51faed93-0804-418c-9057-0d94c3f94a9c.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 03:16:14 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   38C    P0              75W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          On  | 00000000:3B:00.0 Off |                    0 |
| N/A   31C    P0             115W / 700W |    529MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          On  | 00000000:4C:00.0 Off |                    0 |
| N/A   31C    P0             118W / 700W |    529MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   38C    P0             112W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   39C    P0             123W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   30C    P0             110W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   39C    P0             128W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   30C    P0             119W / 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:31769ms step_avg:nanms
step:2/1530 train_loss:10.0645 train_time:31881ms step_avg:nanms
step:3/1530 train_loss:8.3889 train_time:32038ms step_avg:nanms
step:4/1530 train_loss:7.5858 train_time:32199ms step_avg:nanms
step:5/1530 train_loss:7.4682 train_time:32360ms step_avg:nanms
step:6/1530 train_loss:6.9610 train_time:32519ms step_avg:nanms
step:7/1530 train_loss:7.1938 train_time:32680ms step_avg:nanms
step:8/1530 train_loss:6.7272 train_time:32840ms step_avg:nanms
step:9/1530 train_loss:6.6129 train_time:33000ms step_avg:nanms
step:10/1530 train_loss:6.4778 train_time:33161ms step_avg:nanms
step:11/1530 train_loss:6.4571 train_time:115ms step_avg:nanms
step:12/1530 train_loss:6.3979 train_time:276ms step_avg:nanms
step:13/1530 train_loss:6.2246 train_time:436ms step_avg:145.22ms
step:14/1530 train_loss:6.1766 train_time:596ms step_avg:149.00ms
step:15/1530 train_loss:6.1255 train_time:757ms step_avg:151.37ms
step:16/1530 train_loss:6.1241 train_time:916ms step_avg:152.74ms
step:17/1530 train_loss:6.1636 train_time:1078ms step_avg:153.97ms
step:18/1530 train_loss:5.9646 train_time:1238ms step_avg:154.74ms
step:19/1530 train_loss:5.9665 train_time:1398ms step_avg:155.37ms
step:20/1530 train_loss:5.6527 train_time:1558ms step_avg:155.77ms
step:21/1530 train_loss:5.9246 train_time:1720ms step_avg:156.37ms
step:22/1530 train_loss:6.1633 train_time:1881ms step_avg:156.77ms
step:23/1530 train_loss:5.8529 train_time:2041ms step_avg:157.03ms
step:24/1530 train_loss:5.9985 train_time:2202ms step_avg:157.32ms
step:25/1530 train_loss:5.6675 train_time:2362ms step_avg:157.47ms
step:26/1530 train_loss:5.6063 train_time:2523ms step_avg:157.72ms
step:27/1530 train_loss:5.7418 train_time:2684ms step_avg:157.88ms
step:28/1530 train_loss:5.4140 train_time:2844ms step_avg:158.02ms
step:29/1530 train_loss:5.6604 train_time:3005ms step_avg:158.17ms
step:30/1530 train_loss:5.4494 train_time:3165ms step_avg:158.26ms
step:31/1530 train_loss:5.4206 train_time:3327ms step_avg:158.42ms
step:32/1530 train_loss:5.2890 train_time:3488ms step_avg:158.52ms
step:33/1530 train_loss:5.5723 train_time:3648ms step_avg:158.61ms
step:34/1530 train_loss:5.4897 train_time:3810ms step_avg:158.74ms
step:35/1530 train_loss:5.5933 train_time:3969ms step_avg:158.77ms
step:36/1530 train_loss:5.5515 train_time:4130ms step_avg:158.84ms
step:37/1530 train_loss:5.4401 train_time:4290ms step_avg:158.88ms
step:38/1530 train_loss:5.2921 train_time:4451ms step_avg:158.95ms
step:39/1530 train_loss:5.3112 train_time:4610ms step_avg:158.98ms
step:40/1530 train_loss:5.2388 train_time:4770ms step_avg:159.01ms
step:41/1530 train_loss:5.2186 train_time:4931ms step_avg:159.06ms
step:42/1530 train_loss:5.1743 train_time:5091ms step_avg:159.10ms
step:43/1530 train_loss:5.2598 train_time:5252ms step_avg:159.15ms
step:44/1530 train_loss:5.2245 train_time:5412ms step_avg:159.18ms
step:45/1530 train_loss:5.3787 train_time:5571ms step_avg:159.18ms
step:46/1530 train_loss:5.1736 train_time:5731ms step_avg:159.19ms
step:47/1530 train_loss:5.0532 train_time:5891ms step_avg:159.21ms
step:48/1530 train_loss:5.2116 train_time:6051ms step_avg:159.24ms
step:49/1530 train_loss:5.1432 train_time:6212ms step_avg:159.27ms
step:50/1530 train_loss:5.2353 train_time:6371ms step_avg:159.28ms
step:51/1530 train_loss:5.1240 train_time:6531ms step_avg:159.30ms
step:52/1530 train_loss:5.0093 train_time:6691ms step_avg:159.32ms
step:53/1530 train_loss:5.1556 train_time:6851ms step_avg:159.33ms
step:54/1530 train_loss:5.0106 train_time:7011ms step_avg:159.33ms
step:55/1530 train_loss:5.4146 train_time:7171ms step_avg:159.36ms
step:56/1530 train_loss:5.0248 train_time:7332ms step_avg:159.39ms
step:57/1530 train_loss:4.8784 train_time:7492ms step_avg:159.41ms
step:58/1530 train_loss:5.0471 train_time:7653ms step_avg:159.43ms
step:59/1530 train_loss:5.0361 train_time:7813ms step_avg:159.46ms
step:60/1530 train_loss:5.1390 train_time:7974ms step_avg:159.48ms
step:61/1530 train_loss:4.8671 train_time:8134ms step_avg:159.48ms
step:62/1530 train_loss:4.9887 train_time:8294ms step_avg:159.50ms
step:63/1530 train_loss:4.9775 train_time:8455ms step_avg:159.53ms
step:64/1530 train_loss:4.8769 train_time:8616ms step_avg:159.56ms
step:65/1530 train_loss:4.7899 train_time:8776ms step_avg:159.57ms
step:66/1530 train_loss:4.9105 train_time:8937ms step_avg:159.59ms
step:67/1530 train_loss:4.8061 train_time:9097ms step_avg:159.60ms
step:68/1530 train_loss:5.0799 train_time:9257ms step_avg:159.60ms
step:69/1530 train_loss:4.7173 train_time:9417ms step_avg:159.62ms
step:70/1530 train_loss:4.8425 train_time:9579ms step_avg:159.64ms
step:71/1530 train_loss:4.9709 train_time:9739ms step_avg:159.66ms
step:72/1530 train_loss:4.8881 train_time:9900ms step_avg:159.69ms
step:73/1530 train_loss:4.7664 train_time:10061ms step_avg:159.70ms
step:74/1530 train_loss:4.9120 train_time:10222ms step_avg:159.72ms
step:75/1530 train_loss:4.8803 train_time:10383ms step_avg:159.73ms
step:76/1530 train_loss:4.8130 train_time:10543ms step_avg:159.75ms
step:77/1530 train_loss:4.9224 train_time:10704ms step_avg:159.76ms
step:78/1530 train_loss:5.1217 train_time:10864ms step_avg:159.76ms
step:79/1530 train_loss:4.8318 train_time:11026ms step_avg:159.79ms
step:80/1530 train_loss:4.8635 train_time:11187ms step_avg:159.81ms
step:81/1530 train_loss:4.6568 train_time:11348ms step_avg:159.82ms
step:82/1530 train_loss:4.8201 train_time:11508ms step_avg:159.84ms
step:83/1530 train_loss:4.7802 train_time:11668ms step_avg:159.84ms
step:84/1530 train_loss:4.7654 train_time:11829ms step_avg:159.85ms
step:85/1530 train_loss:4.6330 train_time:11989ms step_avg:159.86ms
step:86/1530 train_loss:4.8375 train_time:12152ms step_avg:159.90ms
step:87/1530 train_loss:4.7417 train_time:12313ms step_avg:159.91ms
step:88/1530 train_loss:4.7588 train_time:12472ms step_avg:159.90ms
step:89/1530 train_loss:4.7139 train_time:12633ms step_avg:159.91ms
step:90/1530 train_loss:4.6441 train_time:12793ms step_avg:159.92ms
step:91/1530 train_loss:4.6282 train_time:12954ms step_avg:159.93ms
step:92/1530 train_loss:4.8062 train_time:13116ms step_avg:159.95ms
step:93/1530 train_loss:4.6224 train_time:13276ms step_avg:159.95ms
step:94/1530 train_loss:4.6352 train_time:13436ms step_avg:159.95ms
step:95/1530 train_loss:4.6821 train_time:13596ms step_avg:159.95ms
step:96/1530 train_loss:4.5774 train_time:13756ms step_avg:159.96ms
step:97/1530 train_loss:4.6319 train_time:13917ms step_avg:159.96ms
step:98/1530 train_loss:4.5690 train_time:14077ms step_avg:159.97ms
step:99/1530 train_loss:4.6615 train_time:14237ms step_avg:159.96ms
