_step = (step == args.num_iterations)
    # This effectively ignores timing first 10 steps, which are slower for weird reasons.
    # Alternately, and slightly more correctly in terms of benchmarking, we could do 10
    # steps with dummy data first, and then re-initialize the model and reset the loader.
    if step == 10:
        training_time_ms = 0
        t0 = time.time()
    timed_steps = float('nan') if step <= 11 else (step - 10) + 1 # <= 11 to avoid bug in val

    # Set the attention blocksize for the current step, in chunks of 64. By @fernbear.bsky.social
    attn_blocksize = torch.tensor(64*((step/args.num_iterations * (1792 - 64) + 64)//64), dtype=torch.int, device='cuda')

    # once in a while evaluate the validation dataset
    if (last_step or (args.val_loss_every > 0 and step % args.val_loss_every == 0)):
        # stop the clock
        torch.cuda.synchronize()
        training_time_ms += 1000 * (time.time() - t0)
        # run validation batches
        model.eval()
        val_loader.reset()
        val_loss = 0.0
        for _ in range(val_steps):
            with torch.no_grad():
                x_val, y_val = val_loader.next_batch()
                val_loss += model(x_val, y_val, attn_blocksize=attn_blocksize)
        dist.all_reduce(val_loss, op=dist.ReduceOp.AVG)
        val_loss /= val_steps
        # log val loss to console and to logfile
        print0(f'step:{step}/{args.num_iterations} val_loss:{val_loss:.4f} train_time:{training_time_ms:.0f}ms step_avg:{training_time_ms/(timed_steps-1):.2f}ms')
        # start the clock again
        torch.cuda.synchronize()
        t0 = time.time()

    if master_process and (last_step or (args.save_every > 0 and step % args.save_every == 0)):
        # stop the clock
        torch.cuda.synchronize()
        training_time_ms += 1000 * (time.time() - t0)
        # save the state of the training process
        log = dict(step=step, code=code, model=raw_model.state_dict(), optimizers=[opt.state_dict() for opt in optimizers])
        torch.save(log, 'logs/%s/state_step%06d.pt' % (run_id, step))
        # start the clock again
        torch.cuda.synchronize()
        t0 = time.time()

    # bit confusing: we want to make sure to eval on 0th iteration
    # but also after the very last iteration. so we loop for step <= num_iterations
    # instead of just < num_iterations (one extra due to <=), only to do
    # the validation/sampling one last time, and then we break right here as we're done.
    if last_step:
        break

    # --------------- TRAINING SECTION BEGIN -----------------
    model.train()
    for i in range(1, train_accumulation_steps+1):
        ctx = model.no_sync() if i < train_accumulation_steps else contextlib.nullcontext()
        with ctx: # there's no need to sync gradients every accumulation step
            # forward pass
            loss = model(x, y, attn_blocksize=attn_blocksize)
            # advance the dataset for the next batch
            x, y = train_loader.next_batch()
            # backward pass
            loss.backward()
        train_loss = loss.detach()
    for p in model.parameters():
        p.grad /= train_accumulation_steps
    # momentum warmup for Muon
    frac = min(step/300, 1)
    optimizer3.param_groups[0]['momentum'] = (1 - frac) * 0.85 + frac * 0.95
    # step the optimizers and schedulers
    for opt, sched in zip(optimizers, schedulers):
        opt.step()
        sched.step()
    # null the gradients
    model.zero_grad(set_to_none=True)
    # --------------- TRAINING SECTION END -------------------
    # everything that follows now is just diagnostics, prints, logging, etc.

    #dist.all_reduce(train_loss, op=dist.ReduceOp.AVG) # all-reducing the training loss would be more correct in terms of logging, but slower
    approx_time = training_time_ms + 1000 * (time.time() - t0)
    print0(f"step:{step+1}/{args.num_iterations} train_loss:{train_loss.item():.4f} train_time:{approx_time:.0f}ms step_avg:{approx_time/timed_steps:.2f}ms")

if master_process:
    print(f"peak memory consumption: {torch.cuda.max_memory_allocated() // 1024 // 1024} MiB")

# -------------------------------------------------------------------------
# clean up nice
dist.destroy_process_group()
====================================================================================================
Running pytorch 2.6.0.dev20241203+cu124 compiled for CUDA 12.4
nvidia-smi:
Thu Dec  5 00:51:23 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.06             Driver Version: 535.183.06   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA H100 80GB HBM3          On  | 00000000:19:00.0 Off |                    0 |
| N/A   32C    P0              74W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          On  | 00000000:3B:00.0 Off |                    0 |
| N/A   27C    P0             100W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          On  | 00000000:4C:00.0 Off |                    0 |
| N/A   28C    P0             116W / 700W |    529MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   31C    P0             105W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   32C    P0             117W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   26C    P0             108W / 700W |    529MiB / 81559MiB |      2%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   32C    P0             121W / 700W |    115MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   27C    P0             116W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
+---------------------------------------------------------------------------------------+

====================================================================================================
Training DataLoader: total number of tokens: 1100000000 across 11 files
Validation DataLoader: total number of tokens: 100000000 across 1 files
====================================================================================================
step:0/1530 val_loss:10.8258 train_time:0ms step_avg:nanms
step:1/1530 train_loss:10.8258 train_time:32296ms step_avg:nanms
step:2/1530 train_loss:10.0731 train_time:32407ms step_avg:nanms
step:3/1530 train_loss:8.3621 train_time:32568ms step_avg:nanms
step:4/1530 train_loss:7.5801 train_time:32728ms step_avg:nanms
step:5/1530 train_loss:7.5010 train_time:32887ms step_avg:nanms
step:6/1530 train_loss:6.9784 train_time:33047ms step_avg:nanms
step:7/1530 train_loss:7.2527 train_time:33207ms step_avg:nanms
step:8/1530 train_loss:6.7521 train_time:33367ms step_avg:nanms
step:9/1530 train_loss:6.6315 train_time:33527ms step_avg:nanms
step:10/1530 train_loss:6.5198 train_time:33687ms step_avg:nanms
step:11/1530 train_loss:6.4634 train_time:114ms step_avg:nanms
step:12/1530 train_loss:6.3560 train_time:273ms step_avg:nanms
step:13/1530 train_loss:6.2161 train_time:434ms step_avg:144.65ms
step:14/1530 train_loss:6.1988 train_time:593ms step_avg:148.28ms
step:15/1530 train_loss:6.1283 train_time:753ms step_avg:150.66ms
step:16/1530 train_loss:6.1145 train_time:913ms step_avg:152.15ms
step:17/1530 train_loss:6.1825 train_time:1073ms step_avg:153.30ms
step:18/1530 train_loss:5.9777 train_time:1234ms step_avg:154.20ms
step:19/1530 train_loss:5.9787 train_time:1393ms step_avg:154.73ms
step:20/1530 train_loss:5.6834 train_time:1553ms step_avg:155.27ms
step:21/1530 train_loss:5.9522 train_time:1713ms step_avg:155.75ms
step:22/1530 train_loss:6.1648 train_time:1873ms step_avg:156.09ms
step:23/1530 train_loss:5.8383 train_time:2034ms step_avg:156.46ms
step:24/1530 train_loss:6.0122 train_time:2193ms step_avg:156.67ms
step:25/1530 train_loss:5.6690 train_time:2353ms step_avg:156.89ms
step:26/1530 train_loss:5.5787 train_time:2513ms step_avg:157.05ms
step:27/1530 train_loss:5.7481 train_time:2673ms step_avg:157.24ms
step:28/1530 train_loss:5.4134 train_time:2833ms step_avg:157.39ms
step:29/1530 train_loss:5.6576 train_time:2993ms step_avg:157.54ms
step:30/1530 train_loss:5.4556 train_time:3153ms step_avg:157.63ms
step:31/1530 train_loss:5.4171 train_time:3313ms step_avg:157.74ms
step:32/1530 train_loss:5.2816 train_time:3473ms step_avg:157.87ms
step:33/1530 train_loss:5.5756 train_time:3633ms step_avg:157.96ms
step:34/1530 train_loss:5.4900 train_time:3794ms step_avg:158.07ms
step:35/1530 train_loss:5.6134 train_time:3954ms step_avg:158.14ms
step:36/1530 train_loss:5.5489 train_time:4113ms step_avg:158.21ms
step:37/1530 train_loss:5.4465 train_time:4273ms step_avg:158.27ms
step:38/1530 train_loss:5.3060 train_time:4433ms step_avg:158.33ms
step:39/1530 train_loss:5.3263 train_time:4593ms step_avg:158.39ms
step:40/1530 train_loss:5.2508 train_time:4753ms step_avg:158.44ms
step:41/1530 train_loss:5.2227 train_time:4914ms step_avg:158.50ms
step:42/1530 train_loss:5.1580 train_time:5074ms step_avg:158.55ms
step:43/1530 train_loss:5.2548 train_time:5234ms step_avg:158.60ms
step:44/1530 train_loss:5.2369 train_time:5393ms step_avg:158.62ms
step:45/1530 train_loss:5.3814 train_time:5553ms step_avg:158.65ms
step:46/1530 train_loss:5.1693 train_time:5713ms step_avg:158.70ms
step:47/1530 train_loss:5.0594 train_time:5873ms step_avg:158.72ms
step:48/1530 train_loss:5.2329 train_time:6033ms step_avg:158.76ms
step:49/1530 train_loss:5.1355 train_time:6193ms step_avg:158.78ms
step:50/1530 train_loss:5.2375 train_time:6353ms step_avg:158.82ms
step:51/1530 train_loss:5.1395 train_time:6514ms step_avg:158.87ms
step:52/1530 train_loss:5.0339 train_time:6673ms step_avg:158.88ms
step:53/1530 train_loss:5.1743 train_time:6833ms step_avg:158.91ms
step:54/1530 train_loss:5.0279 train_time:6992ms step_avg:158.92ms
step:55/1530 train_loss:5.4164 train_time:7153ms step_avg:158.96ms
step:56/1530 train_loss:5.0259 train_time:7313ms step_avg:158.98ms
step:57/1530 train_loss:4.8817 train_time:7473ms step_avg:159.00ms
step:58/1530 train_loss:5.0504 train_time:7633ms step_avg:159.02ms
step:59/1530 train_loss:5.0139 train_time:7793ms step_avg:159.05ms
step:60/1530 train_loss:5.1341 train_time:7953ms step_avg:159.06ms
step:61/1530 train_loss:4.8665 train_time:8113ms step_avg:159.07ms
step:62/1530 train_loss:4.9991 train_time:8273ms step_avg:159.09ms
step:63/1530 train_loss:4.9855 train_time:8433ms step_avg:159.12ms
step:64/1530 train_loss:4.9541 train_time:8593ms step_avg:159.12ms
step:65/1530 train_loss:4.8307 train_time:8753ms step_avg:159.15ms
step:66/1530 train_loss:4.9227 train_time:8913ms step_avg:159.17ms
step:67/1530 train_loss:4.8255 train_time:9073ms step_avg:159.17ms
step:68/1530 train_loss:5.1113 train_time:9233ms step_avg:159.19ms
step:69/1530 train_loss:4.7412 train_time:9393ms step_avg:159.21ms
step:70/1530 train_loss:4.8557 train_time:9554ms step_avg:159.23ms
step:71/1530 train_loss:4.9956 train_time:9713ms step_avg:159.24ms
step:72/1530 train_loss:4.9088 train_time:9873ms step_avg:159.25ms
step:73/1530 train_loss:4.7870 train_time:10033ms step_avg:159.26ms
step:74/1530 train_loss:4.9157 train_time:10193ms step_avg:159.27ms
step:75/1530 train_loss:4.8808 train_time:10353ms step_avg:159.28ms
step:76/1530 train_loss:4.8039 train_time:10513ms step_avg:159.29ms
step:77/1530 train_loss:4.9307 train_time:10673ms step_avg:159.30ms
step:78/1530 train_loss:5.1208 train_time:10833ms step_avg:159.31ms
step:79/1530 train_loss:4.8562 train_time:10993ms step_avg:159.31ms
step:80/1530 train_loss:4.8781 train_time:11153ms step_avg:159.32ms
step:81/1530 train_loss:4.6727 train_time:11313ms step_avg:159.34ms
step:82/1530 train_loss:4.8449 train_time:11472ms step_avg:159.34ms
step:83/1530 train_loss:4.8153 train_time:11634ms step_avg:159.36ms
step:84/1530 train_loss:4.8069 train_time:11792ms step_avg:159.36ms
step:85/1530 train_loss:4.6451 train_time:11953ms step_avg:159.37ms
step:86/1530 train_loss:4.8504 train_time:12113ms step_avg:159.39ms
step:87/1530 train_loss:4.7697 train_time:12273ms step_avg:159.39ms
step:88/1530 train_loss:4.7739 train_time:12433ms step_avg:159.40ms
step:89/1530 train_loss:4.7237 train_time:12593ms step_avg:159.41ms
step:90/1530 train_loss:4.6703 train_time:12753ms step_avg:159.41ms
step:91/1530 train_loss:4.6665 train_time:12914ms step_avg:159.43ms
step:92/1530 train_loss:4.8327 train_time:13073ms step_avg:159.43ms
step:93/1530 train_loss:4.6463 train_time:13234ms step_avg:159.44ms
step:94/1530 train_loss:4.6534 train_time:13393ms step_avg:159.44ms
step:95/1530 train_loss:4.7191 train_time:13553ms step_avg:159.45ms
step:96/1530 train_loss:4.6037 train_time:13713ms step_avg:159.45ms
step:97/1530 train_loss:4.6643 train_time:13873ms step_avg:159.46ms
step:98/1530 train_loss:4.6058 train_time:14034ms step_avg:159.47ms
step:99/1530 train_loss:4.6968 train_time:14193ms step_avg:159.48ms
step:100/1530 train_loss:4.6959 train_time:14353ms step_avg:159.48ms
step:101/1530 train_loss:4.5908 train_time:14513ms step_avg:159.49ms
step:102/1530 train_loss:4.7451 train_time:14674ms step_avg:159.49ms
step:103/1530 train_loss:4.6093 train_time:14834ms step_avg:159.51ms
step:104/1530 train_loss:4.5542 train_time:14994ms step_avg:159.51ms
step:105/1530 train_loss:4.5726 train_time:15154ms step_avg:159.51ms
step:106/1530 train_loss:4.6628 train_time:15314ms step_avg:159.52ms
step:107/1530 train_loss:4.5440 train_time:15474ms step_avg:159.53ms
step:108/1530 train_loss:4.3774 train_time:15634ms step_avg:159.53ms
step:109/1530 train_loss:4.5066 train_time:15794ms step_avg:159.53ms
step:110/1530 train_loss:4.5191 train_time:15954ms step_avg:159.54ms
step:111/1530 train_loss:4.4613 train_time:16115ms step_avg:159.55ms
step:112/1530 train_loss:4.6133 train_time:16275ms step_avg:159.55ms
step:113/1530 train_loss:4.5172 train_time:16435ms step_avg:159.56ms
step:114/1530 train_loss:4.4018 train_time:16594ms step_avg:159.55ms
step:115/1530 train_loss:4.5391 train_time:16756ms step_avg:159.58ms
step:116/1530 train_loss:4.4908 train_time:16921ms step_avg:159.63ms
step:117/1530 train_loss:4.3941 train_time:17085ms step_avg:159.67ms
step:118/1530 train_loss:4.6177 train_time:17249ms step_avg:159.72ms
step:119/1530 train_loss:4.4822 train_time:17413ms step_avg:159.76ms
step:120/1530 train_loss:4.3621 train_time:17577ms step_avg:159.79ms
step:121/1530 train_loss:4.3284 train_time:17740ms step_avg:159.82ms