step:100/1530 train_loss:4.6742 train_time:14398ms step_avg:159.98ms
step:101/1530 train_loss:4.5428 train_time:14559ms step_avg:159.98ms
step:102/1530 train_loss:4.7115 train_time:14720ms step_avg:160.00ms
step:103/1530 train_loss:4.5963 train_time:14881ms step_avg:160.01ms
step:104/1530 train_loss:4.5353 train_time:15041ms step_avg:160.01ms
step:105/1530 train_loss:4.5792 train_time:15202ms step_avg:160.02ms
step:106/1530 train_loss:4.6441 train_time:15362ms step_avg:160.03ms
step:107/1530 train_loss:4.4971 train_time:15523ms step_avg:160.04ms
step:108/1530 train_loss:4.3557 train_time:15684ms step_avg:160.04ms
step:109/1530 train_loss:4.4815 train_time:15844ms step_avg:160.04ms
step:110/1530 train_loss:4.4805 train_time:16005ms step_avg:160.05ms
step:111/1530 train_loss:4.4131 train_time:16166ms step_avg:160.06ms
step:112/1530 train_loss:4.5787 train_time:16327ms step_avg:160.07ms
step:113/1530 train_loss:4.4986 train_time:16488ms step_avg:160.07ms
step:114/1530 train_loss:4.3620 train_time:16648ms step_avg:160.08ms
step:115/1530 train_loss:4.4982 train_time:16811ms step_avg:160.10ms
step:116/1530 train_loss:4.4583 train_time:16975ms step_avg:160.14ms
step:117/1530 train_loss:4.3649 train_time:17138ms step_avg:160.17ms
step:118/1530 train_loss:4.5887 train_time:17304ms step_avg:160.22ms
step:119/1530 train_loss:4.4588 train_time:17468ms step_avg:160.26ms
step:120/1530 train_loss:4.3324 train_time:17632ms step_avg:160.29ms
step:121/1530 train_loss:4.3037 train_time:17795ms step_avg:160.31ms
step:122/1530 train_loss:4.4481 train_time:17959ms step_avg:160.35ms
step:123/1530 train_loss:4.2743 train_time:18124ms step_avg:160.39ms
step:124/1530 train_loss:4.5685 train_time:18287ms step_avg:160.41ms
step:125/1530 train_loss:4.4368 train_time:18451ms step_avg:160.44ms
step:125/1530 val_loss:4.3998 train_time:18498ms step_avg:160.86ms
step:126/1530 train_loss:4.4178 train_time:18617ms step_avg:160.49ms
step:127/1530 train_loss:4.4352 train_time:18784ms step_avg:160.54ms
step:128/1530 train_loss:4.3691 train_time:18948ms step_avg:160.58ms
step:129/1530 train_loss:4.6728 train_time:19112ms step_avg:160.60ms
step:130/1530 train_loss:4.3475 train_time:19275ms step_avg:160.63ms
step:131/1530 train_loss:4.3799 train_time:19440ms step_avg:160.66ms
step:132/1530 train_loss:4.3286 train_time:19604ms step_avg:160.69ms
step:133/1530 train_loss:4.4386 train_time:19769ms step_avg:160.72ms
step:134/1530 train_loss:4.2688 train_time:19933ms step_avg:160.75ms
step:135/1530 train_loss:4.4397 train_time:20096ms step_avg:160.77ms
step:136/1530 train_loss:4.2002 train_time:20261ms step_avg:160.80ms
step:137/1530 train_loss:4.3581 train_time:20426ms step_avg:160.83ms
step:138/1530 train_loss:4.2699 train_time:20590ms step_avg:160.86ms
step:139/1530 train_loss:4.3830 train_time:20754ms step_avg:160.88ms
step:140/1530 train_loss:4.4707 train_time:20918ms step_avg:160.91ms
step:141/1530 train_loss:4.2978 train_time:21083ms step_avg:160.94ms
step:142/1530 train_loss:4.2807 train_time:21246ms step_avg:160.96ms
step:143/1530 train_loss:4.2406 train_time:21409ms step_avg:160.97ms
step:144/1530 train_loss:4.3382 train_time:21573ms step_avg:161.00ms
step:145/1530 train_loss:4.3023 train_time:21737ms step_avg:161.01ms
step:146/1530 train_loss:4.1518 train_time:21901ms step_avg:161.03ms
step:147/1530 train_loss:4.3118 train_time:22065ms step_avg:161.06ms
step:148/1530 train_loss:4.3518 train_time:22228ms step_avg:161.08ms
step:149/1530 train_loss:4.2977 train_time:22393ms step_avg:161.10ms
step:150/1530 train_loss:4.4399 train_time:22558ms step_avg:161.13ms
step:151/1530 train_loss:4.2659 train_time:22722ms step_avg:161.15ms
step:152/1530 train_loss:4.2489 train_time:22886ms step_avg:161.17ms
step:153/1530 train_loss:4.3506 train_time:23049ms step_avg:161.18ms
step:154/1530 train_loss:4.3549 train_time:23213ms step_avg:161.20ms
step:155/1530 train_loss:4.2538 train_time:23377ms step_avg:161.22ms
step:156/1530 train_loss:4.3301 train_time:23541ms step_avg:161.24ms
step:157/1530 train_loss:4.3904 train_time:23705ms step_avg:161.26ms
step:158/1530 train_loss:4.2431 train_time:23869ms step_avg:161.28ms
step:159/1530 train_loss:4.3066 train_time:24032ms step_avg:161.29ms
step:160/1530 train_loss:4.1132 train_time:24195ms step_avg:161.30ms
step:161/1530 train_loss:4.3391 train_time:24359ms step_avg:161.32ms
step:162/1530 train_loss:4.3549 train_time:24523ms step_avg:161.34ms
step:163/1530 train_loss:4.3395 train_time:24687ms step_avg:161.35ms
step:164/1530 train_loss:4.1759 train_time:24850ms step_avg:161.36ms
step:165/1530 train_loss:4.2711 train_time:25014ms step_avg:161.38ms
step:166/1530 train_loss:4.3315 train_time:25179ms step_avg:161.40ms
step:167/1530 train_loss:4.2066 train_time:25343ms step_avg:161.42ms
step:168/1530 train_loss:4.2766 train_time:25506ms step_avg:161.43ms
step:169/1530 train_loss:4.1470 train_time:25670ms step_avg:161.45ms
step:170/1530 train_loss:4.0134 train_time:25835ms step_avg:161.47ms
step:171/1530 train_loss:4.1854 train_time:25997ms step_avg:161.47ms
step:172/1530 train_loss:4.1904 train_time:26160ms step_avg:161.48ms
step:173/1530 train_loss:4.2502 train_time:26323ms step_avg:161.49ms
step:174/1530 train_loss:4.4112 train_time:26486ms step_avg:161.50ms
step:175/1530 train_loss:4.2289 train_time:26649ms step_avg:161.51ms
step:176/1530 train_loss:4.0842 train_time:26811ms step_avg:161.51ms
step:177/1530 train_loss:4.0576 train_time:26974ms step_avg:161.52ms
step:178/1530 train_loss:4.1693 train_time:27136ms step_avg:161.52ms
step:179/1530 train_loss:4.1021 train_time:27299ms step_avg:161.54ms
step:180/1530 train_loss:4.0986 train_time:27462ms step_avg:161.54ms
step:181/1530 train_loss:4.2878 train_time:27625ms step_avg:161.55ms
step:182/1530 train_loss:4.1394 train_time:27790ms step_avg:161.57ms
step:183/1530 train_loss:4.1162 train_time:27953ms step_avg:161.58ms
step:184/1530 train_loss:4.1111 train_time:28116ms step_avg:161.59ms
step:185/1530 train_loss:4.1910 train_time:28279ms step_avg:161.59ms
step:186/1530 train_loss:4.1555 train_time:28441ms step_avg:161.60ms
step:187/1530 train_loss:4.2112 train_time:28604ms step_avg:161.60ms
step:188/1530 train_loss:4.1515 train_time:28901ms step_avg:162.36ms
step:189/1530 train_loss:4.0987 train_time:29228ms step_avg:163.29ms
step:190/1530 train_loss:4.1924 train_time:29389ms step_avg:163.27ms
step:191/1530 train_loss:4.0737 train_time:29552ms step_avg:163.27ms
step:192/1530 train_loss:4.0215 train_time:29714ms step_avg:163.26ms
step:193/1530 train_loss:4.2553 train_time:29878ms step_avg:163.27ms
step:194/1530 train_loss:4.1636 train_time:30040ms step_avg:163.26ms
step:195/1530 train_loss:4.3371 train_time:30203ms step_avg:163.26ms
step:196/1530 train_loss:4.1562 train_time:30366ms step_avg:163.26ms
step:197/1530 train_loss:4.0292 train_time:30528ms step_avg:163.25ms
step:198/1530 train_loss:4.1609 train_time:30692ms step_avg:163.26ms
step:199/1530 train_loss:4.0166 train_time:30855ms step_avg:163.25ms
step:200/1530 train_loss:4.0993 train_time:31018ms step_avg:163.25ms
step:201/1530 train_loss:3.9830 train_time:31180ms step_avg:163.24ms
step:202/1530 train_loss:4.2410 train_time:31344ms step_avg:163.25ms
step:203/1530 train_loss:4.0522 train_time:31506ms step_avg:163.25ms
step:204/1530 train_loss:4.1804 train_time:31669ms step_avg:163.24ms
step:205/1530 train_loss:4.2351 train_time:31832ms step_avg:163.24ms
step:206/1530 train_loss:3.9410 train_time:31995ms step_avg:163.24ms
step:207/1530 train_loss:4.0612 train_time:32158ms step_avg:163.24ms