step:122/1530 train_loss:4.4732 train_time:17905ms step_avg:159.87ms
step:123/1530 train_loss:4.2951 train_time:18069ms step_avg:159.90ms
step:124/1530 train_loss:4.6101 train_time:18232ms step_avg:159.93ms
step:125/1530 train_loss:4.4822 train_time:18395ms step_avg:159.95ms
step:125/1530 val_loss:4.4357 train_time:18442ms step_avg:160.37ms
step:126/1530 train_loss:4.4461 train_time:18562ms step_avg:160.02ms
step:127/1530 train_loss:4.4481 train_time:18727ms step_avg:160.06ms
step:128/1530 train_loss:4.3915 train_time:18890ms step_avg:160.09ms
step:129/1530 train_loss:4.7043 train_time:19053ms step_avg:160.11ms
step:130/1530 train_loss:4.3884 train_time:19219ms step_avg:160.16ms
step:131/1530 train_loss:4.4135 train_time:19383ms step_avg:160.19ms
step:132/1530 train_loss:4.3814 train_time:19547ms step_avg:160.22ms
step:133/1530 train_loss:4.4623 train_time:19711ms step_avg:160.25ms
step:134/1530 train_loss:4.2705 train_time:19875ms step_avg:160.28ms
step:135/1530 train_loss:4.4582 train_time:20038ms step_avg:160.31ms
step:136/1530 train_loss:4.2311 train_time:20203ms step_avg:160.34ms
step:137/1530 train_loss:4.3823 train_time:20367ms step_avg:160.37ms
step:138/1530 train_loss:4.3052 train_time:20530ms step_avg:160.39ms
step:139/1530 train_loss:4.4034 train_time:20693ms step_avg:160.41ms
step:140/1530 train_loss:4.4915 train_time:20858ms step_avg:160.44ms
step:141/1530 train_loss:4.3257 train_time:21021ms step_avg:160.47ms
step:142/1530 train_loss:4.3195 train_time:21186ms step_avg:160.50ms
step:143/1530 train_loss:4.2692 train_time:21349ms step_avg:160.52ms
step:144/1530 train_loss:4.3648 train_time:21513ms step_avg:160.54ms
step:145/1530 train_loss:4.3169 train_time:21677ms step_avg:160.57ms
step:146/1530 train_loss:4.1824 train_time:21843ms step_avg:160.61ms
step:147/1530 train_loss:4.3364 train_time:22008ms step_avg:160.64ms
step:148/1530 train_loss:4.3716 train_time:22171ms step_avg:160.66ms
step:149/1530 train_loss:4.3133 train_time:22335ms step_avg:160.69ms
step:150/1530 train_loss:4.4540 train_time:22501ms step_avg:160.72ms
step:151/1530 train_loss:4.2829 train_time:22664ms step_avg:160.74ms
step:152/1530 train_loss:4.2832 train_time:22828ms step_avg:160.76ms
step:153/1530 train_loss:4.3736 train_time:22993ms step_avg:160.79ms
step:154/1530 train_loss:4.3797 train_time:23157ms step_avg:160.82ms
step:155/1530 train_loss:4.2817 train_time:23323ms step_avg:160.85ms
step:156/1530 train_loss:4.3516 train_time:23486ms step_avg:160.86ms
step:157/1530 train_loss:4.3996 train_time:23649ms step_avg:160.88ms
step:158/1530 train_loss:4.2454 train_time:23813ms step_avg:160.90ms
step:159/1530 train_loss:4.3142 train_time:23977ms step_avg:160.92ms
step:160/1530 train_loss:4.1481 train_time:24141ms step_avg:160.94ms
step:161/1530 train_loss:4.3605 train_time:24304ms step_avg:160.95ms
step:162/1530 train_loss:4.3684 train_time:24467ms step_avg:160.97ms
step:163/1530 train_loss:4.3448 train_time:24630ms step_avg:160.98ms
step:164/1530 train_loss:4.1983 train_time:24794ms step_avg:161.00ms
step:165/1530 train_loss:4.2938 train_time:24958ms step_avg:161.02ms
step:166/1530 train_loss:4.3437 train_time:25122ms step_avg:161.04ms
step:167/1530 train_loss:4.2175 train_time:25286ms step_avg:161.06ms
step:168/1530 train_loss:4.3041 train_time:25449ms step_avg:161.07ms
step:169/1530 train_loss:4.1739 train_time:25613ms step_avg:161.09ms
step:170/1530 train_loss:4.0398 train_time:25779ms step_avg:161.12ms
step:171/1530 train_loss:4.2136 train_time:25942ms step_avg:161.13ms
step:172/1530 train_loss:4.2121 train_time:26105ms step_avg:161.14ms
step:173/1530 train_loss:4.2723 train_time:26267ms step_avg:161.15ms
step:174/1530 train_loss:4.4297 train_time:26429ms step_avg:161.15ms
step:175/1530 train_loss:4.2439 train_time:26592ms step_avg:161.16ms
step:176/1530 train_loss:4.0983 train_time:26754ms step_avg:161.17ms
step:177/1530 train_loss:4.0756 train_time:26919ms step_avg:161.19ms
step:178/1530 train_loss:4.1911 train_time:27082ms step_avg:161.20ms
step:179/1530 train_loss:4.1323 train_time:27245ms step_avg:161.21ms
step:180/1530 train_loss:4.1324 train_time:27407ms step_avg:161.22ms
step:181/1530 train_loss:4.3007 train_time:27569ms step_avg:161.23ms
step:182/1530 train_loss:4.1555 train_time:27732ms step_avg:161.23ms
step:183/1530 train_loss:4.1483 train_time:27896ms step_avg:161.25ms
step:184/1530 train_loss:4.1305 train_time:28059ms step_avg:161.26ms
step:185/1530 train_loss:4.2124 train_time:28223ms step_avg:161.27ms
step:186/1530 train_loss:4.1706 train_time:28385ms step_avg:161.28ms
step:187/1530 train_loss:4.2399 train_time:28548ms step_avg:161.29ms
step:188/1530 train_loss:4.1759 train_time:28851ms step_avg:162.08ms
step:189/1530 train_loss:4.1170 train_time:29180ms step_avg:163.02ms
step:190/1530 train_loss:4.2149 train_time:29344ms step_avg:163.02ms
step:191/1530 train_loss:4.0817 train_time:29506ms step_avg:163.02ms
step:192/1530 train_loss:4.0368 train_time:29668ms step_avg:163.01ms
step:193/1530 train_loss:4.2503 train_time:29830ms step_avg:163.01ms
step:194/1530 train_loss:4.1733 train_time:29993ms step_avg:163.01ms
step:195/1530 train_loss:4.3529 train_time:30156ms step_avg:163.01ms
step:196/1530 train_loss:4.1869 train_time:30320ms step_avg:163.01ms
step:197/1530 train_loss:4.0456 train_time:30483ms step_avg:163.01ms
step:198/1530 train_loss:4.1813 train_time:30646ms step_avg:163.01ms
step:199/1530 train_loss:4.0342 train_time:30809ms step_avg:163.01ms
step:200/1530 train_loss:4.1146 train_time:30972ms step_avg:163.01ms
step:201/1530 train_loss:4.0330 train_time:31134ms step_avg:163.01ms
step:202/1530 train_loss:4.2757 train_time:31299ms step_avg:163.02ms
step:203/1530 train_loss:4.0638 train_time:31462ms step_avg:163.02ms
step:204/1530 train_loss:4.1903 train_time:31626ms step_avg:163.02ms
step:205/1530 train_loss:4.2513 train_time:31789ms step_avg:163.02ms
step:206/1530 train_loss:3.9446 train_time:31950ms step_avg:163.01ms
step:207/1530 train_loss:4.0839 train_time:32114ms step_avg:163.02ms
step:208/1530 train_loss:4.1009 train_time:32276ms step_avg:163.01ms
step:209/1530 train_loss:4.2430 train_time:32439ms step_avg:163.01ms
step:210/1530 train_loss:4.1899 train_time:32603ms step_avg:163.02ms
step:211/1530 train_loss:4.0690 train_time:32765ms step_avg:163.01ms
step:212/1530 train_loss:4.1349 train_time:32927ms step_avg:163.00ms
step:213/1530 train_loss:4.0561 train_time:33090ms step_avg:163.01ms
step:214/1530 train_loss:4.1174 train_time:33252ms step_avg:163.00ms
step:215/1530 train_loss:3.9589 train_time:33415ms step_avg:163.00ms
step:216/1530 train_loss:4.0097 train_time:33580ms step_avg:163.01ms
step:217/1530 train_loss:4.0115 train_time:33744ms step_avg:163.02ms
step:218/1530 train_loss:4.0901 train_time:33906ms step_avg:163.01ms
step:219/1530 train_loss:4.0819 train_time:34068ms step_avg:163.01ms
step:220/1530 train_loss:4.0892 train_time:34230ms step_avg:163.00ms
step:221/1530 train_loss:4.1052 train_time:34393ms step_avg:163.00ms
step:222/1530 train_loss:4.0016 train_time:34556ms step_avg:163.00ms
step:223/1530 train_loss:3.9906 train_time:34720ms step_avg:163.00ms
step:224/1530 train_loss:4.3012 train_time:34883ms step_avg:163.00ms
step:225/1530 train_loss:3.9274 train_time:35046ms step_avg:163.00ms
step:226/1530 train_loss:3.9980 train_time:35207ms step_avg:163.00ms
step:227/1530 train_loss:3.9856 train_time:35370ms step_avg:162.99ms
step:228/1530 train_loss:4.1430 train_time:35534ms step_avg:163.00ms
step:229/1530 train_loss:3.9268 train_time:35703ms step_avg:163.03ms
step:230/1530 train_loss:4.0414 train_time:35867ms step_avg:163.03ms
step:231/1530 train_loss:3.9026 train_time:36033ms step_avg:163.05ms
step:232/1530 train_loss:3.9736 train_time:36200ms step_avg:163.06ms
step:233/1530 train_loss:4.0946 train_time:36365ms step_avg:163.07ms
step:234/1530 train_loss:4.0330 train_time:36531ms step_avg:163.08ms
step:235/1530 train_loss:3.9035 train_time:36698ms step_avg:163.10ms
step:236/1530 train_loss:4.0828 train_time:36864ms step_avg:163.11ms
step:237/1530 train_loss:4.0794 train_time:37029ms step_avg:163.12ms
step:238/1530 train_loss:3.9471 train_time:37194ms step_avg:163.13ms
step:239/1530 train_loss:4.0778 train_time:37361ms step_avg:163.15ms
step:240/1530 train_loss:4.1053 train_time:37526ms step_avg:163.16ms
step:241/1530 train_loss:3.9618 train_time:37691ms step_avg:163.17ms
step:242/1530 train_loss:4.1398 train_time:37858ms step_avg:163.18ms
step:243/1530 train_loss:4.0138 train_time:38024ms step_avg:163.19ms
step:244/1530 train_loss:4.0919 train_time:38191ms step_avg:163.21ms
step:245/1530 train_loss:4.1469 train_time:38356ms step_avg:163.22ms
step:246/1530 train_loss:4.0544 train_time:38522ms step_avg:163.23ms
step:247/1530 train_loss:4.0083 train_time:38688ms step_avg:163.24ms
step:248/1530 train_loss:4.1088 train_time:38853ms step_avg:163.25ms
step:249/1530 train_loss:3.9180 train_time:39018ms step_avg:163.26ms
step:250/1530 train_loss:3.9794 train_time:39185ms step_avg:163.27ms
step:250/1530 val_loss:4.0048 train_time:39233ms step_avg:163.47ms
step:251/1530 train_loss:4.0773 train_time:39355ms step_avg:163.30ms
step:252/1530 train_loss:4.1665 train_time:39522ms step_avg:163.32ms
step:253/1530 train_loss:3.9326 train_time:39690ms step_avg:163.33ms
step:254/1530 train_loss:3.8853 train_time:39856ms step_avg:163.34ms
step:255/1530 train_loss:4.0790 train_time:40021ms step_avg:163.35ms
step:256/1530 train_loss:3.9892 train_time:40188ms step_avg:163.37ms
step:257/1530 train_loss:3.9902 train_time:40353ms step_avg:163.37ms
step:258/1530 train_loss:3.9855 train_time:40518ms step_avg:163.38ms
step:259/1530 train_loss:4.0317 train_time:40686ms step_avg:163.40ms
step:260/1530 train_loss:4.0596 train_time:40852ms step_avg:163.41ms
step:261/1530 train_loss:4.0294 train_time:41019ms step_avg:163.42ms
step:262/1530 train_loss:4.0018 train_time:41186ms step_avg:163.44ms
step:263/1530 train_loss:3.8935 train_time:41352ms step_avg:163.45ms
step:264/1530 train_loss:3.9906 train_time:41518ms step_avg:163.46ms
step:265/1530 train_loss:3.8787 train_time:41685ms step_avg:163.47ms
step:266/1530 train_loss:3.9203 train_time:41851ms step_avg:163.48ms
step:267/1530 train_loss:3.9278 train_time:42016ms step_avg:163.49ms
step:268/1530 train_loss:3.9641 train_time:42184ms step_avg:163.50ms
step:269/1530 train_loss:3.8592 train_time:42350ms step_avg:163.51ms
step:270/1530 train_loss:4.1103 train_time:42515ms step_avg:163.52ms
step:271/1530 train_loss:3.9786 train_time:42681ms step_avg:163.53ms
step:272/1530 train_loss:3.9279 train_time:42847ms step_avg:163.54ms
step:273/1530 train_loss:3.9432 train_time:43013ms step_avg:163.55ms
step:274/1530 train_loss:4.0368 train_time:43180ms step_avg:163.56ms
step:275/1530 train_loss:4.0618 train_time:43346ms step_avg:163.57ms
step:276/1530 train_loss:4.2329 train_time:43511ms step_avg:163.57ms
step:277/1530 train_loss:4.0466 train_time:43678ms step_avg:163.59ms
step:278/1530 train_loss:4.0919 train_time:43845ms step_avg:163.60ms
step:279/1530 train_loss:3.9999 train_time:44010ms step_avg:163.61ms
step:280/1530 train_loss:4.1662 train_time:44179ms step_avg:163.62ms
step:281/1530 train_loss:3.9766 train_time:44346ms step_avg:163.64ms
step:282/1530 train_loss:3.9512 train_time:44512ms step_avg:163.65ms
step:283/1530 train_loss:3.9062 train_time:44677ms step_avg:163.65ms
step:284/1530 train_loss:4.0448 train_time:44845ms step_avg:163.67ms
step:285/1530 train_loss:4.0579 train_time:45010ms step_avg:163.67ms
step:286/1530 train_loss:4.0897 train_time:45175ms step_avg:163.68ms
step:287/1530 train_loss:3.9056 train_time:45340ms step_avg:163.68ms
step:288/1530 train_loss:4.0079 train_time:45505ms step_avg:163.69ms
step:289/1530 train_loss:3.8800 train_time:45670ms step_avg:163.69ms
step:290/1530 train_loss:3.8575 train_time:45835ms step_avg:163.70ms
step:291/1530 train_loss:3.9076 train_time:46000ms step_avg:163.70ms
step:292/1530 train_loss:3.8643 train_time:46166ms step_avg:163.71ms
step:293/1530 train_loss:3.8972 train_time:46330ms step_avg:163.71ms
step:294/1530 train_loss:3.9297 train_time:46495ms step_avg:163.71ms
step:295/1530 train_loss:3.8464 train_time:46660ms step_avg:163.72ms
step:296/1530 train_loss:3.8647 train_time:46826ms step_avg:163.73ms
step:297/1530 train_loss:3.8713 train_time:46991ms step_avg:163.73ms
step:298/1530 train_loss:3.9774 train_time:47155ms step_avg:163.73ms
step:299/1530 train_loss:3.8262 train_time:47322ms step_avg:163.75ms
step:300/1530 train_loss:3.9650 train_time:47488ms step_avg:163.75ms
step:301/1530 train_loss:3.9599 train_time:47652ms step_avg:163.75ms
step:302/1530 train_loss:3.9303 train_time:47816ms step_avg:163.75ms
step:303/1530 train_loss:3.9777 train_time:47982ms step_avg:163.76ms
step:304/1530 train_loss:3.9669 train_time:48147ms step_avg:163.77ms
step:305/1530 train_loss:4.4549 train_time:48312ms step_avg:163.77ms
step:306/1530 train_loss:3.9415 train_time:48477ms step_avg:163.78ms
step:307/1530 train_loss:3.8356 train_time:48643ms step_avg:163.78ms
step:308/1530 train_loss:3.9746 train_time:48809ms step_avg:163.79ms
step:309/1530 train_loss:3.8626 train_time:48972ms step_avg:163.79ms
step:310/1530 train_loss:4.0758 train_time:49138ms step_avg:163.79ms
step:311/1530 train_loss:3.9306 train_time:49305ms step_avg:163.80ms
step:312/1530 train_loss:3.8664 train_time:49470ms step_avg:163.81ms
step:313/1530 train_loss:3.9354 train_time:49635ms step_avg:163.81ms
step:314/1530 train_loss:4.0609 train_time:49801ms step_avg:163.82ms
step:315/1530 train_loss:3.9454 train_time:49967ms step_avg:163.83ms
step:316/1530 train_loss:3.7971 train_time:50131ms step_avg:163.83ms
step:317/1530 train_loss:3.8822 train_time:50295ms step_avg:163.83ms
step:318/1530 train_loss:3.9294 train_time:50461ms step_avg:163.83ms
step:319/1530 train_loss:3.8863 train_time:50625ms step_avg:163.84ms
step:320/1530 train_loss:4.0136 train_time:50791ms step_avg:163.84ms
step:321/1530 train_loss:3.9598 train_time:50956ms step_avg:163.85ms
step:322/1530 train_loss:3.9306 train_time:51122ms step_avg:163.85ms
step:323/1530 train_loss:4.0060 train_time:51288ms step_avg:163.86ms
step:324/1530 train_loss:3.9518 train_time:51453ms step_avg:163.86ms
step:325/1530 train_loss:4.0078 train_time:51618ms step_avg:163.87ms
step:326/1530 train_loss:3.8903 train_time:51784ms step_avg:163.87ms
step:327/1530 train_loss:4.3917 train_time:51949ms step_avg:163.88ms
step:328/1530 train_loss:4.0704 train_time:52113ms step_avg:163.88ms
step:329/1530 train_loss:3.8006 train_time:52278ms step_avg:163.88ms
step:330/1530 train_loss:3.7540 train_time:52444ms step_avg:163.89ms
step:331/1530 train_loss:3.9757 train_time:52609ms step_avg:163.89ms
step:332/1530 train_loss:3.9123 train_time:52774ms step_avg:163.89ms
step:333/1530 train_loss:3.8749 train_time:52939ms step_avg:163.90ms
step:334/1530 train_loss:3.8423 train_time:53104ms step_avg:163.90ms
step:335/1530 train_loss:4.0126 train_time:53270ms step_avg:163.91ms
step:336/1530 train_loss:3.9528 train_time:53433ms step_avg:163.91ms
step:337/1530 train_loss:4.4263 train_time:53598ms step_avg:163.91ms
step:338/1530 train_loss:3.9294 train_time:53765ms step_avg:163.92ms
step:339/1530 train_loss:3.8546 train_time:53929ms step_avg:163.92ms
step:340/1530 train_loss:3.9366 train_time:54093ms step_avg:163.92ms
step:341/1530 train_loss:3.8557 train_time:54261ms step_avg:163.93ms
step:342/1530 train_loss:3.8060 train_time:54428ms step_avg:163.94ms
step:343/1530 train_loss:3.8354 train_time:54597ms step_avg:163.95ms
step:344/1530 train_loss:3.9930 train_time:54765ms step_avg:163.97ms
step:345/1530 train_loss:3.8249 train_time:54932ms step_avg:163.98ms
step:346/1530 train_loss:3.7617 train_time:55100ms step_avg:163.99ms
step:347/1530 train_loss:3.7903 train_time:55270ms step_avg:164.00ms
step:348/1530 train_loss:3.8634 train_time:55437ms step_avg:164.01ms
step:349/1530 train_loss:3.8333 train_time:55606ms step_avg:164.03ms
step:350/1530 train_loss:3.5679 train_time:55773ms step_avg:164.04ms
step:351/1530 train_loss:3.8329 train_time:55942ms step_avg:164.05ms
step:352/1530 train_loss:4.1803 train_time:56109ms step_avg:164.06ms
step:353/1530 train_loss:3.6595 train_time:56277ms step_avg:164.07ms
step:354/1530 train_loss:3.9283 train_time:56444ms step_avg:164.08ms
step:355/1530 train_loss:3.7846 train_time:56612ms step_avg:164.09ms
step:356/1530 train_loss:3.8858 train_time:56781ms step_avg:164.11ms
step:357/1530 train_loss:3.7549 train_time:56949ms step_avg:164.12ms
step:358/1530 train_loss:3.8644 train_time:57117ms step_avg:164.13ms
step:359/1530 train_loss:3.8049 train_time:57287ms step_avg:164.15ms
step:360/1530 train_loss:3.4206 train_time:57456ms step_avg:164.16ms
step:361/1530 train_loss:4.0106 train_time:57625ms step_avg:164.17ms
step:362/1530 train_loss:3.9113 train_time:57793ms step_avg:164.18ms
step:363/1530 train_loss:3.8375 train_time:57960ms step_avg:164.19ms
step:364/1530 train_loss:3.7442 train_time:58128ms step_avg:164.20ms
step:365/1530 train_loss:3.9119 train_time:58296ms step_avg:164.22ms
step:366/1530 train_loss:3.8526 train_time:58466ms step_avg:164.23ms
step:367/1530 train_loss:3.8538 train_time:58632ms step_avg:164.23ms
step:368/1530 train_loss:3.8506 train_time:58799ms step_avg:164.24ms
step:369/1530 train_loss:3.7499 train_time:58969ms step_avg:164.26ms
step:370/1530 train_loss:3.8802 train_time:59135ms step_avg:164.26ms
step:371/1530 train_loss:3.7301 train_time:59303ms step_avg:164.27ms
step:372/1530 train_loss:3.6863 train_time:59471ms step_avg:164.29ms
step:373/1530 train_loss:3.9118 train_time:59638ms step_avg:164.29ms
step:374/1530 train_loss:3.8292 train_time:59806ms step_avg:164.30ms
step:375/1530 train_loss:3.7962 train_time:59974ms step_avg:164.31ms
step:375/1530 val_loss:3.8255 train_time:60022ms step_avg:164.45ms
step:376/1530 train_loss:3.8722 train_time:60143ms step_avg:164.33ms
step:377/1530 train_loss:3.7870 train_time:60450ms step_avg:164.71ms
step:378/1530 train_loss:3.8521 train_time:60626ms step_avg:164.75ms
step:379/1530 train_loss:3.8662 train_time:60946ms step_avg:165.17ms
step:380/1530 train_loss:3.9528 train_time:61116ms step_avg:165.18ms
step:381/1530 train_loss:3.8314 train_time:61283ms step_avg:165.18ms
step:382/1530 train_loss:3.8072 train_time:61451ms step_avg:165.19ms
step:383/1530 train_loss:3.7942 train_time:61620ms step_avg:165.20ms
step:384/1530 train_loss:3.8673 train_time:61785ms step_avg:165.20ms
step:385/1530 train_loss:3.7926 train_time:61954ms step_avg:165.21ms
step:386/1530 train_loss:3.8937 train_time:62122ms step_avg:165.22ms
step:387/1530 train_loss:4.0580 train_time:62289ms step_avg:165.22ms
step:388/1530 train_loss:3.7896 train_time:62458ms step_avg:165.23ms
step:389/1530 train_loss:3.7961 train_time:62626ms step_avg:165.24ms
step:390/1530 train_loss:3.9006 train_time:62794ms step_avg:165.25ms
step:391/1530 train_loss:3.8148 train_time:62961ms step_avg:165.25ms
step:392/1530 train_loss:3.9263 train_time:63128ms step_avg:165.26ms
step:393/1530 train_loss:3.7672 train_time:63297ms step_avg:165.27ms
step:394/1530 train_loss:3.8927 train_time:63464ms step_avg:165.27ms
step:395/1530 train_loss:3.6340 train_time:63634ms step_avg:165.28ms
step:396/1530 train_loss:3.8373 train_time:63802ms step_avg:165.29ms
step:397/1530 train_loss:3.8611 train_time:63969ms step_avg:165.29ms
step:398/1530 train_loss:3.8725 train_time:64137ms step_avg:165.30ms
step:399/1530 train_loss:3.7709 train_time:64303ms step_avg:165.30ms
step:400/1530 train_loss:3.8239 train_time:64471ms step_avg:165.31ms
step:401/1530 train_loss:3.9188 train_time:64638ms step_avg:165.32ms
step:402/1530 train_loss:3.8413 train_time:64805ms step_avg:165.32ms
step:403/1530 train_loss:3.9659 train_time:64974ms step_avg:165.33ms
step:404/1530 train_loss:3.6744 train_time:65141ms step_avg:165.33ms
step:405/1530 train_loss:3.7853 train_time:65309ms step_avg:165.34ms
step:406/1530 train_loss:4.0941 train_time:65475ms step_avg:165.34ms
step:407/1530 train_loss:3.7772 train_time:65642ms step_avg:165.35ms
step:408/1530 train_loss:3.8207 train_time:65809ms step_avg:165.35ms
step:409/1530 train_loss:3.8592 train_time:65975ms step_avg:165.35ms
step:410/1530 train_loss:3.7587 train_time:66141ms step_avg:165.35ms
step:411/1530 train_loss:3.7640 train_time:66308ms step_avg:165.36ms
step:412/1530 train_loss:4.1868 train_time:66476ms step_avg:165.36ms
step:413/1530 train_loss:3.6322 train_time:66642ms step_avg:165.37ms
step:414/1530 train_loss:4.0166 train_time:66810ms step_avg:165.37ms
step:415/1530 train_loss:3.7587 train_time:66976ms step_avg:165.37ms
step:416/1530 train_loss:3.7614 train_time:67143ms step_avg:165.38ms
step:417/1530 train_loss:3.9505 train_time:67311ms step_avg:165.38ms
step:418/1530 train_loss:3.6853 train_time:67478ms step_avg:165.39ms
step:419/1530 train_loss:3.8014 train_time:67644ms step_avg:165.39ms
step:420/1530 train_loss:3.7058 train_time:67812ms step_avg:165.39ms
step:421/1530 train_loss:3.6494 train_time:67977ms step_avg:165.40ms
step:422/1530 train_loss:3.7848 train_time:68144ms step_avg:165.40ms
step:423/1530 train_loss:3.8810 train_time:68312ms step_avg:165.41ms
step:424/1530 train_loss:3.6154 train_time:68479ms step_avg:165.41ms
step:425/1530 train_loss:3.7929 train_time:68645ms step_avg:165.41ms
step:426/1530 train_loss:3.6476 train_time:68814ms step_avg:165.42ms
step:427/1530 train_loss:3.8916 train_time:68980ms step_avg:165.42ms
step:428/1530 train_loss:3.8085 train_time:69146ms step_avg:165.42ms
step:429/1530 train_loss:3.7532 train_time:69316ms step_avg:165.43ms
step:430/1530 train_loss:3.7058 train_time:69483ms step_avg:165.44ms
step:431/1530 train_loss:3.6270 train_time:69651ms step_avg:165.44ms
step:432/1530 train_loss:3.7688 train_time:69819ms step_avg:165.45ms
step:433/1530 train_loss:3.8152 train_time:69985ms step_avg:165.45ms
step:434/1530 train_loss:3.7674 train_time:70153ms step_avg:165.46ms
step:435/1530 train_loss:3.8061 train_time:70320ms step_avg:165.46ms
step:436/1530 train_loss:3.8304 train_time:70487ms step_avg:165.46ms
step:437/1530 train_loss:3.7252 train_time:70655ms step_avg:165.47ms
step:438/1530 train_loss:3.6988 train_time:70822ms step_avg:165.47ms
step:439/1530 train_loss:3.7155 train_time:70990ms step_avg:165.48ms
step:440/1530 train_loss:3.8868 train_time:71156ms step_avg:165.48ms
step:441/1530 train_loss:3.7548 train_time:71323ms step_avg:165.48ms
step:442/1530 train_loss:3.7346 train_time:71492ms step_avg:165.49ms
step:443/1530 train_loss:3.6212 train_time:71659ms step_avg:165.49ms
step:444/1530 train_loss:3.9170 train_time:71826ms step_avg:165.50ms
step:445/1530 train_loss:3.8454 train_time:71993ms step_avg:165.50ms
step:446/1530 train_loss:3.8411 train_time:72159ms step_avg:165.50ms
step:447/1530 train_loss:3.7538 train_time:72327ms step_avg:165.51ms
step:448/1530 train_loss:3.8497 train_time:72495ms step_avg:165.51ms
step:449/1530 train_loss:3.6878 train_time:72663ms step_avg:165.52ms
step:450/1530 train_loss:3.7186 train_time:72831ms step_avg:165.53ms
step:451/1530 train_loss:3.5842 train_time:72999ms step_avg:165.53ms
step:452/1530 train_loss:3.7169 train_time:73165ms step_avg:165.53ms
step:453/1530 train_loss:3.6698 train_time:73334ms step_avg:165.54ms
step:454/1530 train_loss:3.6352 train_time:73500ms step_avg:165.54ms
step:455/1530 train_loss:3.8366 train_time:73669ms step_avg:165.55ms
step:456/1530 train_loss:3.7224 train_time:73838ms step_avg:165.56ms
step:457/1530 train_loss:3.7749 train_time:74009ms step_avg:165.57ms
step:458/1530 train_loss:3.8225 train_time:74178ms step_avg:165.58ms
step:459/1530 train_loss:3.6254 train_time:74348ms step_avg:165.59ms
step:460/1530 train_loss:3.7890 train_time:74517ms step_avg:165.59ms
step:461/1530 train_loss:3.6872 train_time:74687ms step_avg:165.60ms
step:462/1530 train_loss:3.7338 train_time:74856ms step_avg:165.61ms
step:463/1530 train_loss:3.7753 train_time:75025ms step_avg:165.62ms
step:464/1530 train_loss:3.7090 train_time:75196ms step_avg:165.63ms
step:465/1530 train_loss:3.7127 train_time:75366ms step_avg:165.64ms
step:466/1530 train_loss:3.8031 train_time:75535ms step_avg:165.65ms
step:467/1530 train_loss:3.8248 train_time:75705ms step_avg:165.66ms
step:468/1530 train_loss:3.7898 train_time:75873ms step_avg:165.66ms
step:469/1530 train_loss:3.6855 train_time:76044ms step_avg:165.67ms
step:470/1530 train_loss:3.7726 train_time:76215ms step_avg:165.68ms
step:471/1530 train_loss:3.8088 train_time:76384ms step_avg:165.69ms
step:472/1530 train_loss:3.7859 train_time:76555ms step_avg:165.70ms
step:473/1530 train_loss:3.7150 train_time:76724ms step_avg:165.71ms
step:474/1530 train_loss:3.5933 train_time:76894ms step_avg:165.72ms
step:475/1530 train_loss:4.0278 train_time:77062ms step_avg:165.72ms
step:476/1530 train_loss:3.7568 train_time:77233ms step_avg:165.74ms
step:477/1530 train_loss:3.5899 train_time:77402ms step_avg:165.74ms
step:478/1530 train_loss:3.8224 train_time:77571ms step_avg:165.75ms
step:479/1530 train_loss:3.7659 train_time:77740ms step_avg:165.76ms
step:480/1530 train_loss:3.9137 train_time:77911ms step_avg:165.77ms
step:481/1530 train_loss:3.7260 train_time:78080ms step_avg:165.78ms
step:482/1530 train_loss:3.5252 train_time:78250ms step_avg:165.78ms
step:483/1530 train_loss:3.7998 train_time:78420ms step_avg:165.79ms
step:484/1530 train_loss:3.6558 train_time:78590ms step_avg:165.80ms
step:485/1530 train_loss:3.6561 train_time:78761ms step_avg:165.81ms
step:486/1530 train_loss:3.5749 train_time:78931ms step_avg:165.82ms
step:487/1530 train_loss:3.6895 train_time:79100ms step_avg:165.83ms
step:488/1530 train_loss:3.8771 train_time:79269ms step_avg:165.84ms
step:489/1530 train_loss:3.7060 train_time:79440ms step_avg:165.85ms
step:490/1530 train_loss:3.5980 train_time:79610ms step_avg:165.85ms
step:491/1530 train_loss:3.6148 train_time:79778ms step_avg:165.86ms
step:492/1530 train_loss:3.7351 train_time:79947ms step_avg:165.87ms
step:493/1530 train_loss:3.5765 train_time:80119ms step_avg:165.88ms
step:494/1530 train_loss:3.6933 train_time:80289ms step_avg:165.89ms
step:495/1530 train_loss:3.6597 train_time:80459ms step_avg:165.90ms
step:496/1530 train_loss:3.5119 train_time:80628ms step_avg:165.90ms
step:497/1530 train_loss:3.7374 train_time:80797ms step_avg:165.91ms
step:498/1530 train_loss:3.7839 train_time:80966ms step_avg:165.91ms
step:499/1530 train_loss:3.8168 train_time:81137ms step_avg:165.92ms
step:500/1530 train_loss:3.7319 train_time:81306ms step_avg:165.93ms
step:500/1530 val_loss:3.7045 train_time:81354ms step_avg:166.03ms
step:501/1530 train_loss:3.8048 train_time:81475ms step_avg:165.94ms
step:502/1530 train_loss:3.7437 train_time:81646ms step_avg:165.95ms
step:503/1530 train_loss:3.7734 train_time:81815ms step_avg:165.95ms
step:504/1530 train_loss:3.7223 train_time:81984ms step_avg:165.96ms
step:505/1530 train_loss:3.8063 train_time:82155ms step_avg:165.97ms
step:506/1530 train_loss:3.6442 train_time:82326ms step_avg:165.98ms
step:507/1530 train_loss:3.7619 train_time:82495ms step_avg:165.99ms
step:508/1530 train_loss:3.8187 train_time:82667ms step_avg:166.00ms
step:509/1530 train_loss:3.7714 train_time:82836ms step_avg:166.00ms
step:510/1530 train_loss:3.5803 train_time:83006ms step_avg:166.01ms
step:511/1530 train_loss:3.7725 train_time:83176ms step_avg:166.02ms
step:512/1530 train_loss:3.7175 train_time:83349ms step_avg:166.03ms
step:513/1530 train_loss:3.6572 train_time:83516ms step_avg:166.04ms
step:514/1530 train_loss:3.8446 train_time:83687ms step_avg:166.04ms
step:515/1530 train_loss:3.7289 train_time:83855ms step_avg:166.05ms
step:516/1530 train_loss:4.0726 train_time:84026ms step_avg:166.06ms
step:517/1530 train_loss:3.6908 train_time:84194ms step_avg:166.06ms
step:518/1530 train_loss:3.7700 train_time:84364ms step_avg:166.07ms
step:519/1530 train_loss:3.6596 train_time:84533ms step_avg:166.08ms
step:520/1530 train_loss:3.6821 train_time:84702ms step_avg:166.08ms
step:521/1530 train_loss:3.6619 train_time:84872ms step_avg:166.09ms
step:522/1530 train_loss:3.6609 train_time:85041ms step_avg:166.10ms
step:523/1530 train_loss:4.2833 train_time:85211ms step_avg:166.10ms
step:524/1530 train_loss:3.7384 train_time:85378ms step_avg:166.11ms
step:525/1530 train_loss:3.6814 train_time:85549ms step_avg:166.11ms
step:526/1530 train_loss:3.6971 train_time:85717ms step_avg:166.12ms
step:527/1530 train_loss:3.6509 train_time:85886ms step_avg:166.12ms
step:528/1530 train_loss:3.6337 train_time:86055ms step_avg:166.13ms
step:529/1530 train_loss:3.8505 train_time:86223ms step_avg:166.13ms
step:530/1530 train_loss:3.6466 train_time:86393ms step_avg:166.14ms