step:208/1530 train_loss:4.0787 train_time:32321ms step_avg:163.24ms
step:209/1530 train_loss:4.2191 train_time:32485ms step_avg:163.24ms
step:210/1530 train_loss:4.1622 train_time:32648ms step_avg:163.24ms
step:211/1530 train_loss:4.0444 train_time:32811ms step_avg:163.24ms
step:212/1530 train_loss:4.1180 train_time:32974ms step_avg:163.24ms
step:213/1530 train_loss:4.0467 train_time:33136ms step_avg:163.23ms
step:214/1530 train_loss:4.1118 train_time:33298ms step_avg:163.22ms
step:215/1530 train_loss:3.9338 train_time:33461ms step_avg:163.23ms
step:216/1530 train_loss:3.9888 train_time:33625ms step_avg:163.23ms
step:217/1530 train_loss:4.0036 train_time:33787ms step_avg:163.22ms
step:218/1530 train_loss:4.0723 train_time:33950ms step_avg:163.22ms
step:219/1530 train_loss:4.0639 train_time:34113ms step_avg:163.22ms
step:220/1530 train_loss:4.0668 train_time:34276ms step_avg:163.22ms
step:221/1530 train_loss:4.0820 train_time:34438ms step_avg:163.21ms
step:222/1530 train_loss:3.9935 train_time:34601ms step_avg:163.21ms
step:223/1530 train_loss:3.9833 train_time:34764ms step_avg:163.21ms
step:224/1530 train_loss:4.2858 train_time:34927ms step_avg:163.21ms
step:225/1530 train_loss:3.9108 train_time:35090ms step_avg:163.21ms
step:226/1530 train_loss:3.9824 train_time:35254ms step_avg:163.21ms
step:227/1530 train_loss:3.9651 train_time:35417ms step_avg:163.21ms
step:228/1530 train_loss:4.1324 train_time:35581ms step_avg:163.21ms
step:229/1530 train_loss:3.9155 train_time:35747ms step_avg:163.23ms
step:230/1530 train_loss:4.0301 train_time:35912ms step_avg:163.24ms
step:231/1530 train_loss:3.8906 train_time:36078ms step_avg:163.25ms
step:232/1530 train_loss:3.9448 train_time:36245ms step_avg:163.26ms
step:233/1530 train_loss:4.0724 train_time:36411ms step_avg:163.28ms
step:234/1530 train_loss:4.0199 train_time:36578ms step_avg:163.29ms
step:235/1530 train_loss:3.8948 train_time:36746ms step_avg:163.32ms
step:236/1530 train_loss:4.0695 train_time:36912ms step_avg:163.33ms
step:237/1530 train_loss:4.0747 train_time:37078ms step_avg:163.34ms
step:238/1530 train_loss:3.9346 train_time:37244ms step_avg:163.35ms
step:239/1530 train_loss:4.0626 train_time:37410ms step_avg:163.36ms
step:240/1530 train_loss:4.0926 train_time:37577ms step_avg:163.38ms
step:241/1530 train_loss:3.9462 train_time:37744ms step_avg:163.39ms
step:242/1530 train_loss:4.1316 train_time:37910ms step_avg:163.40ms
step:243/1530 train_loss:3.9946 train_time:38075ms step_avg:163.41ms
step:244/1530 train_loss:4.0660 train_time:38241ms step_avg:163.43ms
step:245/1530 train_loss:4.1245 train_time:38408ms step_avg:163.44ms
step:246/1530 train_loss:4.0441 train_time:38574ms step_avg:163.45ms
step:247/1530 train_loss:3.9996 train_time:38740ms step_avg:163.46ms
step:248/1530 train_loss:4.1029 train_time:38906ms step_avg:163.47ms
step:249/1530 train_loss:3.9103 train_time:39072ms step_avg:163.48ms
step:250/1530 train_loss:3.9604 train_time:39238ms step_avg:163.49ms
step:250/1530 val_loss:3.9894 train_time:39286ms step_avg:163.69ms
step:251/1530 train_loss:4.0598 train_time:39406ms step_avg:163.51ms
step:252/1530 train_loss:4.1612 train_time:39573ms step_avg:163.53ms
step:253/1530 train_loss:3.9132 train_time:39740ms step_avg:163.54ms
step:254/1530 train_loss:3.8765 train_time:39906ms step_avg:163.55ms
step:255/1530 train_loss:4.0662 train_time:40073ms step_avg:163.56ms
step:256/1530 train_loss:3.9759 train_time:40239ms step_avg:163.57ms
step:257/1530 train_loss:3.9824 train_time:40405ms step_avg:163.58ms
step:258/1530 train_loss:3.9715 train_time:40571ms step_avg:163.59ms
step:259/1530 train_loss:4.0168 train_time:40737ms step_avg:163.60ms
step:260/1530 train_loss:4.0436 train_time:40904ms step_avg:163.61ms
step:261/1530 train_loss:4.0082 train_time:41070ms step_avg:163.63ms
step:262/1530 train_loss:3.9838 train_time:41237ms step_avg:163.64ms
step:263/1530 train_loss:3.8713 train_time:41403ms step_avg:163.65ms
step:264/1530 train_loss:3.9697 train_time:41568ms step_avg:163.66ms
step:265/1530 train_loss:3.8536 train_time:41735ms step_avg:163.67ms
step:266/1530 train_loss:3.9116 train_time:41901ms step_avg:163.68ms
step:267/1530 train_loss:3.9143 train_time:42069ms step_avg:163.69ms
step:268/1530 train_loss:3.9516 train_time:42236ms step_avg:163.70ms
step:269/1530 train_loss:3.8354 train_time:42401ms step_avg:163.71ms
step:270/1530 train_loss:4.0866 train_time:42569ms step_avg:163.73ms
step:271/1530 train_loss:3.9557 train_time:42735ms step_avg:163.73ms
step:272/1530 train_loss:3.9247 train_time:42900ms step_avg:163.74ms
step:273/1530 train_loss:3.9430 train_time:43065ms step_avg:163.75ms
step:274/1530 train_loss:4.0322 train_time:43232ms step_avg:163.76ms
step:275/1530 train_loss:4.0474 train_time:43398ms step_avg:163.77ms
step:276/1530 train_loss:4.2171 train_time:43564ms step_avg:163.78ms
step:277/1530 train_loss:4.0276 train_time:43732ms step_avg:163.79ms
step:278/1530 train_loss:4.0737 train_time:43898ms step_avg:163.80ms
step:279/1530 train_loss:3.9856 train_time:44064ms step_avg:163.81ms
step:280/1530 train_loss:4.1962 train_time:44233ms step_avg:163.83ms
step:281/1530 train_loss:3.9626 train_time:44399ms step_avg:163.83ms
step:282/1530 train_loss:3.9336 train_time:44565ms step_avg:163.84ms
step:283/1530 train_loss:3.9135 train_time:44731ms step_avg:163.85ms
step:284/1530 train_loss:4.0304 train_time:44897ms step_avg:163.86ms
step:285/1530 train_loss:4.0412 train_time:45061ms step_avg:163.86ms
step:286/1530 train_loss:4.0732 train_time:45228ms step_avg:163.87ms
step:287/1530 train_loss:3.8991 train_time:45394ms step_avg:163.88ms
step:288/1530 train_loss:4.0028 train_time:45559ms step_avg:163.88ms
step:289/1530 train_loss:3.8824 train_time:45725ms step_avg:163.89ms
step:290/1530 train_loss:3.8574 train_time:45890ms step_avg:163.89ms
step:291/1530 train_loss:3.9046 train_time:46056ms step_avg:163.90ms
step:292/1530 train_loss:3.8521 train_time:46221ms step_avg:163.90ms
step:293/1530 train_loss:3.8963 train_time:46384ms step_avg:163.90ms
step:294/1530 train_loss:3.9285 train_time:46550ms step_avg:163.91ms
step:295/1530 train_loss:3.8278 train_time:46716ms step_avg:163.92ms
step:296/1530 train_loss:3.8441 train_time:46881ms step_avg:163.92ms
step:297/1530 train_loss:3.8545 train_time:47046ms step_avg:163.92ms
step:298/1530 train_loss:3.9566 train_time:47212ms step_avg:163.93ms
step:299/1530 train_loss:3.8096 train_time:47376ms step_avg:163.93ms
step:300/1530 train_loss:3.9541 train_time:47541ms step_avg:163.94ms
step:301/1530 train_loss:3.9547 train_time:47706ms step_avg:163.94ms
step:302/1530 train_loss:3.9268 train_time:47872ms step_avg:163.95ms
step:303/1530 train_loss:3.9666 train_time:48037ms step_avg:163.95ms
step:304/1530 train_loss:3.9521 train_time:48201ms step_avg:163.95ms
step:305/1530 train_loss:4.4356 train_time:48367ms step_avg:163.96ms
step:306/1530 train_loss:3.9210 train_time:48533ms step_avg:163.96ms
step:307/1530 train_loss:3.8185 train_time:48698ms step_avg:163.97ms
step:308/1530 train_loss:3.9650 train_time:48864ms step_avg:163.97ms
step:309/1530 train_loss:3.8789 train_time:49031ms step_avg:163.98ms
step:310/1530 train_loss:4.0742 train_time:49196ms step_avg:163.99ms
step:311/1530 train_loss:3.9203 train_time:49360ms step_avg:163.99ms
step:312/1530 train_loss:3.8499 train_time:49525ms step_avg:163.99ms
step:313/1530 train_loss:3.9211 train_time:49691ms step_avg:164.00ms
step:314/1530 train_loss:4.0412 train_time:49856ms step_avg:164.00ms
step:315/1530 train_loss:3.9288 train_time:50021ms step_avg:164.00ms
step:316/1530 train_loss:3.7834 train_time:50187ms step_avg:164.01ms