step:531/1530 train_loss:3.9191 train_time:86563ms step_avg:166.15ms
step:532/1530 train_loss:3.7307 train_time:86732ms step_avg:166.15ms
step:533/1530 train_loss:3.6596 train_time:86901ms step_avg:166.16ms
step:534/1530 train_loss:3.6625 train_time:87071ms step_avg:166.17ms
step:535/1530 train_loss:3.6038 train_time:87240ms step_avg:166.17ms
step:536/1530 train_loss:3.7478 train_time:87409ms step_avg:166.18ms
step:537/1530 train_loss:3.7271 train_time:87579ms step_avg:166.18ms
step:538/1530 train_loss:3.6273 train_time:87750ms step_avg:166.19ms
step:539/1530 train_loss:4.1165 train_time:87920ms step_avg:166.20ms
step:540/1530 train_loss:3.6764 train_time:88090ms step_avg:166.21ms
step:541/1530 train_loss:3.7876 train_time:88258ms step_avg:166.21ms
step:542/1530 train_loss:3.5852 train_time:88427ms step_avg:166.22ms
step:543/1530 train_loss:3.5862 train_time:88595ms step_avg:166.22ms
step:544/1530 train_loss:3.6407 train_time:88765ms step_avg:166.23ms
step:545/1530 train_loss:3.5929 train_time:88934ms step_avg:166.23ms
step:546/1530 train_loss:3.6210 train_time:89103ms step_avg:166.24ms
step:547/1530 train_loss:3.6383 train_time:89271ms step_avg:166.24ms
step:548/1530 train_loss:3.6062 train_time:89440ms step_avg:166.25ms
step:549/1530 train_loss:3.7267 train_time:89609ms step_avg:166.25ms
step:550/1530 train_loss:3.6172 train_time:89777ms step_avg:166.25ms
step:551/1530 train_loss:3.6295 train_time:89945ms step_avg:166.26ms
step:552/1530 train_loss:3.9379 train_time:90115ms step_avg:166.26ms
step:553/1530 train_loss:3.7606 train_time:90283ms step_avg:166.27ms
step:554/1530 train_loss:3.7126 train_time:90453ms step_avg:166.27ms
step:555/1530 train_loss:3.6285 train_time:90622ms step_avg:166.28ms
step:556/1530 train_loss:3.6946 train_time:90790ms step_avg:166.28ms
step:557/1530 train_loss:3.3153 train_time:90960ms step_avg:166.29ms
step:558/1530 train_loss:3.6110 train_time:91130ms step_avg:166.29ms
step:559/1530 train_loss:3.6474 train_time:91298ms step_avg:166.30ms
step:560/1530 train_loss:3.6807 train_time:91469ms step_avg:166.31ms
step:561/1530 train_loss:3.6140 train_time:91638ms step_avg:166.31ms
step:562/1530 train_loss:3.5537 train_time:91806ms step_avg:166.32ms
step:563/1530 train_loss:3.7522 train_time:91975ms step_avg:166.32ms
step:564/1530 train_loss:3.5793 train_time:92146ms step_avg:166.33ms
step:565/1530 train_loss:3.6805 train_time:92315ms step_avg:166.33ms
step:566/1530 train_loss:3.6231 train_time:92621ms step_avg:166.59ms
step:567/1530 train_loss:3.5990 train_time:92799ms step_avg:166.60ms
step:568/1530 train_loss:3.6841 train_time:92971ms step_avg:166.61ms
step:569/1530 train_loss:3.6464 train_time:93299ms step_avg:166.90ms
step:570/1530 train_loss:3.6863 train_time:93471ms step_avg:166.91ms
step:571/1530 train_loss:3.7612 train_time:93639ms step_avg:166.91ms
step:572/1530 train_loss:3.7254 train_time:93813ms step_avg:166.93ms
step:573/1530 train_loss:3.7407 train_time:93987ms step_avg:166.94ms
step:574/1530 train_loss:3.7699 train_time:94161ms step_avg:166.95ms
step:575/1530 train_loss:3.7326 train_time:94332ms step_avg:166.96ms
step:576/1530 train_loss:3.7553 train_time:94501ms step_avg:166.96ms
step:577/1530 train_loss:3.6639 train_time:94673ms step_avg:166.97ms
step:578/1530 train_loss:3.6734 train_time:94845ms step_avg:166.98ms
step:579/1530 train_loss:3.6736 train_time:95015ms step_avg:166.99ms
step:580/1530 train_loss:3.5904 train_time:95185ms step_avg:166.99ms
step:581/1530 train_loss:3.6366 train_time:95355ms step_avg:167.00ms
step:582/1530 train_loss:3.8466 train_time:95526ms step_avg:167.00ms
step:583/1530 train_loss:3.6295 train_time:95698ms step_avg:167.01ms
step:584/1530 train_loss:3.5892 train_time:95871ms step_avg:167.02ms
step:585/1530 train_loss:3.7904 train_time:96040ms step_avg:167.03ms
step:586/1530 train_loss:3.5177 train_time:96212ms step_avg:167.03ms
step:587/1530 train_loss:3.6690 train_time:96382ms step_avg:167.04ms
step:588/1530 train_loss:3.6445 train_time:96553ms step_avg:167.05ms
step:589/1530 train_loss:3.9918 train_time:96726ms step_avg:167.06ms
step:590/1530 train_loss:3.7747 train_time:96897ms step_avg:167.06ms
step:591/1530 train_loss:3.5068 train_time:97069ms step_avg:167.07ms
step:592/1530 train_loss:3.5319 train_time:97242ms step_avg:167.08ms
step:593/1530 train_loss:3.4967 train_time:97413ms step_avg:167.09ms
step:594/1530 train_loss:3.5554 train_time:97584ms step_avg:167.10ms
step:595/1530 train_loss:3.9045 train_time:97758ms step_avg:167.11ms
step:596/1530 train_loss:3.6511 train_time:97931ms step_avg:167.12ms
step:597/1530 train_loss:3.5883 train_time:98102ms step_avg:167.12ms
step:598/1530 train_loss:3.6601 train_time:98272ms step_avg:167.13ms
step:599/1530 train_loss:3.4750 train_time:98443ms step_avg:167.14ms
step:600/1530 train_loss:3.5945 train_time:98613ms step_avg:167.14ms
step:601/1530 train_loss:3.6427 train_time:98786ms step_avg:167.15ms
step:602/1530 train_loss:3.6711 train_time:98959ms step_avg:167.16ms
step:603/1530 train_loss:3.7848 train_time:99131ms step_avg:167.17ms
step:604/1530 train_loss:3.6057 train_time:99304ms step_avg:167.18ms
step:605/1530 train_loss:3.6128 train_time:99475ms step_avg:167.19ms
step:606/1530 train_loss:3.5747 train_time:99648ms step_avg:167.20ms
step:607/1530 train_loss:3.8379 train_time:99819ms step_avg:167.20ms
step:608/1530 train_loss:3.6326 train_time:99990ms step_avg:167.21ms
step:609/1530 train_loss:3.6182 train_time:100161ms step_avg:167.21ms
step:610/1530 train_loss:3.7009 train_time:100331ms step_avg:167.22ms
step:611/1530 train_loss:3.6022 train_time:100501ms step_avg:167.22ms
step:612/1530 train_loss:3.5705 train_time:100673ms step_avg:167.23ms
step:613/1530 train_loss:3.7624 train_time:100844ms step_avg:167.24ms
step:614/1530 train_loss:3.6997 train_time:101014ms step_avg:167.24ms
step:615/1530 train_loss:3.7309 train_time:101184ms step_avg:167.25ms
step:616/1530 train_loss:3.6289 train_time:101355ms step_avg:167.25ms
step:617/1530 train_loss:3.5441 train_time:101528ms step_avg:167.26ms
step:618/1530 train_loss:3.6900 train_time:101697ms step_avg:167.27ms
step:619/1530 train_loss:3.5514 train_time:101872ms step_avg:167.28ms
step:620/1530 train_loss:3.5901 train_time:102042ms step_avg:167.28ms
step:621/1530 train_loss:3.9244 train_time:102215ms step_avg:167.29ms
step:622/1530 train_loss:3.5760 train_time:102388ms step_avg:167.30ms
step:623/1530 train_loss:3.6005 train_time:102560ms step_avg:167.31ms
step:624/1530 train_loss:3.6889 train_time:102730ms step_avg:167.31ms
step:625/1530 train_loss:3.7010 train_time:102900ms step_avg:167.32ms
step:625/1530 val_loss:3.6209 train_time:102949ms step_avg:167.40ms
step:626/1530 train_loss:3.7391 train_time:103072ms step_avg:167.32ms
step:627/1530 train_loss:3.7152 train_time:103243ms step_avg:167.33ms
step:628/1530 train_loss:3.7644 train_time:103414ms step_avg:167.34ms
step:629/1530 train_loss:3.5885 train_time:103584ms step_avg:167.34ms
step:630/1530 train_loss:3.7259 train_time:103756ms step_avg:167.35ms
step:631/1530 train_loss:3.7424 train_time:103926ms step_avg:167.35ms
step:632/1530 train_loss:3.6431 train_time:104098ms step_avg:167.36ms
step:633/1530 train_loss:3.6022 train_time:104268ms step_avg:167.36ms
step:634/1530 train_loss:3.6981 train_time:104438ms step_avg:167.37ms
step:635/1530 train_loss:3.9505 train_time:104608ms step_avg:167.37ms
step:636/1530 train_loss:3.5490 train_time:104779ms step_avg:167.38ms
step:637/1530 train_loss:3.3541 train_time:104949ms step_avg:167.38ms
step:638/1530 train_loss:3.5966 train_time:105119ms step_avg:167.39ms
step:639/1530 train_loss:3.6304 train_time:105289ms step_avg:167.39ms
step:640/1530 train_loss:3.5664 train_time:105460ms step_avg:167.40ms
step:641/1530 train_loss:3.5842 train_time:105629ms step_avg:167.40ms
step:642/1530 train_loss:3.6305 train_time:105800ms step_avg:167.40ms
step:643/1530 train_loss:3.5979 train_time:105970ms step_avg:167.41ms
step:644/1530 train_loss:3.5523 train_time:106140ms step_avg:167.41ms
step:645/1530 train_loss:3.7786 train_time:106310ms step_avg:167.42ms
step:646/1530 train_loss:3.6718 train_time:106482ms step_avg:167.42ms
step:647/1530 train_loss:3.6677 train_time:106652ms step_avg:167.43ms
step:648/1530 train_loss:3.7108 train_time:106825ms step_avg:167.44ms
step:649/1530 train_loss:3.7670 train_time:106995ms step_avg:167.44ms
step:650/1530 train_loss:3.6192 train_time:107166ms step_avg:167.45ms
step:651/1530 train_loss:3.7635 train_time:107337ms step_avg:167.45ms
step:652/1530 train_loss:3.5846 train_time:107507ms step_avg:167.46ms
step:653/1530 train_loss:3.6574 train_time:107678ms step_avg:167.46ms
step:654/1530 train_loss:3.4247 train_time:107850ms step_avg:167.47ms
step:655/1530 train_loss:3.5797 train_time:108019ms step_avg:167.47ms
step:656/1530 train_loss:3.5732 train_time:108188ms step_avg:167.47ms
step:657/1530 train_loss:3.4966 train_time:108361ms step_avg:167.48ms
step:658/1530 train_loss:3.6902 train_time:108531ms step_avg:167.49ms
step:659/1530 train_loss:3.5865 train_time:108702ms step_avg:167.49ms
step:660/1530 train_loss:3.6797 train_time:108872ms step_avg:167.50ms
step:661/1530 train_loss:3.7525 train_time:109043ms step_avg:167.50ms
step:662/1530 train_loss:3.6671 train_time:109213ms step_avg:167.50ms
step:663/1530 train_loss:3.5544 train_time:109382ms step_avg:167.51ms
step:664/1530 train_loss:3.6073 train_time:109553ms step_avg:167.51ms
step:665/1530 train_loss:3.4932 train_time:109723ms step_avg:167.52ms
step:666/1530 train_loss:3.7787 train_time:109894ms step_avg:167.52ms
step:667/1530 train_loss:3.5996 train_time:110063ms step_avg:167.52ms
step:668/1530 train_loss:3.6471 train_time:110233ms step_avg:167.53ms
step:669/1530 train_loss:3.4805 train_time:110404ms step_avg:167.53ms
step:670/1530 train_loss:3.5980 train_time:110575ms step_avg:167.54ms
step:671/1530 train_loss:3.5656 train_time:110745ms step_avg:167.54ms
step:672/1530 train_loss:3.5658 train_time:110917ms step_avg:167.55ms
step:673/1530 train_loss:3.8496 train_time:111088ms step_avg:167.55ms
step:674/1530 train_loss:3.6281 train_time:111258ms step_avg:167.56ms
step:675/1530 train_loss:3.7137 train_time:111429ms step_avg:167.56ms
step:676/1530 train_loss:3.4916 train_time:111601ms step_avg:167.57ms
step:677/1530 train_loss:3.6027 train_time:111772ms step_avg:167.57ms
step:678/1530 train_loss:3.5562 train_time:111942ms step_avg:167.58ms
step:679/1530 train_loss:3.6713 train_time:112113ms step_avg:167.58ms
step:680/1530 train_loss:3.5824 train_time:112284ms step_avg:167.59ms
step:681/1530 train_loss:3.6143 train_time:112456ms step_avg:167.60ms
step:682/1530 train_loss:3.6639 train_time:112631ms step_avg:167.61ms
step:683/1530 train_loss:3.7354 train_time:112803ms step_avg:167.61ms
step:684/1530 train_loss:3.6456 train_time:112977ms step_avg:167.62ms
step:685/1530 train_loss:3.6821 train_time:113151ms step_avg:167.63ms
step:686/1530 train_loss:3.6341 train_time:113322ms step_avg:167.64ms
step:687/1530 train_loss:3.6622 train_time:113495ms step_avg:167.64ms
step:688/1530 train_loss:3.2236 train_time:113670ms step_avg:167.65ms
step:689/1530 train_loss:3.4104 train_time:113843ms step_avg:167.66ms
step:690/1530 train_loss:3.5363 train_time:114018ms step_avg:167.67ms
step:691/1530 train_loss:3.4092 train_time:114190ms step_avg:167.68ms
step:692/1530 train_loss:3.6270 train_time:114362ms step_avg:167.69ms
step:693/1530 train_loss:3.6449 train_time:114535ms step_avg:167.69ms
step:694/1530 train_loss:3.5521 train_time:114706ms step_avg:167.70ms
step:695/1530 train_loss:3.5342 train_time:114878ms step_avg:167.70ms
step:696/1530 train_loss:3.8529 train_time:115048ms step_avg:167.71ms
step:697/1530 train_loss:3.5858 train_time:115223ms step_avg:167.72ms
step:698/1530 train_loss:3.6460 train_time:115395ms step_avg:167.73ms
step:699/1530 train_loss:3.7733 train_time:115569ms step_avg:167.73ms
step:700/1530 train_loss:3.5655 train_time:115741ms step_avg:167.74ms
step:701/1530 train_loss:3.5399 train_time:115913ms step_avg:167.75ms
step:702/1530 train_loss:3.5158 train_time:116087ms step_avg:167.76ms
step:703/1530 train_loss:3.4973 train_time:116259ms step_avg:167.76ms
step:704/1530 train_loss:3.5745 train_time:116433ms step_avg:167.77ms
step:705/1530 train_loss:3.5639 train_time:116609ms step_avg:167.78ms
step:706/1530 train_loss:3.5809 train_time:116783ms step_avg:167.79ms
step:707/1530 train_loss:3.6496 train_time:116958ms step_avg:167.80ms
step:708/1530 train_loss:3.6066 train_time:117130ms step_avg:167.81ms
step:709/1530 train_loss:3.5777 train_time:117304ms step_avg:167.82ms
step:710/1530 train_loss:3.5408 train_time:117476ms step_avg:167.82ms
step:711/1530 train_loss:3.5909 train_time:117649ms step_avg:167.83ms
step:712/1530 train_loss:3.6458 train_time:117824ms step_avg:167.84ms
step:713/1530 train_loss:3.6552 train_time:118000ms step_avg:167.85ms
step:714/1530 train_loss:3.5613 train_time:118173ms step_avg:167.86ms
step:715/1530 train_loss:3.5746 train_time:118345ms step_avg:167.87ms
step:716/1530 train_loss:3.5887 train_time:118517ms step_avg:167.87ms
step:717/1530 train_loss:3.7075 train_time:118691ms step_avg:167.88ms
step:718/1530 train_loss:3.5963 train_time:118862ms step_avg:167.88ms
step:719/1530 train_loss:3.6781 train_time:119035ms step_avg:167.89ms
step:720/1530 train_loss:3.8460 train_time:119209ms step_avg:167.90ms
step:721/1530 train_loss:3.4681 train_time:119382ms step_avg:167.91ms
step:722/1530 train_loss:3.7330 train_time:119553ms step_avg:167.91ms
step:723/1530 train_loss:3.7667 train_time:119724ms step_avg:167.92ms
step:724/1530 train_loss:3.5685 train_time:119897ms step_avg:167.92ms
step:725/1530 train_loss:3.6578 train_time:120070ms step_avg:167.93ms
step:726/1530 train_loss:3.5331 train_time:120242ms step_avg:167.94ms
step:727/1530 train_loss:3.5805 train_time:120417ms step_avg:167.95ms
step:728/1530 train_loss:3.7324 train_time:120590ms step_avg:167.95ms
step:729/1530 train_loss:3.6725 train_time:120762ms step_avg:167.96ms
step:730/1530 train_loss:3.6632 train_time:120935ms step_avg:167.96ms
step:731/1530 train_loss:3.5558 train_time:121106ms step_avg:167.97ms
step:732/1530 train_loss:3.6020 train_time:121278ms step_avg:167.97ms
step:733/1530 train_loss:3.8311 train_time:121452ms step_avg:167.98ms
step:734/1530 train_loss:3.5627 train_time:121626ms step_avg:167.99ms
step:735/1530 train_loss:3.6211 train_time:121798ms step_avg:168.00ms
step:736/1530 train_loss:3.7361 train_time:121971ms step_avg:168.00ms
step:737/1530 train_loss:3.6749 train_time:122141ms step_avg:168.01ms
step:738/1530 train_loss:3.5997 train_time:122313ms step_avg:168.01ms
step:739/1530 train_loss:3.5148 train_time:122485ms step_avg:168.02ms
step:740/1530 train_loss:4.1124 train_time:122662ms step_avg:168.03ms
step:741/1530 train_loss:3.4870 train_time:122835ms step_avg:168.04ms
step:742/1530 train_loss:3.5517 train_time:123008ms step_avg:168.04ms
step:743/1530 train_loss:3.5817 train_time:123181ms step_avg:168.05ms
step:744/1530 train_loss:3.6470 train_time:123354ms step_avg:168.06ms
step:745/1530 train_loss:3.5848 train_time:123528ms step_avg:168.07ms
step:746/1530 train_loss:3.5974 train_time:123700ms step_avg:168.07ms
step:747/1530 train_loss:3.6443 train_time:123873ms step_avg:168.08ms
step:748/1530 train_loss:3.5647 train_time:124048ms step_avg:168.09ms
step:749/1530 train_loss:3.5606 train_time:124221ms step_avg:168.09ms
step:750/1530 train_loss:3.5976 train_time:124394ms step_avg:168.10ms
step:750/1530 val_loss:3.5680 train_time:124443ms step_avg:168.17ms
step:751/1530 train_loss:3.5755 train_time:124566ms step_avg:168.11ms
step:752/1530 train_loss:3.6146 train_time:124740ms step_avg:168.11ms
step:753/1530 train_loss:3.6152 train_time:124914ms step_avg:168.12ms
step:754/1530 train_loss:3.6003 train_time:125085ms step_avg:168.13ms
step:755/1530 train_loss:3.6812 train_time:125396ms step_avg:168.32ms
step:756/1530 train_loss:3.4592 train_time:125582ms step_avg:168.34ms
step:757/1530 train_loss:3.7240 train_time:125756ms step_avg:168.35ms
step:758/1530 train_loss:3.6506 train_time:125926ms step_avg:168.35ms
step:759/1530 train_loss:3.5902 train_time:126257ms step_avg:168.57ms
step:760/1530 train_loss:3.7054 train_time:126426ms step_avg:168.57ms
step:761/1530 train_loss:3.4048 train_time:126599ms step_avg:168.57ms
step:762/1530 train_loss:3.5508 train_time:126771ms step_avg:168.58ms
step:763/1530 train_loss:3.6618 train_time:126943ms step_avg:168.58ms
step:764/1530 train_loss:3.3178 train_time:127117ms step_avg:168.59ms
step:765/1530 train_loss:3.7325 train_time:127287ms step_avg:168.59ms
step:766/1530 train_loss:3.5698 train_time:127460ms step_avg:168.60ms
step:767/1530 train_loss:3.5638 train_time:127632ms step_avg:168.60ms
step:768/1530 train_loss:3.5726 train_time:127806ms step_avg:168.61ms
step:769/1530 train_loss:3.5877 train_time:127980ms step_avg:168.62ms
step:770/1530 train_loss:3.6387 train_time:128151ms step_avg:168.62ms
step:771/1530 train_loss:3.8911 train_time:128323ms step_avg:168.62ms
step:772/1530 train_loss:3.4560 train_time:128495ms step_avg:168.63ms
step:773/1530 train_loss:3.6279 train_time:128666ms step_avg:168.63ms
step:774/1530 train_loss:3.6411 train_time:128838ms step_avg:168.64ms
step:775/1530 train_loss:3.6063 train_time:129010ms step_avg:168.64ms
step:776/1530 train_loss:3.3955 train_time:129183ms step_avg:168.65ms
step:777/1530 train_loss:3.3906 train_time:129358ms step_avg:168.65ms
step:778/1530 train_loss:3.4903 train_time:129529ms step_avg:168.66ms
step:779/1530 train_loss:3.5794 train_time:129703ms step_avg:168.66ms
step:780/1530 train_loss:3.5880 train_time:129877ms step_avg:168.67ms
step:781/1530 train_loss:3.6756 train_time:130049ms step_avg:168.68ms
step:782/1530 train_loss:3.5858 train_time:130221ms step_avg:168.68ms
step:783/1530 train_loss:3.5680 train_time:130392ms step_avg:168.68ms
step:784/1530 train_loss:3.6090 train_time:130563ms step_avg:168.69ms
step:785/1530 train_loss:3.5599 train_time:130734ms step_avg:168.69ms
step:786/1530 train_loss:3.4410 train_time:130906ms step_avg:168.69ms
step:787/1530 train_loss:3.7250 train_time:131078ms step_avg:168.70ms
step:788/1530 train_loss:3.5044 train_time:131252ms step_avg:168.70ms
step:789/1530 train_loss:3.5526 train_time:131423ms step_avg:168.71ms
step:790/1530 train_loss:3.6238 train_time:131597ms step_avg:168.71ms
step:791/1530 train_loss:3.7706 train_time:131772ms step_avg:168.72ms
step:792/1530 train_loss:3.7601 train_time:131944ms step_avg:168.73ms
step:793/1530 train_loss:3.4467 train_time:132115ms step_avg:168.73ms
step:794/1530 train_loss:3.5958 train_time:132287ms step_avg:168.73ms
step:795/1530 train_loss:3.6706 train_time:132462ms step_avg:168.74ms
step:796/1530 train_loss:3.7463 train_time:132639ms step_avg:168.75ms
step:797/1530 train_loss:3.5255 train_time:132814ms step_avg:168.76ms
step:798/1530 train_loss:3.6456 train_time:132989ms step_avg:168.77ms
step:799/1530 train_loss:3.5353 train_time:133164ms step_avg:168.78ms
step:800/1530 train_loss:3.5306 train_time:133337ms step_avg:168.78ms
step:801/1530 train_loss:3.6310 train_time:133511ms step_avg:168.79ms
step:802/1530 train_loss:3.4954 train_time:133689ms step_avg:168.80ms
step:803/1530 train_loss:3.4919 train_time:133864ms step_avg:168.81ms
step:804/1530 train_loss:3.6251 train_time:134038ms step_avg:168.81ms
step:805/1530 train_loss:3.5259 train_time:134213ms step_avg:168.82ms
step:806/1530 train_loss:3.5595 train_time:134388ms step_avg:168.83ms
step:807/1530 train_loss:3.6448 train_time:134561ms step_avg:168.83ms
step:808/1530 train_loss:3.5488 train_time:134737ms step_avg:168.84ms
step:809/1530 train_loss:3.4961 train_time:134910ms step_avg:168.85ms
step:810/1530 train_loss:3.5590 train_time:135083ms step_avg:168.85ms
step:811/1530 train_loss:3.5783 train_time:135257ms step_avg:168.86ms
step:812/1530 train_loss:3.6023 train_time:135430ms step_avg:168.86ms
step:813/1530 train_loss:3.6284 train_time:135602ms step_avg:168.87ms
step:814/1530 train_loss:3.5661 train_time:135777ms step_avg:168.88ms
step:815/1530 train_loss:3.5657 train_time:135949ms step_avg:168.88ms
step:816/1530 train_loss:3.6863 train_time:136124ms step_avg:168.89ms
step:817/1530 train_loss:3.7682 train_time:136299ms step_avg:168.90ms
step:818/1530 train_loss:3.5204 train_time:136471ms step_avg:168.90ms
step:819/1530 train_loss:3.7198 train_time:136645ms step_avg:168.91ms
step:820/1530 train_loss:3.4957 train_time:136821ms step_avg:168.91ms
step:821/1530 train_loss:3.5646 train_time:136994ms step_avg:168.92ms
step:822/1530 train_loss:3.6997 train_time:137170ms step_avg:168.93ms
step:823/1530 train_loss:3.5728 train_time:137344ms step_avg:168.93ms
step:824/1530 train_loss:3.5154 train_time:137519ms step_avg:168.94ms
step:825/1530 train_loss:3.6208 train_time:137694ms step_avg:168.95ms
step:826/1530 train_loss:3.4832 train_time:137869ms step_avg:168.96ms
step:827/1530 train_loss:3.7367 train_time:138043ms step_avg:168.96ms
step:828/1530 train_loss:3.6214 train_time:138217ms step_avg:168.97ms
step:829/1530 train_loss:3.6287 train_time:138392ms step_avg:168.98ms
step:830/1530 train_loss:3.5378 train_time:138567ms step_avg:168.98ms
step:831/1530 train_loss:3.6013 train_time:138742ms step_avg:168.99ms
step:832/1530 train_loss:3.5168 train_time:138918ms step_avg:169.00ms
step:833/1530 train_loss:3.6479 train_time:139093ms step_avg:169.01ms
step:834/1530 train_loss:3.4727 train_time:139266ms step_avg:169.01ms
step:835/1530 train_loss:3.4530 train_time:139440ms step_avg:169.02ms
step:836/1530 train_loss:3.7178 train_time:139615ms step_avg:169.03ms
step:837/1530 train_loss:3.3942 train_time:139788ms step_avg:169.03ms
step:838/1530 train_loss:3.5921 train_time:139961ms step_avg:169.03ms
step:839/1530 train_loss:3.4235 train_time:140136ms step_avg:169.04ms
step:840/1530 train_loss:3.4664 train_time:140307ms step_avg:169.05ms
step:841/1530 train_loss:3.5670 train_time:140480ms step_avg:169.05ms
step:842/1530 train_loss:3.5833 train_time:140657ms step_avg:169.06ms
step:843/1530 train_loss:3.5634 train_time:140828ms step_avg:169.06ms
step:844/1530 train_loss:3.4316 train_time:141002ms step_avg:169.07ms
step:845/1530 train_loss:3.6641 train_time:141178ms step_avg:169.08ms
step:846/1530 train_loss:3.5193 train_time:141353ms step_avg:169.08ms
step:847/1530 train_loss:3.4933 train_time:141528ms step_avg:169.09ms
step:848/1530 train_loss:3.6441 train_time:141702ms step_avg:169.10ms
step:849/1530 train_loss:3.4918 train_time:141879ms step_avg:169.10ms
step:850/1530 train_loss:3.4391 train_time:142052ms step_avg:169.11ms
step:851/1530 train_loss:3.7344 train_time:142226ms step_avg:169.12ms
step:852/1530 train_loss:3.4365 train_time:142399ms step_avg:169.12ms
step:853/1530 train_loss:3.5626 train_time:142571ms step_avg:169.12ms
step:854/1530 train_loss:3.6528 train_time:142745ms step_avg:169.13ms
step:855/1530 train_loss:3.5138 train_time:142919ms step_avg:169.13ms
step:856/1530 train_loss:3.5428 train_time:143091ms step_avg:169.14ms
step:857/1530 train_loss:3.6120 train_time:143265ms step_avg:169.14ms
step:858/1530 train_loss:3.4691 train_time:143440ms step_avg:169.15ms
step:859/1530 train_loss:3.5649 train_time:143615ms step_avg:169.16ms
step:860/1530 train_loss:3.5875 train_time:143786ms step_avg:169.16ms
step:861/1530 train_loss:3.6261 train_time:143963ms step_avg:169.17ms
step:862/1530 train_loss:3.6042 train_time:144142ms step_avg:169.18ms
step:863/1530 train_loss:3.5667 train_time:144318ms step_avg:169.19ms
step:864/1530 train_loss:3.3797 train_time:144492ms step_avg:169.19ms
step:865/1530 train_loss:3.5979 train_time:144662ms step_avg:169.20ms
step:866/1530 train_loss:3.8908 train_time:144840ms step_avg:169.21ms
step:867/1530 train_loss:3.4637 train_time:145014ms step_avg:169.21ms
step:868/1530 train_loss:3.6436 train_time:145185ms step_avg:169.21ms
step:869/1530 train_loss:3.6160 train_time:145359ms step_avg:169.22ms
step:870/1530 train_loss:3.4512 train_time:145533ms step_avg:169.22ms
step:871/1530 train_loss:3.3995 train_time:145709ms step_avg:169.23ms
step:872/1530 train_loss:3.6488 train_time:145884ms step_avg:169.24ms
step:873/1530 train_loss:3.4631 train_time:146057ms step_avg:169.24ms
step:874/1530 train_loss:3.2218 train_time:146234ms step_avg:169.25ms
step:875/1530 train_loss:3.6320 train_time:146409ms step_avg:169.26ms
step:875/1530 val_loss:3.5194 train_time:146458ms step_avg:169.32ms
step:876/1530 train_loss:3.4343 train_time:146582ms step_avg:169.26ms
step:877/1530 train_loss:3.6148 train_time:146757ms step_avg:169.27ms
step:878/1530 train_loss:3.4702 train_time:146931ms step_avg:169.28ms
step:879/1530 train_loss:3.6521 train_time:147104ms step_avg:169.28ms
step:880/1530 train_loss:3.3061 train_time:147276ms step_avg:169.28ms
step:881/1530 train_loss:3.4779 train_time:147447ms step_avg:169.29ms
step:882/1530 train_loss:3.6973 train_time:147621ms step_avg:169.29ms
step:883/1530 train_loss:3.8404 train_time:147792ms step_avg:169.29ms
step:884/1530 train_loss:3.5594 train_time:147967ms step_avg:169.30ms
step:885/1530 train_loss:3.4956 train_time:148141ms step_avg:169.30ms
step:886/1530 train_loss:3.5691 train_time:148314ms step_avg:169.31ms
step:887/1530 train_loss:4.0926 train_time:148488ms step_avg:169.31ms
step:888/1530 train_loss:3.8365 train_time:148667ms step_avg:169.32ms
step:889/1530 train_loss:3.5199 train_time:148840ms step_avg:169.33ms
step:890/1530 train_loss:3.5330 train_time:149013ms step_avg:169.33ms
step:891/1530 train_loss:3.3571 train_time:149186ms step_avg:169.34ms
step:892/1530 train_loss:3.7236 train_time:149359ms step_avg:169.34ms
step:893/1530 train_loss:3.4196 train_time:149532ms step_avg:169.35ms
step:894/1530 train_loss:3.6348 train_time:149708ms step_avg:169.35ms
step:895/1530 train_loss:3.6754 train_time:149883ms step_avg:169.36ms
step:896/1530 train_loss:3.4936 train_time:150056ms step_avg:169.36ms
step:897/1530 train_loss:3.5419 train_time:150232ms step_avg:169.37ms
step:898/1530 train_loss:3.5905 train_time:150407ms step_avg:169.38ms
step:899/1530 train_loss:3.4770 train_time:150579ms step_avg:169.38ms
step:900/1530 train_loss:3.4246 train_time:150751ms step_avg:169.38ms
step:901/1530 train_loss:3.6199 train_time:150924ms step_avg:169.39ms
step:902/1530 train_loss:3.6334 train_time:151098ms step_avg:169.39ms
step:903/1530 train_loss:3.5447 train_time:151274ms step_avg:169.40ms
step:904/1530 train_loss:3.4973 train_time:151446ms step_avg:169.40ms
step:905/1530 train_loss:3.5004 train_time:151617ms step_avg:169.40ms
step:906/1530 train_loss:3.7045 train_time:151791ms step_avg:169.41ms
step:907/1530 train_loss:3.5242 train_time:151963ms step_avg:169.41ms
step:908/1530 train_loss:3.5687 train_time:152138ms step_avg:169.42ms
step:909/1530 train_loss:3.4535 train_time:152315ms step_avg:169.43ms
step:910/1530 train_loss:3.5259 train_time:152495ms step_avg:169.44ms
step:911/1530 train_loss:3.6403 train_time:152670ms step_avg:169.45ms
step:912/1530 train_loss:3.6118 train_time:152848ms step_avg:169.45ms
step:913/1530 train_loss:3.4626 train_time:153027ms step_avg:169.46ms
step:914/1530 train_loss:3.7463 train_time:153204ms step_avg:169.47ms
step:915/1530 train_loss:3.5363 train_time:153382ms step_avg:169.48ms
step:916/1530 train_loss:3.6193 train_time:153559ms step_avg:169.49ms
step:917/1530 train_loss:3.6017 train_time:153734ms step_avg:169.50ms
step:918/1530 train_loss:4.8231 train_time:153914ms step_avg:169.51ms
step:919/1530 train_loss:3.4930 train_time:154094ms step_avg:169.52ms
step:920/1530 train_loss:3.5908 train_time:154268ms step_avg:169.53ms
step:921/1530 train_loss:3.5550 train_time:154445ms step_avg:169.53ms
step:922/1530 train_loss:3.5813 train_time:154621ms step_avg:169.54ms
step:923/1530 train_loss:3.6140 train_time:154796ms step_avg:169.55ms
step:924/1530 train_loss:3.6825 train_time:154971ms step_avg:169.55ms
step:925/1530 train_loss:3.6436 train_time:155144ms step_avg:169.56ms
step:926/1530 train_loss:3.5494 train_time:155318ms step_avg:169.56ms
step:927/1530 train_loss:3.5493 train_time:155493ms step_avg:169.57ms
step:928/1530 train_loss:3.7779 train_time:155670ms step_avg:169.58ms
step:929/1530 train_loss:3.6080 train_time:155844ms step_avg:169.58ms
step:930/1530 train_loss:3.4064 train_time:156022ms step_avg:169.59ms
step:931/1530 train_loss:3.4930 train_time:156196ms step_avg:169.59ms
step:932/1530 train_loss:3.6497 train_time:156372ms step_avg:169.60ms
step:933/1530 train_loss:3.3638 train_time:156548ms step_avg:169.61ms
step:934/1530 train_loss:3.5808 train_time:156725ms step_avg:169.62ms