step:317/1530 train_loss:3.8693 train_time:50353ms step_avg:164.02ms
step:318/1530 train_loss:3.9114 train_time:50518ms step_avg:164.02ms
step:319/1530 train_loss:3.8800 train_time:50682ms step_avg:164.02ms
step:320/1530 train_loss:4.0024 train_time:50848ms step_avg:164.03ms
step:321/1530 train_loss:3.9444 train_time:51013ms step_avg:164.03ms
step:322/1530 train_loss:3.9175 train_time:51178ms step_avg:164.03ms
step:323/1530 train_loss:3.9959 train_time:51344ms step_avg:164.04ms
step:324/1530 train_loss:3.9361 train_time:51511ms step_avg:164.05ms
step:325/1530 train_loss:4.0039 train_time:51676ms step_avg:164.05ms
step:326/1530 train_loss:3.8848 train_time:51842ms step_avg:164.06ms
step:327/1530 train_loss:4.3787 train_time:52008ms step_avg:164.06ms
step:328/1530 train_loss:4.0634 train_time:52174ms step_avg:164.07ms
step:329/1530 train_loss:3.7923 train_time:52341ms step_avg:164.08ms
step:330/1530 train_loss:3.7421 train_time:52508ms step_avg:164.09ms
step:331/1530 train_loss:3.9644 train_time:52673ms step_avg:164.09ms
step:332/1530 train_loss:3.8993 train_time:52838ms step_avg:164.09ms
step:333/1530 train_loss:3.8676 train_time:53003ms step_avg:164.10ms
step:334/1530 train_loss:3.8289 train_time:53168ms step_avg:164.10ms
step:335/1530 train_loss:4.0021 train_time:53334ms step_avg:164.10ms
step:336/1530 train_loss:3.9589 train_time:53498ms step_avg:164.10ms
step:337/1530 train_loss:4.4109 train_time:53663ms step_avg:164.11ms
step:338/1530 train_loss:3.9255 train_time:53830ms step_avg:164.12ms
step:339/1530 train_loss:3.8500 train_time:53995ms step_avg:164.12ms
step:340/1530 train_loss:3.9277 train_time:54160ms step_avg:164.12ms
step:341/1530 train_loss:3.8527 train_time:54329ms step_avg:164.14ms
step:342/1530 train_loss:3.7976 train_time:54496ms step_avg:164.15ms
step:343/1530 train_loss:3.8235 train_time:54664ms step_avg:164.16ms
step:344/1530 train_loss:3.9853 train_time:54833ms step_avg:164.17ms
step:345/1530 train_loss:3.8117 train_time:55001ms step_avg:164.18ms
step:346/1530 train_loss:3.7553 train_time:55169ms step_avg:164.19ms
step:347/1530 train_loss:3.7772 train_time:55338ms step_avg:164.21ms
step:348/1530 train_loss:3.8457 train_time:55505ms step_avg:164.22ms
step:349/1530 train_loss:3.8252 train_time:55673ms step_avg:164.23ms
step:350/1530 train_loss:3.5651 train_time:55841ms step_avg:164.24ms
step:351/1530 train_loss:3.8221 train_time:56009ms step_avg:164.25ms
step:352/1530 train_loss:4.1781 train_time:56177ms step_avg:164.26ms
step:353/1530 train_loss:3.6505 train_time:56345ms step_avg:164.27ms
step:354/1530 train_loss:3.9296 train_time:56513ms step_avg:164.28ms
step:355/1530 train_loss:3.7819 train_time:56680ms step_avg:164.29ms
step:356/1530 train_loss:3.8803 train_time:56848ms step_avg:164.30ms
step:357/1530 train_loss:3.7491 train_time:57017ms step_avg:164.31ms
step:358/1530 train_loss:3.8541 train_time:57186ms step_avg:164.33ms
step:359/1530 train_loss:3.7865 train_time:57355ms step_avg:164.34ms
step:360/1530 train_loss:3.4130 train_time:57525ms step_avg:164.36ms
step:361/1530 train_loss:4.0106 train_time:57695ms step_avg:164.37ms
step:362/1530 train_loss:3.9016 train_time:57862ms step_avg:164.38ms
step:363/1530 train_loss:3.8288 train_time:58030ms step_avg:164.39ms
step:364/1530 train_loss:3.7316 train_time:58199ms step_avg:164.40ms
step:365/1530 train_loss:3.9085 train_time:58368ms step_avg:164.42ms
step:366/1530 train_loss:3.8505 train_time:58536ms step_avg:164.43ms
step:367/1530 train_loss:3.8425 train_time:58703ms step_avg:164.43ms
step:368/1530 train_loss:3.8464 train_time:58873ms step_avg:164.45ms
step:369/1530 train_loss:3.7406 train_time:59041ms step_avg:164.46ms
step:370/1530 train_loss:3.8638 train_time:59209ms step_avg:164.47ms
step:371/1530 train_loss:3.7243 train_time:59376ms step_avg:164.48ms
step:372/1530 train_loss:3.6887 train_time:59544ms step_avg:164.49ms
step:373/1530 train_loss:3.9027 train_time:59713ms step_avg:164.50ms
step:374/1530 train_loss:3.8200 train_time:59882ms step_avg:164.51ms
step:375/1530 train_loss:3.7891 train_time:60051ms step_avg:164.52ms
step:375/1530 val_loss:3.8173 train_time:60099ms step_avg:164.65ms
step:376/1530 train_loss:3.8597 train_time:60219ms step_avg:164.53ms
step:377/1530 train_loss:3.7844 train_time:60523ms step_avg:164.91ms
step:378/1530 train_loss:3.8325 train_time:60700ms step_avg:164.95ms
step:379/1530 train_loss:3.8569 train_time:61024ms step_avg:165.38ms
step:380/1530 train_loss:3.9543 train_time:61191ms step_avg:165.38ms
step:381/1530 train_loss:3.8341 train_time:61360ms step_avg:165.39ms
step:382/1530 train_loss:3.7967 train_time:61529ms step_avg:165.40ms
step:383/1530 train_loss:3.7863 train_time:61697ms step_avg:165.41ms
step:384/1530 train_loss:3.8663 train_time:61865ms step_avg:165.41ms
step:385/1530 train_loss:3.7894 train_time:62032ms step_avg:165.42ms
step:386/1530 train_loss:3.8790 train_time:62200ms step_avg:165.43ms
step:387/1530 train_loss:4.0457 train_time:62369ms step_avg:165.44ms
step:388/1530 train_loss:3.7844 train_time:62538ms step_avg:165.44ms
step:389/1530 train_loss:3.7881 train_time:62705ms step_avg:165.45ms
step:390/1530 train_loss:3.8878 train_time:62873ms step_avg:165.46ms
step:391/1530 train_loss:3.8007 train_time:63039ms step_avg:165.46ms
step:392/1530 train_loss:3.9191 train_time:63207ms step_avg:165.46ms
step:393/1530 train_loss:3.7568 train_time:63374ms step_avg:165.47ms
step:394/1530 train_loss:3.8741 train_time:63543ms step_avg:165.48ms
step:395/1530 train_loss:3.6272 train_time:63711ms step_avg:165.48ms
step:396/1530 train_loss:3.8258 train_time:63878ms step_avg:165.49ms
step:397/1530 train_loss:3.8533 train_time:64047ms step_avg:165.50ms
step:398/1530 train_loss:3.8714 train_time:64214ms step_avg:165.50ms
step:399/1530 train_loss:3.7623 train_time:64381ms step_avg:165.50ms
step:400/1530 train_loss:3.8231 train_time:64549ms step_avg:165.51ms
step:401/1530 train_loss:3.9063 train_time:64715ms step_avg:165.51ms
step:402/1530 train_loss:3.8379 train_time:64882ms step_avg:165.52ms
step:403/1530 train_loss:3.9515 train_time:65051ms step_avg:165.52ms
step:404/1530 train_loss:3.6706 train_time:65217ms step_avg:165.53ms
step:405/1530 train_loss:3.7757 train_time:65385ms step_avg:165.53ms
step:406/1530 train_loss:4.0863 train_time:65552ms step_avg:165.54ms
step:407/1530 train_loss:3.7708 train_time:65720ms step_avg:165.54ms
step:408/1530 train_loss:3.8161 train_time:65887ms step_avg:165.54ms
step:409/1530 train_loss:3.8465 train_time:66053ms step_avg:165.55ms
step:410/1530 train_loss:3.7486 train_time:66220ms step_avg:165.55ms
step:411/1530 train_loss:3.7556 train_time:66387ms step_avg:165.55ms
step:412/1530 train_loss:4.1754 train_time:66555ms step_avg:165.56ms
step:413/1530 train_loss:3.6308 train_time:66722ms step_avg:165.56ms
step:414/1530 train_loss:3.9971 train_time:66888ms step_avg:165.57ms
step:415/1530 train_loss:3.7461 train_time:67056ms step_avg:165.57ms
step:416/1530 train_loss:3.7494 train_time:67224ms step_avg:165.58ms
step:417/1530 train_loss:3.9473 train_time:67391ms step_avg:165.58ms
step:418/1530 train_loss:3.6782 train_time:67559ms step_avg:165.59ms
step:419/1530 train_loss:3.7965 train_time:67726ms step_avg:165.59ms
step:420/1530 train_loss:3.6901 train_time:67893ms step_avg:165.59ms
step:421/1530 train_loss:3.6451 train_time:68061ms step_avg:165.60ms
step:422/1530 train_loss:3.7768 train_time:68228ms step_avg:165.60ms
step:423/1530 train_loss:3.8702 train_time:68394ms step_avg:165.60ms
step:424/1530 train_loss:3.6032 train_time:68562ms step_avg:165.61ms
step:425/1530 train_loss:3.7814 train_time:68729ms step_avg:165.61ms
step:426/1530 train_loss:3.6435 train_time:68895ms step_avg:165.61ms
step:427/1530 train_loss:3.8808 train_time:69065ms step_avg:165.62ms
step:428/1530 train_loss:3.8020 train_time:69232ms step_avg:165.63ms
step:429/1530 train_loss:3.7571 train_time:69400ms step_avg:165.63ms
step:430/1530 train_loss:3.7013 train_time:69568ms step_avg:165.64ms
step:431/1530 train_loss:3.6139 train_time:69735ms step_avg:165.64ms
step:432/1530 train_loss:3.7529 train_time:69903ms step_avg:165.65ms
step:433/1530 train_loss:3.8056 train_time:70072ms step_avg:165.65ms
step:434/1530 train_loss:3.7657 train_time:70240ms step_avg:165.66ms
step:435/1530 train_loss:3.8067 train_time:70407ms step_avg:165.66ms
step:436/1530 train_loss:3.8308 train_time:70573ms step_avg:165.66ms
step:437/1530 train_loss:3.7206 train_time:70740ms step_avg:165.67ms
step:438/1530 train_loss:3.6853 train_time:70907ms step_avg:165.67ms
step:439/1530 train_loss:3.7034 train_time:71076ms step_avg:165.68ms
step:440/1530 train_loss:3.8864 train_time:71245ms step_avg:165.69ms
step:441/1530 train_loss:3.7487 train_time:71411ms step_avg:165.69ms
step:442/1530 train_loss:3.7313 train_time:71579ms step_avg:165.69ms
step:443/1530 train_loss:3.6183 train_time:71748ms step_avg:165.70ms
step:444/1530 train_loss:3.9262 train_time:71915ms step_avg:165.70ms
step:445/1530 train_loss:3.8392 train_time:72082ms step_avg:165.71ms
step:446/1530 train_loss:3.8211 train_time:72249ms step_avg:165.71ms
step:447/1530 train_loss:3.7420 train_time:72417ms step_avg:165.71ms
step:448/1530 train_loss:3.8442 train_time:72584ms step_avg:165.72ms
step:449/1530 train_loss:3.6842 train_time:72751ms step_avg:165.72ms
step:450/1530 train_loss:3.7097 train_time:72918ms step_avg:165.72ms
step:451/1530 train_loss:3.5802 train_time:73086ms step_avg:165.73ms
step:452/1530 train_loss:3.7091 train_time:73253ms step_avg:165.73ms
step:453/1530 train_loss:3.6625 train_time:73420ms step_avg:165.73ms
step:454/1530 train_loss:3.6285 train_time:73587ms step_avg:165.74ms
step:455/1530 train_loss:3.8299 train_time:73757ms step_avg:165.75ms
step:456/1530 train_loss:3.7216 train_time:73926ms step_avg:165.75ms
step:457/1530 train_loss:3.7716 train_time:74095ms step_avg:165.76ms
step:458/1530 train_loss:3.8181 train_time:74265ms step_avg:165.77ms
step:459/1530 train_loss:3.6238 train_time:74435ms step_avg:165.78ms
step:460/1530 train_loss:3.7860 train_time:74605ms step_avg:165.79ms
step:461/1530 train_loss:3.6863 train_time:74775ms step_avg:165.80ms
step:462/1530 train_loss:3.7319 train_time:74947ms step_avg:165.81ms
step:463/1530 train_loss:3.7679 train_time:75116ms step_avg:165.82ms
step:464/1530 train_loss:3.7040 train_time:75285ms step_avg:165.83ms
step:465/1530 train_loss:3.7078 train_time:75453ms step_avg:165.83ms
step:466/1530 train_loss:3.7863 train_time:75623ms step_avg:165.84ms
step:467/1530 train_loss:3.8160 train_time:75794ms step_avg:165.85ms
step:468/1530 train_loss:3.7868 train_time:75963ms step_avg:165.86ms
step:469/1530 train_loss:3.6742 train_time:76132ms step_avg:165.87ms
step:470/1530 train_loss:3.7677 train_time:76302ms step_avg:165.87ms
step:471/1530 train_loss:3.8070 train_time:76472ms step_avg:165.88ms
step:472/1530 train_loss:3.7854 train_time:76644ms step_avg:165.90ms
step:473/1530 train_loss:3.7073 train_time:76812ms step_avg:165.90ms
step:474/1530 train_loss:3.5819 train_time:76982ms step_avg:165.91ms
step:475/1530 train_loss:4.0040 train_time:77152ms step_avg:165.92ms
step:476/1530 train_loss:3.7463 train_time:77321ms step_avg:165.92ms
step:477/1530 train_loss:3.5877 train_time:77492ms step_avg:165.94ms
step:478/1530 train_loss:3.8163 train_time:77661ms step_avg:165.94ms
step:479/1530 train_loss:3.7620 train_time:77831ms step_avg:165.95ms
step:480/1530 train_loss:3.9127 train_time:78001ms step_avg:165.96ms
step:481/1530 train_loss:3.7200 train_time:78171ms step_avg:165.97ms
step:482/1530 train_loss:3.5201 train_time:78341ms step_avg:165.98ms
step:483/1530 train_loss:3.8044 train_time:78510ms step_avg:165.98ms
step:484/1530 train_loss:3.6521 train_time:78680ms step_avg:165.99ms
step:485/1530 train_loss:3.6546 train_time:78850ms step_avg:166.00ms
step:486/1530 train_loss:3.5722 train_time:79021ms step_avg:166.01ms
step:487/1530 train_loss:3.6794 train_time:79190ms step_avg:166.02ms
step:488/1530 train_loss:3.8709 train_time:79361ms step_avg:166.03ms
step:489/1530 train_loss:3.6977 train_time:79531ms step_avg:166.04ms
step:490/1530 train_loss:3.5837 train_time:79701ms step_avg:166.04ms
step:491/1530 train_loss:3.6070 train_time:79871ms step_avg:166.05ms
step:492/1530 train_loss:3.7317 train_time:80041ms step_avg:166.06ms
step:493/1530 train_loss:3.5712 train_time:80210ms step_avg:166.07ms
step:494/1530 train_loss:3.6962 train_time:80381ms step_avg:166.08ms
step:495/1530 train_loss:3.6550 train_time:80553ms step_avg:166.09ms
step:496/1530 train_loss:3.5043 train_time:80726ms step_avg:166.10ms
step:497/1530 train_loss:3.7279 train_time:80893ms step_avg:166.11ms
step:498/1530 train_loss:3.7742 train_time:81065ms step_avg:166.12ms
step:499/1530 train_loss:3.8089 train_time:81235ms step_avg:166.12ms
step:500/1530 train_loss:3.7247 train_time:81406ms step_avg:166.13ms
step:500/1530 val_loss:3.6986 train_time:81454ms step_avg:166.23ms
step:501/1530 train_loss:3.7974 train_time:81576ms step_avg:166.14ms
step:502/1530 train_loss:3.7439 train_time:81748ms step_avg:166.15ms
step:503/1530 train_loss:3.7690 train_time:81917ms step_avg:166.16ms
step:504/1530 train_loss:3.7123 train_time:82087ms step_avg:166.17ms
step:505/1530 train_loss:3.8023 train_time:82256ms step_avg:166.17ms
step:506/1530 train_loss:3.6547 train_time:82425ms step_avg:166.18ms
step:507/1530 train_loss:3.7550 train_time:82594ms step_avg:166.19ms
step:508/1530 train_loss:3.8132 train_time:82763ms step_avg:166.19ms
step:509/1530 train_loss:3.7619 train_time:82932ms step_avg:166.20ms
step:510/1530 train_loss:3.5738 train_time:83101ms step_avg:166.20ms
step:511/1530 train_loss:3.7685 train_time:83271ms step_avg:166.21ms
step:512/1530 train_loss:3.7143 train_time:83442ms step_avg:166.22ms
step:513/1530 train_loss:3.6571 train_time:83611ms step_avg:166.22ms
step:514/1530 train_loss:3.7807 train_time:83781ms step_avg:166.23ms
step:515/1530 train_loss:3.7273 train_time:83951ms step_avg:166.24ms
step:516/1530 train_loss:4.0674 train_time:84121ms step_avg:166.25ms
step:517/1530 train_loss:3.6852 train_time:84292ms step_avg:166.26ms
step:518/1530 train_loss:3.7551 train_time:84460ms step_avg:166.26ms
step:519/1530 train_loss:3.6401 train_time:84630ms step_avg:166.27ms
step:520/1530 train_loss:3.6703 train_time:84798ms step_avg:166.27ms
step:521/1530 train_loss:3.6595 train_time:84969ms step_avg:166.28ms
step:522/1530 train_loss:3.6451 train_time:85138ms step_avg:166.28ms
step:523/1530 train_loss:4.2730 train_time:85307ms step_avg:166.29ms
step:524/1530 train_loss:3.7295 train_time:85476ms step_avg:166.30ms
step:525/1530 train_loss:3.6734 train_time:85646ms step_avg:166.30ms
step:526/1530 train_loss:3.6920 train_time:85815ms step_avg:166.31ms
step:527/1530 train_loss:3.6485 train_time:85983ms step_avg:166.31ms
step:528/1530 train_loss:3.6208 train_time:86152ms step_avg:166.32ms
step:529/1530 train_loss:3.8446 train_time:86321ms step_avg:166.32ms
step:530/1530 train_loss:3.6416 train_time:86492ms step_avg:166.33ms
step:531/1530 train_loss:3.9112 train_time:86661ms step_avg:166.33ms
step:532/1530 train_loss:3.7257 train_time:86828ms step_avg:166.34ms