step:935/1530 train_loss:3.4385 train_time:156903ms step_avg:169.62ms
step:936/1530 train_loss:3.5165 train_time:157081ms step_avg:169.63ms
step:937/1530 train_loss:3.6255 train_time:157261ms step_avg:169.64ms
step:938/1530 train_loss:3.5433 train_time:157436ms step_avg:169.65ms
step:939/1530 train_loss:3.6669 train_time:157615ms step_avg:169.66ms
step:940/1530 train_loss:3.4788 train_time:157790ms step_avg:169.67ms
step:941/1530 train_loss:3.5452 train_time:157964ms step_avg:169.67ms
step:942/1530 train_loss:3.3503 train_time:158140ms step_avg:169.68ms
step:943/1530 train_loss:3.7111 train_time:158320ms step_avg:169.69ms
step:944/1530 train_loss:3.4004 train_time:158641ms step_avg:169.85ms
step:945/1530 train_loss:3.4256 train_time:158823ms step_avg:169.86ms
step:946/1530 train_loss:5.0641 train_time:159003ms step_avg:169.87ms
step:947/1530 train_loss:3.5994 train_time:159179ms step_avg:169.88ms
step:948/1530 train_loss:3.4841 train_time:159356ms step_avg:169.89ms
step:949/1530 train_loss:3.3752 train_time:159688ms step_avg:170.06ms
step:950/1530 train_loss:3.4423 train_time:159862ms step_avg:170.07ms
step:951/1530 train_loss:3.4067 train_time:160040ms step_avg:170.07ms
step:952/1530 train_loss:3.4762 train_time:160216ms step_avg:170.08ms
step:953/1530 train_loss:3.5681 train_time:160393ms step_avg:170.09ms
step:954/1530 train_loss:3.4425 train_time:160571ms step_avg:170.10ms
step:955/1530 train_loss:3.4745 train_time:160745ms step_avg:170.10ms
step:956/1530 train_loss:3.4434 train_time:160921ms step_avg:170.11ms
step:957/1530 train_loss:3.4929 train_time:161101ms step_avg:170.12ms
step:958/1530 train_loss:3.5061 train_time:161281ms step_avg:170.13ms
step:959/1530 train_loss:3.5079 train_time:161458ms step_avg:170.13ms
step:960/1530 train_loss:3.4078 train_time:161635ms step_avg:170.14ms
step:961/1530 train_loss:3.6479 train_time:161811ms step_avg:170.15ms
step:962/1530 train_loss:3.5889 train_time:161986ms step_avg:170.15ms
step:963/1530 train_loss:3.4022 train_time:162163ms step_avg:170.16ms
step:964/1530 train_loss:3.4265 train_time:162342ms step_avg:170.17ms
step:965/1530 train_loss:3.4800 train_time:162515ms step_avg:170.17ms
step:966/1530 train_loss:3.7008 train_time:162689ms step_avg:170.18ms
step:967/1530 train_loss:3.5200 train_time:162863ms step_avg:170.18ms
step:968/1530 train_loss:3.5073 train_time:163040ms step_avg:170.19ms
step:969/1530 train_loss:3.5853 train_time:163216ms step_avg:170.19ms
step:970/1530 train_loss:3.3739 train_time:163389ms step_avg:170.20ms
step:971/1530 train_loss:3.5312 train_time:163562ms step_avg:170.20ms
step:972/1530 train_loss:3.4806 train_time:163735ms step_avg:170.20ms
step:973/1530 train_loss:3.5377 train_time:163909ms step_avg:170.21ms
step:974/1530 train_loss:3.5910 train_time:164085ms step_avg:170.21ms
step:975/1530 train_loss:3.4631 train_time:164261ms step_avg:170.22ms
step:976/1530 train_loss:3.6701 train_time:164434ms step_avg:170.22ms
step:977/1530 train_loss:3.5655 train_time:164607ms step_avg:170.22ms
step:978/1530 train_loss:3.3560 train_time:164781ms step_avg:170.23ms
step:979/1530 train_loss:3.6228 train_time:164957ms step_avg:170.23ms
step:980/1530 train_loss:3.4144 train_time:165133ms step_avg:170.24ms
step:981/1530 train_loss:3.5748 train_time:165313ms step_avg:170.25ms
step:982/1530 train_loss:3.5406 train_time:165486ms step_avg:170.25ms
step:983/1530 train_loss:3.5102 train_time:165662ms step_avg:170.26ms
step:984/1530 train_loss:3.4971 train_time:165837ms step_avg:170.26ms
step:985/1530 train_loss:3.5768 train_time:166015ms step_avg:170.27ms
step:986/1530 train_loss:3.4122 train_time:166189ms step_avg:170.28ms
step:987/1530 train_loss:3.4851 train_time:166361ms step_avg:170.28ms
step:988/1530 train_loss:3.4610 train_time:166537ms step_avg:170.28ms
step:989/1530 train_loss:3.4168 train_time:166710ms step_avg:170.29ms
step:990/1530 train_loss:3.6570 train_time:166886ms step_avg:170.29ms
step:991/1530 train_loss:3.4708 train_time:167060ms step_avg:170.30ms
step:992/1530 train_loss:3.4440 train_time:167240ms step_avg:170.31ms
step:993/1530 train_loss:3.5020 train_time:167419ms step_avg:170.31ms
step:994/1530 train_loss:3.5928 train_time:167593ms step_avg:170.32ms
step:995/1530 train_loss:3.5267 train_time:167765ms step_avg:170.32ms
step:996/1530 train_loss:3.4558 train_time:167938ms step_avg:170.32ms
step:997/1530 train_loss:3.7555 train_time:168111ms step_avg:170.33ms
step:998/1530 train_loss:3.4399 train_time:168282ms step_avg:170.33ms
step:999/1530 train_loss:3.5898 train_time:168458ms step_avg:170.33ms
step:1000/1530 train_loss:3.4396 train_time:168634ms step_avg:170.34ms
step:1000/1530 val_loss:3.4662 train_time:168685ms step_avg:170.39ms
step:1001/1530 train_loss:3.4983 train_time:168810ms step_avg:170.34ms
step:1002/1530 train_loss:3.3725 train_time:168985ms step_avg:170.35ms
step:1003/1530 train_loss:3.5485 train_time:169162ms step_avg:170.35ms
step:1004/1530 train_loss:3.6007 train_time:169339ms step_avg:170.36ms
step:1005/1530 train_loss:3.3948 train_time:169514ms step_avg:170.37ms
step:1006/1530 train_loss:3.4655 train_time:169692ms step_avg:170.37ms
step:1007/1530 train_loss:3.4378 train_time:169867ms step_avg:170.38ms
step:1008/1530 train_loss:3.5554 train_time:170043ms step_avg:170.38ms
step:1009/1530 train_loss:3.6625 train_time:170221ms step_avg:170.39ms
step:1010/1530 train_loss:3.5616 train_time:170394ms step_avg:170.39ms
step:1011/1530 train_loss:3.5340 train_time:170566ms step_avg:170.40ms
step:1012/1530 train_loss:3.3919 train_time:170740ms step_avg:170.40ms
step:1013/1530 train_loss:3.5408 train_time:170916ms step_avg:170.40ms
step:1014/1530 train_loss:3.6212 train_time:171090ms step_avg:170.41ms
step:1015/1530 train_loss:3.3273 train_time:171267ms step_avg:170.41ms
step:1016/1530 train_loss:3.4106 train_time:171440ms step_avg:170.42ms
step:1017/1530 train_loss:3.3957 train_time:171616ms step_avg:170.42ms
step:1018/1530 train_loss:3.3952 train_time:171791ms step_avg:170.43ms
step:1019/1530 train_loss:3.5225 train_time:171966ms step_avg:170.43ms
step:1020/1530 train_loss:3.3857 train_time:172143ms step_avg:170.44ms
step:1021/1530 train_loss:3.3564 train_time:172318ms step_avg:170.44ms
step:1022/1530 train_loss:3.4818 train_time:172492ms step_avg:170.45ms
step:1023/1530 train_loss:3.5060 train_time:172667ms step_avg:170.45ms
step:1024/1530 train_loss:3.4816 train_time:172844ms step_avg:170.46ms
step:1025/1530 train_loss:3.4769 train_time:173022ms step_avg:170.47ms
step:1026/1530 train_loss:3.6159 train_time:173199ms step_avg:170.47ms
step:1027/1530 train_loss:3.3232 train_time:173376ms step_avg:170.48ms
step:1028/1530 train_loss:3.3975 train_time:173555ms step_avg:170.49ms
step:1029/1530 train_loss:3.3093 train_time:173738ms step_avg:170.50ms
step:1030/1530 train_loss:3.5370 train_time:173914ms step_avg:170.50ms
step:1031/1530 train_loss:3.5079 train_time:174090ms step_avg:170.51ms
step:1032/1530 train_loss:3.6905 train_time:174269ms step_avg:170.52ms
step:1033/1530 train_loss:3.4915 train_time:174445ms step_avg:170.52ms
step:1034/1530 train_loss:3.3977 train_time:174622ms step_avg:170.53ms
step:1035/1530 train_loss:3.4457 train_time:174800ms step_avg:170.54ms
step:1036/1530 train_loss:3.4812 train_time:174977ms step_avg:170.54ms
step:1037/1530 train_loss:3.7875 train_time:175156ms step_avg:170.55ms
step:1038/1530 train_loss:3.6198 train_time:175337ms step_avg:170.56ms
step:1039/1530 train_loss:3.5115 train_time:175518ms step_avg:170.57ms
step:1040/1530 train_loss:3.4111 train_time:175696ms step_avg:170.58ms
step:1041/1530 train_loss:3.4883 train_time:175871ms step_avg:170.58ms
step:1042/1530 train_loss:3.5234 train_time:176044ms step_avg:170.59ms
step:1043/1530 train_loss:3.4482 train_time:176220ms step_avg:170.59ms
step:1044/1530 train_loss:3.4554 train_time:176396ms step_avg:170.60ms
step:1045/1530 train_loss:3.5181 train_time:176574ms step_avg:170.60ms
step:1046/1530 train_loss:3.4217 train_time:176750ms step_avg:170.61ms
step:1047/1530 train_loss:3.6387 train_time:176927ms step_avg:170.61ms
step:1048/1530 train_loss:3.4970 train_time:177102ms step_avg:170.62ms
step:1049/1530 train_loss:3.4019 train_time:177278ms step_avg:170.62ms
step:1050/1530 train_loss:3.3938 train_time:177456ms step_avg:170.63ms
step:1051/1530 train_loss:3.4930 train_time:177634ms step_avg:170.64ms
step:1052/1530 train_loss:3.3640 train_time:177810ms step_avg:170.64ms
step:1053/1530 train_loss:3.6898 train_time:177986ms step_avg:170.65ms
step:1054/1530 train_loss:3.5365 train_time:178166ms step_avg:170.66ms
step:1055/1530 train_loss:3.3841 train_time:178340ms step_avg:170.66ms
step:1056/1530 train_loss:3.4957 train_time:178517ms step_avg:170.67ms
step:1057/1530 train_loss:3.5782 train_time:178694ms step_avg:170.67ms
step:1058/1530 train_loss:3.3026 train_time:178871ms step_avg:170.68ms
step:1059/1530 train_loss:3.3700 train_time:179051ms step_avg:170.69ms
step:1060/1530 train_loss:3.4325 train_time:179226ms step_avg:170.69ms
step:1061/1530 train_loss:3.4184 train_time:179402ms step_avg:170.70ms
step:1062/1530 train_loss:3.3831 train_time:179578ms step_avg:170.70ms
step:1063/1530 train_loss:3.4588 train_time:179753ms step_avg:170.71ms
step:1064/1530 train_loss:3.3847 train_time:179926ms step_avg:170.71ms
step:1065/1530 train_loss:3.3611 train_time:180104ms step_avg:170.72ms
step:1066/1530 train_loss:3.4106 train_time:180281ms step_avg:170.72ms
step:1067/1530 train_loss:3.2787 train_time:180460ms step_avg:170.73ms
step:1068/1530 train_loss:3.4320 train_time:180637ms step_avg:170.73ms
step:1069/1530 train_loss:3.2966 train_time:180818ms step_avg:170.74ms
step:1070/1530 train_loss:3.5684 train_time:180994ms step_avg:170.75ms
step:1071/1530 train_loss:3.5111 train_time:181173ms step_avg:170.76ms
step:1072/1530 train_loss:3.4415 train_time:181348ms step_avg:170.76ms
step:1073/1530 train_loss:3.5211 train_time:181522ms step_avg:170.76ms
step:1074/1530 train_loss:3.4308 train_time:181699ms step_avg:170.77ms
step:1075/1530 train_loss:3.3998 train_time:181877ms step_avg:170.78ms
step:1076/1530 train_loss:3.7956 train_time:182053ms step_avg:170.78ms
step:1077/1530 train_loss:3.4354 train_time:182229ms step_avg:170.79ms
step:1078/1530 train_loss:3.0935 train_time:182413ms step_avg:170.80ms
step:1079/1530 train_loss:3.5278 train_time:182588ms step_avg:170.80ms
step:1080/1530 train_loss:3.4301 train_time:182766ms step_avg:170.81ms
step:1081/1530 train_loss:3.5009 train_time:182941ms step_avg:170.81ms
step:1082/1530 train_loss:3.5917 train_time:183117ms step_avg:170.82ms
step:1083/1530 train_loss:3.4949 train_time:183290ms step_avg:170.82ms
step:1084/1530 train_loss:3.4620 train_time:183466ms step_avg:170.82ms
step:1085/1530 train_loss:3.4288 train_time:183641ms step_avg:170.83ms
step:1086/1530 train_loss:3.6267 train_time:183818ms step_avg:170.83ms
step:1087/1530 train_loss:3.5039 train_time:183992ms step_avg:170.84ms
step:1088/1530 train_loss:3.3697 train_time:184170ms step_avg:170.84ms
step:1089/1530 train_loss:3.3728 train_time:184348ms step_avg:170.85ms
step:1090/1530 train_loss:3.4815 train_time:184526ms step_avg:170.86ms
step:1091/1530 train_loss:3.2852 train_time:184703ms step_avg:170.86ms
step:1092/1530 train_loss:3.4832 train_time:184880ms step_avg:170.87ms
step:1093/1530 train_loss:3.6033 train_time:185057ms step_avg:170.87ms
step:1094/1530 train_loss:3.4463 train_time:185232ms step_avg:170.88ms
step:1095/1530 train_loss:3.4211 train_time:185405ms step_avg:170.88ms
step:1096/1530 train_loss:3.4284 train_time:185583ms step_avg:170.89ms
step:1097/1530 train_loss:3.4886 train_time:185762ms step_avg:170.89ms
step:1098/1530 train_loss:3.5757 train_time:185940ms step_avg:170.90ms
step:1099/1530 train_loss:3.5283 train_time:186117ms step_avg:170.91ms
step:1100/1530 train_loss:3.4270 train_time:186295ms step_avg:170.91ms
step:1101/1530 train_loss:3.2909 train_time:186472ms step_avg:170.92ms
step:1102/1530 train_loss:3.3124 train_time:186651ms step_avg:170.93ms
step:1103/1530 train_loss:3.4398 train_time:186837ms step_avg:170.94ms
step:1104/1530 train_loss:3.3154 train_time:187014ms step_avg:170.95ms
step:1105/1530 train_loss:4.0657 train_time:187193ms step_avg:170.95ms
step:1106/1530 train_loss:3.2217 train_time:187368ms step_avg:170.96ms
step:1107/1530 train_loss:3.5713 train_time:187544ms step_avg:170.96ms
step:1108/1530 train_loss:3.3423 train_time:187719ms step_avg:170.96ms
step:1109/1530 train_loss:3.4988 train_time:187895ms step_avg:170.97ms
step:1110/1530 train_loss:3.4289 train_time:188068ms step_avg:170.97ms
step:1111/1530 train_loss:3.4860 train_time:188244ms step_avg:170.98ms
step:1112/1530 train_loss:3.5579 train_time:188424ms step_avg:170.98ms
step:1113/1530 train_loss:3.4327 train_time:188607ms step_avg:170.99ms
step:1114/1530 train_loss:3.3689 train_time:188786ms step_avg:171.00ms
step:1115/1530 train_loss:3.2367 train_time:188965ms step_avg:171.01ms
step:1116/1530 train_loss:3.4252 train_time:189139ms step_avg:171.01ms
step:1117/1530 train_loss:3.5913 train_time:189318ms step_avg:171.02ms
step:1118/1530 train_loss:3.6253 train_time:189496ms step_avg:171.03ms
step:1119/1530 train_loss:3.4752 train_time:189669ms step_avg:171.03ms
step:1120/1530 train_loss:3.4900 train_time:189847ms step_avg:171.03ms
step:1121/1530 train_loss:3.3926 train_time:190025ms step_avg:171.04ms
step:1122/1530 train_loss:3.4603 train_time:190201ms step_avg:171.04ms
step:1123/1530 train_loss:3.5773 train_time:190377ms step_avg:171.05ms
step:1124/1530 train_loss:3.3335 train_time:190552ms step_avg:171.05ms
step:1125/1530 train_loss:3.2294 train_time:190729ms step_avg:171.06ms
step:1125/1530 val_loss:3.4084 train_time:190779ms step_avg:171.10ms
step:1126/1530 train_loss:3.4759 train_time:190904ms step_avg:171.06ms
step:1127/1530 train_loss:3.6733 train_time:191082ms step_avg:171.07ms
step:1128/1530 train_loss:3.2290 train_time:191260ms step_avg:171.07ms
step:1129/1530 train_loss:3.5559 train_time:191439ms step_avg:171.08ms
step:1130/1530 train_loss:3.3778 train_time:191618ms step_avg:171.09ms
step:1131/1530 train_loss:3.4025 train_time:191801ms step_avg:171.10ms
step:1132/1530 train_loss:3.3699 train_time:191975ms step_avg:171.10ms
step:1133/1530 train_loss:3.4851 train_time:192292ms step_avg:171.23ms