step:533/1530 train_loss:3.6388 train_time:86999ms step_avg:166.35ms
step:534/1530 train_loss:3.6597 train_time:87168ms step_avg:166.35ms
step:535/1530 train_loss:3.5986 train_time:87336ms step_avg:166.35ms
step:536/1530 train_loss:3.7429 train_time:87507ms step_avg:166.36ms
step:537/1530 train_loss:3.7199 train_time:87676ms step_avg:166.37ms
step:538/1530 train_loss:3.6158 train_time:87847ms step_avg:166.38ms
step:539/1530 train_loss:4.0999 train_time:88017ms step_avg:166.38ms
step:540/1530 train_loss:3.6666 train_time:88187ms step_avg:166.39ms
step:541/1530 train_loss:3.7753 train_time:88355ms step_avg:166.39ms
step:542/1530 train_loss:3.5757 train_time:88524ms step_avg:166.40ms
step:543/1530 train_loss:3.5822 train_time:88693ms step_avg:166.40ms
step:544/1530 train_loss:3.6298 train_time:88861ms step_avg:166.41ms
step:545/1530 train_loss:3.5850 train_time:89030ms step_avg:166.41ms
step:546/1530 train_loss:3.6129 train_time:89200ms step_avg:166.42ms
step:547/1530 train_loss:3.6330 train_time:89370ms step_avg:166.43ms
step:548/1530 train_loss:3.6002 train_time:89539ms step_avg:166.43ms
step:549/1530 train_loss:3.7167 train_time:89708ms step_avg:166.43ms
step:550/1530 train_loss:3.6038 train_time:89876ms step_avg:166.44ms
step:551/1530 train_loss:3.6225 train_time:90045ms step_avg:166.44ms
step:552/1530 train_loss:3.9207 train_time:90213ms step_avg:166.45ms
step:553/1530 train_loss:3.7561 train_time:90383ms step_avg:166.45ms
step:554/1530 train_loss:3.7043 train_time:90552ms step_avg:166.46ms
step:555/1530 train_loss:3.6222 train_time:90720ms step_avg:166.46ms
step:556/1530 train_loss:3.6920 train_time:90890ms step_avg:166.47ms
step:557/1530 train_loss:3.2960 train_time:91059ms step_avg:166.47ms
step:558/1530 train_loss:3.6070 train_time:91227ms step_avg:166.47ms
step:559/1530 train_loss:3.6436 train_time:91396ms step_avg:166.48ms
step:560/1530 train_loss:3.6806 train_time:91565ms step_avg:166.48ms
step:561/1530 train_loss:3.6018 train_time:91732ms step_avg:166.48ms
step:562/1530 train_loss:3.5449 train_time:91901ms step_avg:166.49ms
step:563/1530 train_loss:3.7566 train_time:92071ms step_avg:166.49ms
step:564/1530 train_loss:3.5676 train_time:92241ms step_avg:166.50ms
step:565/1530 train_loss:3.6737 train_time:92411ms step_avg:166.51ms
step:566/1530 train_loss:3.6117 train_time:92712ms step_avg:166.75ms
step:567/1530 train_loss:3.5954 train_time:92895ms step_avg:166.78ms
step:568/1530 train_loss:3.6771 train_time:93064ms step_avg:166.78ms
step:569/1530 train_loss:3.6396 train_time:93387ms step_avg:167.06ms
step:570/1530 train_loss:3.6777 train_time:93557ms step_avg:167.07ms
step:571/1530 train_loss:3.7518 train_time:93727ms step_avg:167.07ms
step:572/1530 train_loss:3.7209 train_time:93899ms step_avg:167.08ms
step:573/1530 train_loss:3.7253 train_time:94073ms step_avg:167.09ms
step:574/1530 train_loss:3.7740 train_time:94244ms step_avg:167.10ms
step:575/1530 train_loss:3.7248 train_time:94417ms step_avg:167.11ms
step:576/1530 train_loss:3.7490 train_time:94591ms step_avg:167.12ms
step:577/1530 train_loss:3.6581 train_time:94762ms step_avg:167.13ms
step:578/1530 train_loss:3.6642 train_time:94934ms step_avg:167.14ms
step:579/1530 train_loss:3.6660 train_time:95106ms step_avg:167.15ms
step:580/1530 train_loss:3.5921 train_time:95276ms step_avg:167.15ms
step:581/1530 train_loss:3.6337 train_time:95448ms step_avg:167.16ms
step:582/1530 train_loss:3.8468 train_time:95619ms step_avg:167.17ms
step:583/1530 train_loss:3.6259 train_time:95793ms step_avg:167.18ms
step:584/1530 train_loss:3.5872 train_time:95963ms step_avg:167.18ms
step:585/1530 train_loss:3.7801 train_time:96133ms step_avg:167.19ms
step:586/1530 train_loss:3.5071 train_time:96305ms step_avg:167.20ms
step:587/1530 train_loss:3.6621 train_time:96476ms step_avg:167.20ms
step:588/1530 train_loss:3.6324 train_time:96646ms step_avg:167.21ms
step:589/1530 train_loss:3.9831 train_time:96817ms step_avg:167.21ms
step:590/1530 train_loss:3.7682 train_time:96991ms step_avg:167.23ms
step:591/1530 train_loss:3.5045 train_time:97161ms step_avg:167.23ms
step:592/1530 train_loss:3.5274 train_time:97334ms step_avg:167.24ms
step:593/1530 train_loss:3.4971 train_time:97507ms step_avg:167.25ms
step:594/1530 train_loss:3.5448 train_time:97679ms step_avg:167.26ms
step:595/1530 train_loss:3.9083 train_time:97854ms step_avg:167.27ms
step:596/1530 train_loss:3.6399 train_time:98026ms step_avg:167.28ms
step:597/1530 train_loss:3.5845 train_time:98198ms step_avg:167.29ms
step:598/1530 train_loss:3.6511 train_time:98368ms step_avg:167.29ms
step:599/1530 train_loss:3.4736 train_time:98538ms step_avg:167.30ms
step:600/1530 train_loss:3.5940 train_time:98709ms step_avg:167.30ms
step:601/1530 train_loss:3.6426 train_time:98880ms step_avg:167.31ms
step:602/1530 train_loss:3.6618 train_time:99053ms step_avg:167.32ms
step:603/1530 train_loss:3.7728 train_time:99223ms step_avg:167.32ms
step:604/1530 train_loss:3.6004 train_time:99394ms step_avg:167.33ms
step:605/1530 train_loss:3.6028 train_time:99568ms step_avg:167.34ms
step:606/1530 train_loss:3.5678 train_time:99740ms step_avg:167.35ms
step:607/1530 train_loss:3.8377 train_time:99912ms step_avg:167.36ms
step:608/1530 train_loss:3.6295 train_time:100084ms step_avg:167.36ms
step:609/1530 train_loss:3.6051 train_time:100255ms step_avg:167.37ms
step:610/1530 train_loss:3.6885 train_time:100425ms step_avg:167.38ms
step:611/1530 train_loss:3.5908 train_time:100597ms step_avg:167.38ms
step:612/1530 train_loss:3.5682 train_time:100770ms step_avg:167.39ms
step:613/1530 train_loss:3.7544 train_time:100940ms step_avg:167.40ms
step:614/1530 train_loss:3.6889 train_time:101113ms step_avg:167.41ms
step:615/1530 train_loss:3.7010 train_time:101283ms step_avg:167.41ms
step:616/1530 train_loss:3.6170 train_time:101454ms step_avg:167.42ms
step:617/1530 train_loss:3.5544 train_time:101627ms step_avg:167.43ms
step:618/1530 train_loss:3.6846 train_time:101797ms step_avg:167.43ms
step:619/1530 train_loss:3.5404 train_time:101970ms step_avg:167.44ms
step:620/1530 train_loss:3.5829 train_time:102141ms step_avg:167.44ms
step:621/1530 train_loss:3.9143 train_time:102314ms step_avg:167.45ms
step:622/1530 train_loss:3.5657 train_time:102487ms step_avg:167.46ms
step:623/1530 train_loss:3.5964 train_time:102658ms step_avg:167.47ms
step:624/1530 train_loss:3.6847 train_time:102831ms step_avg:167.48ms
step:625/1530 train_loss:3.6947 train_time:103000ms step_avg:167.48ms
step:625/1530 val_loss:3.6163 train_time:103050ms step_avg:167.56ms
step:626/1530 train_loss:3.7358 train_time:103175ms step_avg:167.49ms
step:627/1530 train_loss:3.7101 train_time:103348ms step_avg:167.50ms
step:628/1530 train_loss:3.7624 train_time:103518ms step_avg:167.50ms
step:629/1530 train_loss:3.5924 train_time:103690ms step_avg:167.51ms
step:630/1530 train_loss:3.7166 train_time:103860ms step_avg:167.52ms
step:631/1530 train_loss:3.7367 train_time:104030ms step_avg:167.52ms
step:632/1530 train_loss:3.6375 train_time:104204ms step_avg:167.53ms
step:633/1530 train_loss:3.6017 train_time:104375ms step_avg:167.54ms
step:634/1530 train_loss:3.6952 train_time:104546ms step_avg:167.54ms
step:635/1530 train_loss:3.9520 train_time:104716ms step_avg:167.55ms
step:636/1530 train_loss:3.5483 train_time:104886ms step_avg:167.55ms
step:637/1530 train_loss:3.3473 train_time:105057ms step_avg:167.55ms
step:638/1530 train_loss:3.5867 train_time:105227ms step_avg:167.56ms
step:639/1530 train_loss:3.6258 train_time:105397ms step_avg:167.56ms
step:640/1530 train_loss:3.5650 train_time:105567ms step_avg:167.57ms