step:1134/1530 train_loss:3.4474 train_time:192476ms step_avg:171.24ms
step:1135/1530 train_loss:3.5220 train_time:192653ms step_avg:171.25ms
step:1136/1530 train_loss:3.5648 train_time:192832ms step_avg:171.25ms
step:1137/1530 train_loss:3.4559 train_time:193009ms step_avg:171.26ms
step:1138/1530 train_loss:3.3518 train_time:193187ms step_avg:171.27ms
step:1139/1530 train_loss:3.6539 train_time:193521ms step_avg:171.41ms
step:1140/1530 train_loss:3.4551 train_time:193697ms step_avg:171.41ms
step:1141/1530 train_loss:3.5970 train_time:193879ms step_avg:171.42ms
step:1142/1530 train_loss:3.4476 train_time:194056ms step_avg:171.43ms
step:1143/1530 train_loss:3.3641 train_time:194236ms step_avg:171.44ms
step:1144/1530 train_loss:3.4423 train_time:194414ms step_avg:171.44ms
step:1145/1530 train_loss:3.5859 train_time:194587ms step_avg:171.44ms
step:1146/1530 train_loss:3.5598 train_time:194768ms step_avg:171.45ms
step:1147/1530 train_loss:3.5217 train_time:194946ms step_avg:171.46ms
step:1148/1530 train_loss:3.4951 train_time:195124ms step_avg:171.46ms
step:1149/1530 train_loss:3.3257 train_time:195303ms step_avg:171.47ms
step:1150/1530 train_loss:3.3768 train_time:195480ms step_avg:171.47ms
step:1151/1530 train_loss:3.3215 train_time:195659ms step_avg:171.48ms
step:1152/1530 train_loss:3.4006 train_time:195843ms step_avg:171.49ms
step:1153/1530 train_loss:3.4309 train_time:196024ms step_avg:171.50ms
step:1154/1530 train_loss:3.5132 train_time:196199ms step_avg:171.50ms
step:1155/1530 train_loss:3.3237 train_time:196380ms step_avg:171.51ms
step:1156/1530 train_loss:3.5361 train_time:196564ms step_avg:171.52ms
step:1157/1530 train_loss:3.4936 train_time:196742ms step_avg:171.53ms
step:1158/1530 train_loss:3.2462 train_time:196920ms step_avg:171.53ms
step:1159/1530 train_loss:3.3448 train_time:197097ms step_avg:171.54ms
step:1160/1530 train_loss:3.3360 train_time:197272ms step_avg:171.54ms
step:1161/1530 train_loss:3.0890 train_time:197451ms step_avg:171.55ms
step:1162/1530 train_loss:3.4212 train_time:197628ms step_avg:171.55ms
step:1163/1530 train_loss:3.3877 train_time:197806ms step_avg:171.56ms
step:1164/1530 train_loss:3.2909 train_time:197983ms step_avg:171.56ms
step:1165/1530 train_loss:3.2450 train_time:198159ms step_avg:171.57ms
step:1166/1530 train_loss:3.3918 train_time:198339ms step_avg:171.57ms
step:1167/1530 train_loss:3.4092 train_time:198514ms step_avg:171.58ms
step:1168/1530 train_loss:3.7141 train_time:198690ms step_avg:171.58ms
step:1169/1530 train_loss:3.3718 train_time:198866ms step_avg:171.58ms
step:1170/1530 train_loss:3.3920 train_time:199043ms step_avg:171.59ms
step:1171/1530 train_loss:3.2824 train_time:199219ms step_avg:171.59ms
step:1172/1530 train_loss:3.4237 train_time:199394ms step_avg:171.60ms
step:1173/1530 train_loss:3.5393 train_time:199574ms step_avg:171.60ms
step:1174/1530 train_loss:3.3805 train_time:199760ms step_avg:171.61ms
step:1175/1530 train_loss:3.3626 train_time:199940ms step_avg:171.62ms
step:1176/1530 train_loss:3.4235 train_time:200120ms step_avg:171.63ms
step:1177/1530 train_loss:3.4521 train_time:200302ms step_avg:171.64ms
step:1178/1530 train_loss:3.4930 train_time:200478ms step_avg:171.64ms
step:1179/1530 train_loss:3.4006 train_time:200654ms step_avg:171.65ms
step:1180/1530 train_loss:3.3603 train_time:200841ms step_avg:171.66ms
step:1181/1530 train_loss:3.3346 train_time:201020ms step_avg:171.66ms
step:1182/1530 train_loss:3.3730 train_time:201196ms step_avg:171.67ms
step:1183/1530 train_loss:3.3358 train_time:201375ms step_avg:171.68ms
step:1184/1530 train_loss:3.5087 train_time:201553ms step_avg:171.68ms
step:1185/1530 train_loss:3.5386 train_time:201732ms step_avg:171.69ms
step:1186/1530 train_loss:3.3645 train_time:201912ms step_avg:171.69ms
step:1187/1530 train_loss:3.4194 train_time:202097ms step_avg:171.71ms
step:1188/1530 train_loss:3.4405 train_time:202273ms step_avg:171.71ms
step:1189/1530 train_loss:3.2723 train_time:202453ms step_avg:171.72ms
step:1190/1530 train_loss:3.4453 train_time:202633ms step_avg:171.72ms
step:1191/1530 train_loss:3.5821 train_time:202814ms step_avg:171.73ms
step:1192/1530 train_loss:3.3893 train_time:202987ms step_avg:171.73ms
step:1193/1530 train_loss:3.2721 train_time:203163ms step_avg:171.74ms
step:1194/1530 train_loss:3.5561 train_time:203342ms step_avg:171.74ms
step:1195/1530 train_loss:3.3739 train_time:203522ms step_avg:171.75ms
step:1196/1530 train_loss:3.3814 train_time:203708ms step_avg:171.76ms
step:1197/1530 train_loss:3.2929 train_time:203886ms step_avg:171.77ms
step:1198/1530 train_loss:3.2996 train_time:204071ms step_avg:171.78ms
step:1199/1530 train_loss:3.3455 train_time:204251ms step_avg:171.78ms
step:1200/1530 train_loss:3.4480 train_time:204429ms step_avg:171.79ms
step:1201/1530 train_loss:3.4774 train_time:204607ms step_avg:171.79ms
step:1202/1530 train_loss:3.6148 train_time:204797ms step_avg:171.81ms
step:1203/1530 train_loss:3.4034 train_time:204978ms step_avg:171.82ms
step:1204/1530 train_loss:3.3076 train_time:205159ms step_avg:171.82ms
step:1205/1530 train_loss:3.4359 train_time:205336ms step_avg:171.83ms
step:1206/1530 train_loss:3.4730 train_time:205512ms step_avg:171.83ms
step:1207/1530 train_loss:3.5156 train_time:205690ms step_avg:171.84ms
step:1208/1530 train_loss:3.3983 train_time:205865ms step_avg:171.84ms
step:1209/1530 train_loss:3.2456 train_time:206045ms step_avg:171.85ms
step:1210/1530 train_loss:3.3059 train_time:206225ms step_avg:171.85ms
step:1211/1530 train_loss:3.3993 train_time:206403ms step_avg:171.86ms
step:1212/1530 train_loss:3.3971 train_time:206581ms step_avg:171.86ms
step:1213/1530 train_loss:3.4098 train_time:206762ms step_avg:171.87ms
step:1214/1530 train_loss:3.2517 train_time:206944ms step_avg:171.88ms
step:1215/1530 train_loss:3.3951 train_time:207120ms step_avg:171.88ms
step:1216/1530 train_loss:3.3314 train_time:207298ms step_avg:171.89ms
step:1217/1530 train_loss:3.3220 train_time:207476ms step_avg:171.89ms
step:1218/1530 train_loss:3.4017 train_time:207655ms step_avg:171.90ms
step:1219/1530 train_loss:3.2600 train_time:207840ms step_avg:171.91ms
step:1220/1530 train_loss:3.4784 train_time:208017ms step_avg:171.91ms
step:1221/1530 train_loss:3.5062 train_time:208191ms step_avg:171.92ms
step:1222/1530 train_loss:3.4268 train_time:208365ms step_avg:171.92ms
step:1223/1530 train_loss:3.2966 train_time:208543ms step_avg:171.92ms
step:1224/1530 train_loss:3.2523 train_time:208725ms step_avg:171.93ms
step:1225/1530 train_loss:3.3661 train_time:208902ms step_avg:171.94ms
step:1226/1530 train_loss:3.3305 train_time:209081ms step_avg:171.94ms
step:1227/1530 train_loss:3.2732 train_time:209262ms step_avg:171.95ms
step:1228/1530 train_loss:3.4466 train_time:209439ms step_avg:171.95ms
step:1229/1530 train_loss:3.3712 train_time:209618ms step_avg:171.96ms
step:1230/1530 train_loss:3.3990 train_time:209802ms step_avg:171.97ms
step:1231/1530 train_loss:3.5781 train_time:209981ms step_avg:171.97ms
step:1232/1530 train_loss:3.4937 train_time:210161ms step_avg:171.98ms
step:1233/1530 train_loss:3.4276 train_time:210341ms step_avg:171.99ms
step:1234/1530 train_loss:3.5889 train_time:210519ms step_avg:171.99ms
step:1235/1530 train_loss:3.3247 train_time:210700ms step_avg:172.00ms
step:1236/1530 train_loss:3.2895 train_time:210876ms step_avg:172.00ms
step:1237/1530 train_loss:3.2739 train_time:211053ms step_avg:172.01ms
step:1238/1530 train_loss:3.2751 train_time:211238ms step_avg:172.02ms
step:1239/1530 train_loss:3.3378 train_time:211416ms step_avg:172.02ms
step:1240/1530 train_loss:3.3794 train_time:211592ms step_avg:172.03ms
step:1241/1530 train_loss:3.4268 train_time:211771ms step_avg:172.03ms
step:1242/1530 train_loss:3.2975 train_time:211948ms step_avg:172.04ms
step:1243/1530 train_loss:3.4085 train_time:212126ms step_avg:172.04ms
step:1244/1530 train_loss:3.4065 train_time:212300ms step_avg:172.04ms
step:1245/1530 train_loss:3.4069 train_time:212478ms step_avg:172.05ms
step:1246/1530 train_loss:3.2422 train_time:212657ms step_avg:172.05ms
step:1247/1530 train_loss:3.3738 train_time:212834ms step_avg:172.06ms
step:1248/1530 train_loss:3.4263 train_time:213010ms step_avg:172.06ms
step:1249/1530 train_loss:3.4217 train_time:213188ms step_avg:172.06ms
step:1250/1530 train_loss:3.3033 train_time:213367ms step_avg:172.07ms
step:1250/1530 val_loss:3.3553 train_time:213421ms step_avg:172.11ms
step:1251/1530 train_loss:3.4935 train_time:213553ms step_avg:172.08ms
step:1252/1530 train_loss:3.3599 train_time:213729ms step_avg:172.08ms
step:1253/1530 train_loss:3.3108 train_time:213906ms step_avg:172.09ms
step:1254/1530 train_loss:3.4179 train_time:214088ms step_avg:172.10ms
step:1255/1530 train_loss:3.5177 train_time:214277ms step_avg:172.11ms
step:1256/1530 train_loss:3.3047 train_time:214460ms step_avg:172.12ms
step:1257/1530 train_loss:3.3715 train_time:214638ms step_avg:172.12ms
step:1258/1530 train_loss:3.3636 train_time:214821ms step_avg:172.13ms
step:1259/1530 train_loss:3.3309 train_time:215000ms step_avg:172.14ms
step:1260/1530 train_loss:3.2103 train_time:215177ms step_avg:172.14ms
step:1261/1530 train_loss:3.3033 train_time:215357ms step_avg:172.15ms
step:1262/1530 train_loss:3.3233 train_time:215538ms step_avg:172.16ms
step:1263/1530 train_loss:3.2415 train_time:215723ms step_avg:172.17ms
step:1264/1530 train_loss:3.4421 train_time:215898ms step_avg:172.17ms
step:1265/1530 train_loss:3.4223 train_time:216074ms step_avg:172.17ms
step:1266/1530 train_loss:3.4423 train_time:216253ms step_avg:172.18ms
step:1267/1530 train_loss:3.3700 train_time:216432ms step_avg:172.18ms
step:1268/1530 train_loss:3.4106 train_time:216612ms step_avg:172.19ms
step:1269/1530 train_loss:3.2488 train_time:216795ms step_avg:172.20ms
step:1270/1530 train_loss:3.1064 train_time:216971ms step_avg:172.20ms
step:1271/1530 train_loss:3.4057 train_time:217150ms step_avg:172.20ms
step:1272/1530 train_loss:3.3530 train_time:217326ms step_avg:172.21ms
step:1273/1530 train_loss:3.3789 train_time:217508ms step_avg:172.22ms
step:1274/1530 train_loss:3.3576 train_time:217688ms step_avg:172.22ms
step:1275/1530 train_loss:3.4333 train_time:217864ms step_avg:172.22ms
step:1276/1530 train_loss:3.4743 train_time:218038ms step_avg:172.23ms
step:1277/1530 train_loss:3.4089 train_time:218217ms step_avg:172.23ms
step:1278/1530 train_loss:3.4071 train_time:218392ms step_avg:172.23ms
step:1279/1530 train_loss:3.2619 train_time:218574ms step_avg:172.24ms
step:1280/1530 train_loss:3.3637 train_time:218758ms step_avg:172.25ms
step:1281/1530 train_loss:3.4271 train_time:218935ms step_avg:172.25ms
step:1282/1530 train_loss:3.4630 train_time:219110ms step_avg:172.26ms
step:1283/1530 train_loss:3.3323 train_time:219290ms step_avg:172.26ms
step:1284/1530 train_loss:3.3673 train_time:219468ms step_avg:172.27ms
step:1285/1530 train_loss:3.3656 train_time:219647ms step_avg:172.27ms
step:1286/1530 train_loss:3.3315 train_time:219825ms step_avg:172.28ms
step:1287/1530 train_loss:3.4900 train_time:220004ms step_avg:172.28ms
step:1288/1530 train_loss:3.2951 train_time:220184ms step_avg:172.29ms
step:1289/1530 train_loss:3.3837 train_time:220370ms step_avg:172.30ms
step:1290/1530 train_loss:3.4595 train_time:220555ms step_avg:172.31ms
step:1291/1530 train_loss:3.3796 train_time:220736ms step_avg:172.32ms
step:1292/1530 train_loss:3.4812 train_time:220918ms step_avg:172.32ms
step:1293/1530 train_loss:3.5132 train_time:221099ms step_avg:172.33ms
step:1294/1530 train_loss:3.4583 train_time:221277ms step_avg:172.33ms
step:1295/1530 train_loss:3.2822 train_time:221455ms step_avg:172.34ms
step:1296/1530 train_loss:3.3766 train_time:221636ms step_avg:172.34ms
step:1297/1530 train_loss:3.2771 train_time:221817ms step_avg:172.35ms
step:1298/1530 train_loss:3.2739 train_time:221997ms step_avg:172.36ms
step:1299/1530 train_loss:3.3972 train_time:222175ms step_avg:172.36ms
step:1300/1530 train_loss:3.4074 train_time:222351ms step_avg:172.37ms
step:1301/1530 train_loss:3.4049 train_time:222528ms step_avg:172.37ms
step:1302/1530 train_loss:3.5759 train_time:222710ms step_avg:172.38ms
step:1303/1530 train_loss:3.3025 train_time:222893ms step_avg:172.38ms
step:1304/1530 train_loss:3.5103 train_time:223074ms step_avg:172.39ms
step:1305/1530 train_loss:3.2626 train_time:223249ms step_avg:172.39ms
step:1306/1530 train_loss:3.4515 train_time:223431ms step_avg:172.40ms
step:1307/1530 train_loss:3.4547 train_time:223606ms step_avg:172.40ms
step:1308/1530 train_loss:3.2827 train_time:223784ms step_avg:172.41ms
step:1309/1530 train_loss:3.3112 train_time:223963ms step_avg:172.41ms
step:1310/1530 train_loss:3.2881 train_time:224140ms step_avg:172.42ms
step:1311/1530 train_loss:3.2969 train_time:224318ms step_avg:172.42ms
step:1312/1530 train_loss:3.3739 train_time:224496ms step_avg:172.42ms
step:1313/1530 train_loss:3.3405 train_time:224672ms step_avg:172.43ms
step:1314/1530 train_loss:3.0474 train_time:224853ms step_avg:172.43ms
step:1315/1530 train_loss:3.2788 train_time:225030ms step_avg:172.44ms
step:1316/1530 train_loss:3.3997 train_time:225206ms step_avg:172.44ms
step:1317/1530 train_loss:3.4229 train_time:225385ms step_avg:172.44ms
step:1318/1530 train_loss:3.3023 train_time:225570ms step_avg:172.45ms
step:1319/1530 train_loss:3.4258 train_time:225749ms step_avg:172.46ms
step:1320/1530 train_loss:3.4609 train_time:225932ms step_avg:172.47ms
step:1321/1530 train_loss:3.3686 train_time:226111ms step_avg:172.47ms
step:1322/1530 train_loss:3.3253 train_time:226430ms step_avg:172.58ms
step:1323/1530 train_loss:3.3210 train_time:226616ms step_avg:172.59ms
step:1324/1530 train_loss:3.4394 train_time:226796ms step_avg:172.60ms
step:1325/1530 train_loss:3.4920 train_time:226979ms step_avg:172.61ms
step:1326/1530 train_loss:3.2155 train_time:227158ms step_avg:172.61ms
step:1327/1530 train_loss:3.1633 train_time:227335ms step_avg:172.62ms
step:1328/1530 train_loss:3.4948 train_time:227514ms step_avg:172.62ms
step:1329/1530 train_loss:3.2975 train_time:227859ms step_avg:172.75ms
step:1330/1530 train_loss:3.4295 train_time:228042ms step_avg:172.76ms
step:1331/1530 train_loss:3.3290 train_time:228219ms step_avg:172.76ms
step:1332/1530 train_loss:3.7369 train_time:228399ms step_avg:172.77ms
step:1333/1530 train_loss:3.4798 train_time:228579ms step_avg:172.77ms