step:641/1530 train_loss:3.5814 train_time:105737ms step_avg:167.57ms
step:642/1530 train_loss:3.6335 train_time:105907ms step_avg:167.57ms
step:643/1530 train_loss:3.5868 train_time:106078ms step_avg:167.58ms
step:644/1530 train_loss:3.5485 train_time:106249ms step_avg:167.59ms
step:645/1530 train_loss:3.7732 train_time:106419ms step_avg:167.59ms
step:646/1530 train_loss:3.6649 train_time:106592ms step_avg:167.60ms
step:647/1530 train_loss:3.6565 train_time:106762ms step_avg:167.60ms
step:648/1530 train_loss:3.7073 train_time:106935ms step_avg:167.61ms
step:649/1530 train_loss:3.7612 train_time:107105ms step_avg:167.61ms
step:650/1530 train_loss:3.6146 train_time:107276ms step_avg:167.62ms
step:651/1530 train_loss:3.7571 train_time:107447ms step_avg:167.62ms
step:652/1530 train_loss:3.5768 train_time:107617ms step_avg:167.63ms
step:653/1530 train_loss:3.6553 train_time:107788ms step_avg:167.63ms
step:654/1530 train_loss:3.4200 train_time:107958ms step_avg:167.64ms
step:655/1530 train_loss:3.5748 train_time:108129ms step_avg:167.64ms
step:656/1530 train_loss:3.5640 train_time:108299ms step_avg:167.64ms
step:657/1530 train_loss:3.4925 train_time:108470ms step_avg:167.65ms
step:658/1530 train_loss:3.6823 train_time:108640ms step_avg:167.65ms
step:659/1530 train_loss:3.5759 train_time:108813ms step_avg:167.66ms
step:660/1530 train_loss:3.6775 train_time:108983ms step_avg:167.67ms
step:661/1530 train_loss:3.7474 train_time:109154ms step_avg:167.67ms
step:662/1530 train_loss:3.6685 train_time:109324ms step_avg:167.67ms
step:663/1530 train_loss:3.5450 train_time:109494ms step_avg:167.68ms
step:664/1530 train_loss:3.6004 train_time:109664ms step_avg:167.68ms
step:665/1530 train_loss:3.4866 train_time:109835ms step_avg:167.69ms
step:666/1530 train_loss:3.7691 train_time:110005ms step_avg:167.69ms
step:667/1530 train_loss:3.5945 train_time:110175ms step_avg:167.69ms
step:668/1530 train_loss:3.6416 train_time:110346ms step_avg:167.70ms
step:669/1530 train_loss:3.4824 train_time:110518ms step_avg:167.71ms
step:670/1530 train_loss:3.5912 train_time:110689ms step_avg:167.71ms
step:671/1530 train_loss:3.5582 train_time:110859ms step_avg:167.71ms
step:672/1530 train_loss:3.5582 train_time:111032ms step_avg:167.72ms
step:673/1530 train_loss:3.8401 train_time:111203ms step_avg:167.73ms
step:674/1530 train_loss:3.6123 train_time:111374ms step_avg:167.73ms
step:675/1530 train_loss:3.7038 train_time:111545ms step_avg:167.74ms
step:676/1530 train_loss:3.4891 train_time:111717ms step_avg:167.74ms
step:677/1530 train_loss:3.5975 train_time:111888ms step_avg:167.75ms
step:678/1530 train_loss:3.5497 train_time:112059ms step_avg:167.75ms
step:679/1530 train_loss:3.6656 train_time:112232ms step_avg:167.76ms
step:680/1530 train_loss:3.5834 train_time:112401ms step_avg:167.76ms
step:681/1530 train_loss:3.6127 train_time:112574ms step_avg:167.77ms
step:682/1530 train_loss:3.6573 train_time:112749ms step_avg:167.78ms
step:683/1530 train_loss:3.7309 train_time:112922ms step_avg:167.79ms
step:684/1530 train_loss:3.6454 train_time:113094ms step_avg:167.80ms
step:685/1530 train_loss:3.6834 train_time:113268ms step_avg:167.80ms
step:686/1530 train_loss:3.6271 train_time:113440ms step_avg:167.81ms
step:687/1530 train_loss:3.6597 train_time:113614ms step_avg:167.82ms
step:688/1530 train_loss:3.2017 train_time:113788ms step_avg:167.83ms
step:689/1530 train_loss:3.4101 train_time:113961ms step_avg:167.84ms
step:690/1530 train_loss:3.5371 train_time:114136ms step_avg:167.85ms
step:691/1530 train_loss:3.4066 train_time:114310ms step_avg:167.86ms
step:692/1530 train_loss:3.6251 train_time:114483ms step_avg:167.86ms
step:693/1530 train_loss:3.6369 train_time:114656ms step_avg:167.87ms
step:694/1530 train_loss:3.5460 train_time:114828ms step_avg:167.88ms
step:695/1530 train_loss:3.5296 train_time:114998ms step_avg:167.88ms
step:696/1530 train_loss:3.8456 train_time:115171ms step_avg:167.89ms
step:697/1530 train_loss:3.5788 train_time:115343ms step_avg:167.89ms
step:698/1530 train_loss:3.6416 train_time:115515ms step_avg:167.90ms
step:699/1530 train_loss:3.7640 train_time:115690ms step_avg:167.91ms
step:700/1530 train_loss:3.5612 train_time:115862ms step_avg:167.92ms
step:701/1530 train_loss:3.5380 train_time:116035ms step_avg:167.92ms
step:702/1530 train_loss:3.5042 train_time:116209ms step_avg:167.93ms
step:703/1530 train_loss:3.4959 train_time:116380ms step_avg:167.94ms
step:704/1530 train_loss:3.5647 train_time:116554ms step_avg:167.95ms
step:705/1530 train_loss:3.5594 train_time:116730ms step_avg:167.96ms
step:706/1530 train_loss:3.5713 train_time:116905ms step_avg:167.97ms
step:707/1530 train_loss:3.6417 train_time:117079ms step_avg:167.98ms
step:708/1530 train_loss:3.5993 train_time:117253ms step_avg:167.98ms
step:709/1530 train_loss:3.5755 train_time:117427ms step_avg:167.99ms
step:710/1530 train_loss:3.5322 train_time:117598ms step_avg:168.00ms
step:711/1530 train_loss:3.5898 train_time:117772ms step_avg:168.01ms
step:712/1530 train_loss:3.6469 train_time:117947ms step_avg:168.02ms
step:713/1530 train_loss:3.6452 train_time:118122ms step_avg:168.03ms
step:714/1530 train_loss:3.5550 train_time:118295ms step_avg:168.03ms
step:715/1530 train_loss:3.5654 train_time:118468ms step_avg:168.04ms
step:716/1530 train_loss:3.5812 train_time:118639ms step_avg:168.04ms
step:717/1530 train_loss:3.7007 train_time:118815ms step_avg:168.06ms
step:718/1530 train_loss:3.5896 train_time:118986ms step_avg:168.06ms
step:719/1530 train_loss:3.6746 train_time:119157ms step_avg:168.06ms
step:720/1530 train_loss:3.8376 train_time:119333ms step_avg:168.07ms
step:721/1530 train_loss:3.4645 train_time:119507ms step_avg:168.08ms
step:722/1530 train_loss:3.7298 train_time:119678ms step_avg:168.09ms
step:723/1530 train_loss:3.7627 train_time:119850ms step_avg:168.09ms
step:724/1530 train_loss:3.5639 train_time:120022ms step_avg:168.10ms
step:725/1530 train_loss:3.6495 train_time:120196ms step_avg:168.11ms
step:726/1530 train_loss:3.5272 train_time:120369ms step_avg:168.11ms
step:727/1530 train_loss:3.5791 train_time:120543ms step_avg:168.12ms
step:728/1530 train_loss:3.7290 train_time:120716ms step_avg:168.13ms
step:729/1530 train_loss:3.6684 train_time:120890ms step_avg:168.14ms
step:730/1530 train_loss:3.6598 train_time:121062ms step_avg:168.14ms
step:731/1530 train_loss:3.5480 train_time:121236ms step_avg:168.15ms
step:732/1530 train_loss:3.5897 train_time:121406ms step_avg:168.15ms
step:733/1530 train_loss:3.8260 train_time:121580ms step_avg:168.16ms
step:734/1530 train_loss:3.5597 train_time:121755ms step_avg:168.17ms
step:735/1530 train_loss:3.6135 train_time:121927ms step_avg:168.17ms
step:736/1530 train_loss:3.7328 train_time:122100ms step_avg:168.18ms
step:737/1530 train_loss:3.6709 train_time:122273ms step_avg:168.19ms
step:738/1530 train_loss:3.5985 train_time:122445ms step_avg:168.19ms
step:739/1530 train_loss:3.4951 train_time:122617ms step_avg:168.20ms
step:740/1530 train_loss:4.1082 train_time:122797ms step_avg:168.22ms
step:741/1530 train_loss:3.4876 train_time:122970ms step_avg:168.22ms
step:742/1530 train_loss:3.5449 train_time:123142ms step_avg:168.23ms
step:743/1530 train_loss:3.5735 train_time:123315ms step_avg:168.23ms
step:744/1530 train_loss:3.6454 train_time:123487ms step_avg:168.24ms
step:745/1530 train_loss:3.5735 train_time:123660ms step_avg:168.25ms
step:746/1530 train_loss:3.5916 train_time:123833ms step_avg:168.25ms
step:747/1530 train_loss:3.6418 train_time:124009ms step_avg:168.26ms
step:748/1530 train_loss:3.5613 train_time:124184ms step_avg:168.27ms