step:1334/1530 train_loss:3.3684 train_time:228757ms step_avg:172.78ms
step:1335/1530 train_loss:3.2962 train_time:228934ms step_avg:172.78ms
step:1336/1530 train_loss:3.2987 train_time:229118ms step_avg:172.79ms
step:1337/1530 train_loss:3.5469 train_time:229297ms step_avg:172.79ms
step:1338/1530 train_loss:3.5193 train_time:229474ms step_avg:172.80ms
step:1339/1530 train_loss:3.3405 train_time:229654ms step_avg:172.80ms
step:1340/1530 train_loss:3.2854 train_time:229833ms step_avg:172.81ms
step:1341/1530 train_loss:3.5961 train_time:230010ms step_avg:172.81ms
step:1342/1530 train_loss:3.3558 train_time:230190ms step_avg:172.82ms
step:1343/1530 train_loss:3.3676 train_time:230367ms step_avg:172.82ms
step:1344/1530 train_loss:3.4158 train_time:230547ms step_avg:172.82ms
step:1345/1530 train_loss:3.3842 train_time:230730ms step_avg:172.83ms
step:1346/1530 train_loss:3.2953 train_time:230906ms step_avg:172.83ms
step:1347/1530 train_loss:3.2756 train_time:231082ms step_avg:172.84ms
step:1348/1530 train_loss:3.3488 train_time:231260ms step_avg:172.84ms
step:1349/1530 train_loss:3.2760 train_time:231436ms step_avg:172.84ms
step:1350/1530 train_loss:3.3890 train_time:231616ms step_avg:172.85ms
step:1351/1530 train_loss:3.2452 train_time:231791ms step_avg:172.85ms
step:1352/1530 train_loss:3.3143 train_time:231969ms step_avg:172.85ms
step:1353/1530 train_loss:3.4046 train_time:232147ms step_avg:172.86ms
step:1354/1530 train_loss:3.2614 train_time:232327ms step_avg:172.86ms
step:1355/1530 train_loss:3.1946 train_time:232503ms step_avg:172.86ms
step:1356/1530 train_loss:3.5119 train_time:232683ms step_avg:172.87ms
step:1357/1530 train_loss:3.4241 train_time:232864ms step_avg:172.88ms
step:1358/1530 train_loss:3.1851 train_time:233041ms step_avg:172.88ms
step:1359/1530 train_loss:3.4424 train_time:233221ms step_avg:172.88ms
step:1360/1530 train_loss:3.3574 train_time:233402ms step_avg:172.89ms
step:1361/1530 train_loss:3.1283 train_time:233588ms step_avg:172.90ms
step:1362/1530 train_loss:3.3937 train_time:233770ms step_avg:172.91ms
step:1363/1530 train_loss:3.2812 train_time:233955ms step_avg:172.92ms
step:1364/1530 train_loss:3.3091 train_time:234134ms step_avg:172.92ms
step:1365/1530 train_loss:3.3144 train_time:234313ms step_avg:172.92ms
step:1366/1530 train_loss:3.4257 train_time:234493ms step_avg:172.93ms
step:1367/1530 train_loss:3.4011 train_time:234669ms step_avg:172.93ms
step:1368/1530 train_loss:3.3512 train_time:234850ms step_avg:172.94ms
step:1369/1530 train_loss:3.2769 train_time:235037ms step_avg:172.95ms
step:1370/1530 train_loss:3.6085 train_time:235218ms step_avg:172.95ms
step:1371/1530 train_loss:3.3116 train_time:235399ms step_avg:172.96ms
step:1372/1530 train_loss:3.3737 train_time:235584ms step_avg:172.97ms
step:1373/1530 train_loss:3.3702 train_time:235763ms step_avg:172.97ms
step:1374/1530 train_loss:3.1530 train_time:235946ms step_avg:172.98ms
step:1375/1530 train_loss:3.5352 train_time:236126ms step_avg:172.99ms
step:1375/1530 val_loss:3.3129 train_time:236176ms step_avg:173.02ms
step:1376/1530 train_loss:3.3450 train_time:236305ms step_avg:172.99ms
step:1377/1530 train_loss:3.4798 train_time:236481ms step_avg:172.99ms
step:1378/1530 train_loss:3.4714 train_time:236660ms step_avg:173.00ms
step:1379/1530 train_loss:3.1228 train_time:236840ms step_avg:173.00ms
step:1380/1530 train_loss:3.3134 train_time:237019ms step_avg:173.01ms
step:1381/1530 train_loss:3.7008 train_time:237204ms step_avg:173.02ms
step:1382/1530 train_loss:3.2152 train_time:237383ms step_avg:173.02ms
step:1383/1530 train_loss:3.3945 train_time:237563ms step_avg:173.02ms
step:1384/1530 train_loss:3.4771 train_time:237745ms step_avg:173.03ms
step:1385/1530 train_loss:3.4057 train_time:237921ms step_avg:173.03ms
step:1386/1530 train_loss:3.3390 train_time:238100ms step_avg:173.04ms
step:1387/1530 train_loss:3.2026 train_time:238279ms step_avg:173.04ms
step:1388/1530 train_loss:3.3488 train_time:238456ms step_avg:173.04ms
step:1389/1530 train_loss:3.3196 train_time:238639ms step_avg:173.05ms
step:1390/1530 train_loss:3.5689 train_time:238816ms step_avg:173.06ms
step:1391/1530 train_loss:3.2887 train_time:238994ms step_avg:173.06ms
step:1392/1530 train_loss:3.2871 train_time:239174ms step_avg:173.06ms
step:1393/1530 train_loss:3.2392 train_time:239354ms step_avg:173.07ms
step:1394/1530 train_loss:3.5043 train_time:239532ms step_avg:173.07ms
step:1395/1530 train_loss:3.3911 train_time:239712ms step_avg:173.08ms
step:1396/1530 train_loss:3.4061 train_time:239890ms step_avg:173.08ms
step:1397/1530 train_loss:3.3109 train_time:240066ms step_avg:173.08ms
step:1398/1530 train_loss:3.2576 train_time:240242ms step_avg:173.08ms
step:1399/1530 train_loss:3.3223 train_time:240420ms step_avg:173.09ms
step:1400/1530 train_loss:3.3206 train_time:240604ms step_avg:173.10ms
step:1401/1530 train_loss:3.3470 train_time:240780ms step_avg:173.10ms
step:1402/1530 train_loss:3.2990 train_time:240959ms step_avg:173.10ms
step:1403/1530 train_loss:3.4945 train_time:241143ms step_avg:173.11ms
step:1404/1530 train_loss:3.2822 train_time:241321ms step_avg:173.11ms
step:1405/1530 train_loss:3.3161 train_time:241502ms step_avg:173.12ms
step:1406/1530 train_loss:3.3156 train_time:241682ms step_avg:173.12ms
step:1407/1530 train_loss:3.1744 train_time:241857ms step_avg:173.13ms
step:1408/1530 train_loss:3.3146 train_time:242037ms step_avg:173.13ms
step:1409/1530 train_loss:3.3002 train_time:242223ms step_avg:173.14ms
step:1410/1530 train_loss:3.2877 train_time:242402ms step_avg:173.14ms
step:1411/1530 train_loss:3.3676 train_time:242578ms step_avg:173.15ms
step:1412/1530 train_loss:3.3354 train_time:242754ms step_avg:173.15ms
step:1413/1530 train_loss:3.3612 train_time:242934ms step_avg:173.15ms
step:1414/1530 train_loss:3.3293 train_time:243114ms step_avg:173.16ms
step:1415/1530 train_loss:3.4028 train_time:243298ms step_avg:173.17ms
step:1416/1530 train_loss:3.2272 train_time:243486ms step_avg:173.18ms
step:1417/1530 train_loss:3.2841 train_time:243670ms step_avg:173.18ms
step:1418/1530 train_loss:3.3912 train_time:243850ms step_avg:173.19ms
step:1419/1530 train_loss:3.3425 train_time:244032ms step_avg:173.19ms
step:1420/1530 train_loss:3.3667 train_time:244215ms step_avg:173.20ms
step:1421/1530 train_loss:3.3759 train_time:244394ms step_avg:173.21ms
step:1422/1530 train_loss:3.3375 train_time:244574ms step_avg:173.21ms
step:1423/1530 train_loss:3.3184 train_time:244754ms step_avg:173.22ms
step:1424/1530 train_loss:3.3321 train_time:244937ms step_avg:173.22ms
step:1425/1530 train_loss:3.1949 train_time:245123ms step_avg:173.23ms
step:1426/1530 train_loss:3.3272 train_time:245301ms step_avg:173.24ms
step:1427/1530 train_loss:3.2856 train_time:245485ms step_avg:173.24ms
step:1428/1530 train_loss:3.3782 train_time:245661ms step_avg:173.24ms
step:1429/1530 train_loss:3.3540 train_time:245838ms step_avg:173.25ms
step:1430/1530 train_loss:3.2620 train_time:246019ms step_avg:173.25ms
step:1431/1530 train_loss:3.3257 train_time:246201ms step_avg:173.26ms
step:1432/1530 train_loss:3.3421 train_time:246383ms step_avg:173.27ms
step:1433/1530 train_loss:3.1426 train_time:246565ms step_avg:173.27ms
step:1434/1530 train_loss:3.2881 train_time:246751ms step_avg:173.28ms
step:1435/1530 train_loss:3.1229 train_time:246931ms step_avg:173.29ms
step:1436/1530 train_loss:3.2317 train_time:247112ms step_avg:173.29ms
step:1437/1530 train_loss:3.4081 train_time:247289ms step_avg:173.29ms
step:1438/1530 train_loss:3.3815 train_time:247464ms step_avg:173.29ms
step:1439/1530 train_loss:3.3130 train_time:247643ms step_avg:173.30ms
step:1440/1530 train_loss:3.1927 train_time:247818ms step_avg:173.30ms
step:1441/1530 train_loss:3.3376 train_time:247997ms step_avg:173.30ms
step:1442/1530 train_loss:3.3851 train_time:248179ms step_avg:173.31ms
step:1443/1530 train_loss:3.4904 train_time:248366ms step_avg:173.32ms
step:1444/1530 train_loss:3.4511 train_time:248543ms step_avg:173.32ms
step:1445/1530 train_loss:3.3394 train_time:248720ms step_avg:173.32ms
step:1446/1530 train_loss:3.1957 train_time:248900ms step_avg:173.33ms
step:1447/1530 train_loss:3.2947 train_time:249080ms step_avg:173.33ms
step:1448/1530 train_loss:3.2951 train_time:249258ms step_avg:173.34ms
step:1449/1530 train_loss:3.4021 train_time:249437ms step_avg:173.34ms
step:1450/1530 train_loss:3.3897 train_time:249617ms step_avg:173.35ms
step:1451/1530 train_loss:3.2044 train_time:249796ms step_avg:173.35ms
step:1452/1530 train_loss:3.3289 train_time:249977ms step_avg:173.35ms
step:1453/1530 train_loss:3.2633 train_time:250153ms step_avg:173.36ms
step:1454/1530 train_loss:3.2894 train_time:250332ms step_avg:173.36ms
step:1455/1530 train_loss:3.3289 train_time:250516ms step_avg:173.37ms
step:1456/1530 train_loss:3.2839 train_time:250693ms step_avg:173.37ms
step:1457/1530 train_loss:3.1541 train_time:250870ms step_avg:173.37ms
step:1458/1530 train_loss:3.4239 train_time:251048ms step_avg:173.38ms
step:1459/1530 train_loss:3.2713 train_time:251231ms step_avg:173.38ms
step:1460/1530 train_loss:3.3155 train_time:251412ms step_avg:173.39ms
step:1461/1530 train_loss:3.4297 train_time:251593ms step_avg:173.39ms
step:1462/1530 train_loss:3.2641 train_time:251771ms step_avg:173.40ms
step:1463/1530 train_loss:3.4648 train_time:251953ms step_avg:173.40ms
step:1464/1530 train_loss:3.3619 train_time:252132ms step_avg:173.41ms
step:1465/1530 train_loss:3.3579 train_time:252314ms step_avg:173.41ms
step:1466/1530 train_loss:3.2844 train_time:252491ms step_avg:173.41ms
step:1467/1530 train_loss:3.3936 train_time:252671ms step_avg:173.42ms
step:1468/1530 train_loss:3.2890 train_time:252848ms step_avg:173.42ms
step:1469/1530 train_loss:3.2784 train_time:253027ms step_avg:173.42ms
step:1470/1530 train_loss:3.3345 train_time:253210ms step_avg:173.43ms
step:1471/1530 train_loss:3.2597 train_time:253395ms step_avg:173.44ms
step:1472/1530 train_loss:3.2506 train_time:253578ms step_avg:173.45ms
step:1473/1530 train_loss:3.4458 train_time:253756ms step_avg:173.45ms
step:1474/1530 train_loss:3.3134 train_time:253939ms step_avg:173.46ms
step:1475/1530 train_loss:3.1514 train_time:254124ms step_avg:173.46ms
step:1476/1530 train_loss:3.2634 train_time:254303ms step_avg:173.47ms
step:1477/1530 train_loss:3.2419 train_time:254490ms step_avg:173.48ms
step:1478/1530 train_loss:3.3077 train_time:254675ms step_avg:173.48ms
step:1479/1530 train_loss:3.3989 train_time:254857ms step_avg:173.49ms
step:1480/1530 train_loss:3.2694 train_time:255035ms step_avg:173.49ms
step:1481/1530 train_loss:3.4514 train_time:255217ms step_avg:173.50ms
step:1482/1530 train_loss:3.3712 train_time:255403ms step_avg:173.51ms
step:1483/1530 train_loss:3.2831 train_time:255595ms step_avg:173.52ms
step:1484/1530 train_loss:3.2668 train_time:255780ms step_avg:173.53ms
step:1485/1530 train_loss:3.2830 train_time:255960ms step_avg:173.53ms
step:1486/1530 train_loss:3.2275 train_time:256145ms step_avg:173.54ms
step:1487/1530 train_loss:3.3449 train_time:256327ms step_avg:173.55ms
step:1488/1530 train_loss:3.2440 train_time:256511ms step_avg:173.55ms
step:1489/1530 train_loss:3.3161 train_time:256693ms step_avg:173.56ms
step:1490/1530 train_loss:3.2545 train_time:256873ms step_avg:173.56ms
step:1491/1530 train_loss:3.1603 train_time:257053ms step_avg:173.57ms
step:1492/1530 train_loss:3.2706 train_time:257234ms step_avg:173.57ms
step:1493/1530 train_loss:3.4329 train_time:257413ms step_avg:173.58ms
step:1494/1530 train_loss:3.3001 train_time:257593ms step_avg:173.58ms
step:1495/1530 train_loss:3.0307 train_time:257776ms step_avg:173.59ms
step:1496/1530 train_loss:3.3631 train_time:257957ms step_avg:173.59ms
step:1497/1530 train_loss:3.3150 train_time:258139ms step_avg:173.60ms
step:1498/1530 train_loss:3.3479 train_time:258323ms step_avg:173.60ms
step:1499/1530 train_loss:3.3110 train_time:258512ms step_avg:173.61ms
step:1500/1530 train_loss:3.2953 train_time:258704ms step_avg:173.63ms
step:1500/1530 val_loss:3.2816 train_time:258758ms step_avg:173.66ms
step:1501/1530 train_loss:3.0845 train_time:258893ms step_avg:173.64ms
step:1502/1530 train_loss:3.3635 train_time:259086ms step_avg:173.65ms
step:1503/1530 train_loss:3.2451 train_time:259266ms step_avg:173.65ms
step:1504/1530 train_loss:3.2498 train_time:259446ms step_avg:173.66ms
step:1505/1530 train_loss:3.2154 train_time:259625ms step_avg:173.66ms
step:1506/1530 train_loss:3.2851 train_time:259807ms step_avg:173.67ms
step:1507/1530 train_loss:3.1790 train_time:260003ms step_avg:173.68ms
step:1508/1530 train_loss:3.4880 train_time:260188ms step_avg:173.69ms
step:1509/1530 train_loss:3.2811 train_time:260366ms step_avg:173.69ms
step:1510/1530 train_loss:3.2696 train_time:260545ms step_avg:173.70ms
step:1511/1530 train_loss:3.4172 train_time:260861ms step_avg:173.79ms
step:1512/1530 train_loss:3.4191 train_time:261049ms step_avg:173.80ms
step:1513/1530 train_loss:3.2716 train_time:261233ms step_avg:173.81ms
step:1514/1530 train_loss:3.0889 train_time:261412ms step_avg:173.81ms
step:1515/1530 train_loss:3.2411 train_time:261592ms step_avg:173.82ms
step:1516/1530 train_loss:3.2555 train_time:261777ms step_avg:173.82ms
step:1517/1530 train_loss:3.3041 train_time:261958ms step_avg:173.83ms
step:1518/1530 train_loss:3.2053 train_time:262143ms step_avg:173.83ms
step:1519/1530 train_loss:3.5026 train_time:262481ms step_avg:173.94ms
step:1520/1530 train_loss:3.1303 train_time:262664ms step_avg:173.95ms
step:1521/1530 train_loss:3.2090 train_time:262841ms step_avg:173.95ms
step:1522/1530 train_loss:3.3578 train_time:263024ms step_avg:173.96ms
step:1523/1530 train_loss:3.2318 train_time:263202ms step_avg:173.96ms
step:1524/1530 train_loss:3.3479 train_time:263384ms step_avg:173.97ms
step:1525/1530 train_loss:3.3379 train_time:263572ms step_avg:173.97ms
step:1526/1530 train_loss:3.2778 train_time:263763ms step_avg:173.99ms
step:1527/1530 train_loss:3.2926 train_time:263944ms step_avg:173.99ms
step:1528/1530 train_loss:3.4094 train_time:264124ms step_avg:173.99ms
step:1529/1530 train_loss:3.4106 train_time:264303ms step_avg:174.00ms
step:1530/1530 train_loss:3.2380 train_time:264482ms step_avg:174.00ms
step:1530/1530 val_loss:3.2791 train_time:264535ms step_avg:174.04ms