step:749/1530 train_loss:3.5541 train_time:124358ms step_avg:168.28ms
step:750/1530 train_loss:3.5934 train_time:124529ms step_avg:168.28ms
step:750/1530 val_loss:3.5594 train_time:124578ms step_avg:168.35ms
step:751/1530 train_loss:3.5631 train_time:124704ms step_avg:168.29ms
step:752/1530 train_loss:3.6103 train_time:124875ms step_avg:168.29ms
step:753/1530 train_loss:3.6168 train_time:125048ms step_avg:168.30ms
step:754/1530 train_loss:3.5916 train_time:125220ms step_avg:168.31ms
step:755/1530 train_loss:3.6730 train_time:125524ms step_avg:168.49ms
step:756/1530 train_loss:3.4520 train_time:125708ms step_avg:168.51ms
step:757/1530 train_loss:3.7198 train_time:125882ms step_avg:168.52ms
step:758/1530 train_loss:3.6448 train_time:126052ms step_avg:168.52ms
step:759/1530 train_loss:3.5893 train_time:126384ms step_avg:168.74ms
step:760/1530 train_loss:3.7028 train_time:126555ms step_avg:168.74ms
step:761/1530 train_loss:3.3914 train_time:126728ms step_avg:168.75ms
step:762/1530 train_loss:3.5497 train_time:126900ms step_avg:168.75ms
step:763/1530 train_loss:3.6566 train_time:127072ms step_avg:168.75ms
step:764/1530 train_loss:3.3141 train_time:127244ms step_avg:168.76ms
step:765/1530 train_loss:3.7268 train_time:127416ms step_avg:168.76ms
step:766/1530 train_loss:3.5600 train_time:127592ms step_avg:168.77ms
step:767/1530 train_loss:3.5659 train_time:127764ms step_avg:168.78ms
step:768/1530 train_loss:3.5644 train_time:127937ms step_avg:168.78ms
step:769/1530 train_loss:3.5804 train_time:128111ms step_avg:168.79ms
step:770/1530 train_loss:3.6312 train_time:128283ms step_avg:168.79ms
step:771/1530 train_loss:3.8778 train_time:128454ms step_avg:168.80ms
step:772/1530 train_loss:3.4393 train_time:128627ms step_avg:168.80ms
step:773/1530 train_loss:3.6247 train_time:128798ms step_avg:168.80ms
step:774/1530 train_loss:3.6377 train_time:128969ms step_avg:168.81ms
step:775/1530 train_loss:3.6009 train_time:129141ms step_avg:168.81ms
step:776/1530 train_loss:3.4022 train_time:129314ms step_avg:168.82ms
step:777/1530 train_loss:3.3663 train_time:129489ms step_avg:168.83ms
step:778/1530 train_loss:3.4901 train_time:129661ms step_avg:168.83ms
step:779/1530 train_loss:3.5782 train_time:129833ms step_avg:168.83ms
step:780/1530 train_loss:3.5784 train_time:130007ms step_avg:168.84ms
step:781/1530 train_loss:3.6670 train_time:130177ms step_avg:168.84ms
step:782/1530 train_loss:3.5835 train_time:130350ms step_avg:168.85ms
step:783/1530 train_loss:3.5579 train_time:130521ms step_avg:168.85ms
step:784/1530 train_loss:3.5976 train_time:130693ms step_avg:168.85ms
step:785/1530 train_loss:3.5593 train_time:130865ms step_avg:168.86ms
step:786/1530 train_loss:3.4362 train_time:131036ms step_avg:168.86ms
step:787/1530 train_loss:3.7279 train_time:131209ms step_avg:168.87ms
step:788/1530 train_loss:3.4936 train_time:131383ms step_avg:168.87ms
step:789/1530 train_loss:3.5423 train_time:131554ms step_avg:168.87ms
step:790/1530 train_loss:3.6219 train_time:131729ms step_avg:168.88ms
step:791/1530 train_loss:3.7677 train_time:131905ms step_avg:168.89ms
step:792/1530 train_loss:3.7560 train_time:132077ms step_avg:168.90ms
step:793/1530 train_loss:3.4571 train_time:132249ms step_avg:168.90ms
step:794/1530 train_loss:3.5905 train_time:132422ms step_avg:168.91ms
step:795/1530 train_loss:3.6730 train_time:132596ms step_avg:168.91ms
step:796/1530 train_loss:3.7221 train_time:132772ms step_avg:168.92ms
step:797/1530 train_loss:3.5175 train_time:132946ms step_avg:168.93ms
step:798/1530 train_loss:3.6333 train_time:133120ms step_avg:168.93ms
step:799/1530 train_loss:3.5282 train_time:133298ms step_avg:168.95ms
step:800/1530 train_loss:3.5203 train_time:133472ms step_avg:168.95ms
step:801/1530 train_loss:3.6196 train_time:133648ms step_avg:168.96ms
step:802/1530 train_loss:3.4875 train_time:133826ms step_avg:168.97ms
step:803/1530 train_loss:3.4826 train_time:133999ms step_avg:168.98ms
step:804/1530 train_loss:3.6147 train_time:134173ms step_avg:168.98ms
step:805/1530 train_loss:3.5118 train_time:134349ms step_avg:168.99ms
step:806/1530 train_loss:3.5551 train_time:134522ms step_avg:169.00ms
step:807/1530 train_loss:3.6381 train_time:134695ms step_avg:169.00ms
step:808/1530 train_loss:3.5350 train_time:134870ms step_avg:169.01ms
step:809/1530 train_loss:3.4950 train_time:135044ms step_avg:169.02ms
step:810/1530 train_loss:3.5546 train_time:135216ms step_avg:169.02ms
step:811/1530 train_loss:3.5748 train_time:135390ms step_avg:169.03ms
step:812/1530 train_loss:3.5907 train_time:135563ms step_avg:169.03ms
step:813/1530 train_loss:3.6215 train_time:135735ms step_avg:169.03ms
step:814/1530 train_loss:3.5634 train_time:135910ms step_avg:169.04ms
step:815/1530 train_loss:3.5581 train_time:136084ms step_avg:169.05ms
step:816/1530 train_loss:3.6767 train_time:136258ms step_avg:169.05ms
step:817/1530 train_loss:3.7601 train_time:136432ms step_avg:169.06ms
step:818/1530 train_loss:3.5185 train_time:136605ms step_avg:169.07ms
step:819/1530 train_loss:3.7157 train_time:136780ms step_avg:169.07ms
step:820/1530 train_loss:3.4901 train_time:136955ms step_avg:169.08ms
step:821/1530 train_loss:3.5519 train_time:137128ms step_avg:169.08ms
step:822/1530 train_loss:3.6967 train_time:137302ms step_avg:169.09ms
step:823/1530 train_loss:3.5735 train_time:137475ms step_avg:169.10ms
step:824/1530 train_loss:3.5072 train_time:137650ms step_avg:169.10ms
step:825/1530 train_loss:3.6083 train_time:137827ms step_avg:169.11ms
step:826/1530 train_loss:3.4688 train_time:138002ms step_avg:169.12ms
step:827/1530 train_loss:3.7299 train_time:138176ms step_avg:169.13ms
step:828/1530 train_loss:3.6096 train_time:138350ms step_avg:169.13ms
step:829/1530 train_loss:3.6226 train_time:138526ms step_avg:169.14ms
step:830/1530 train_loss:3.5316 train_time:138700ms step_avg:169.15ms
step:831/1530 train_loss:3.5939 train_time:138874ms step_avg:169.15ms
step:832/1530 train_loss:3.5052 train_time:139051ms step_avg:169.16ms
step:833/1530 train_loss:3.6447 train_time:139229ms step_avg:169.17ms
step:834/1530 train_loss:3.4685 train_time:139401ms step_avg:169.18ms
step:835/1530 train_loss:3.4537 train_time:139575ms step_avg:169.18ms
step:836/1530 train_loss:3.7068 train_time:139752ms step_avg:169.19ms
step:837/1530 train_loss:3.3901 train_time:139928ms step_avg:169.20ms
step:838/1530 train_loss:3.5882 train_time:140100ms step_avg:169.20ms
step:839/1530 train_loss:3.4179 train_time:140274ms step_avg:169.21ms
step:840/1530 train_loss:3.4665 train_time:140448ms step_avg:169.21ms
step:841/1530 train_loss:3.5671 train_time:140621ms step_avg:169.22ms
step:842/1530 train_loss:3.5770 train_time:140796ms step_avg:169.23ms
step:843/1530 train_loss:3.5571 train_time:140969ms step_avg:169.23ms
step:844/1530 train_loss:3.4272 train_time:141144ms step_avg:169.24ms
step:845/1530 train_loss:3.6576 train_time:141317ms step_avg:169.24ms
step:846/1530 train_loss:3.5071 train_time:141492ms step_avg:169.25ms
step:847/1530 train_loss:3.4911 train_time:141668ms step_avg:169.26ms
step:848/1530 train_loss:3.6351 train_time:141839ms step_avg:169.26ms
step:849/1530 train_loss:3.4839 train_time:142013ms step_avg:169.26ms
step:850/1530 train_loss:3.4361 train_time:142187ms step_avg:169.27ms
step:851/1530 train_loss:3.7234 train_time:142360ms step_avg:169.27ms
step:852/1530 train_loss:3.4289 train_time:142535ms step_avg:169.28ms
step:853/1530 train_loss:3.5587 train_time:142708ms step_avg:169.29ms
step:854/1530 train_loss:3.6444 train_time:142882ms step_avg:169.29ms
step:855/1530 train_loss:3.5099 train_time:143055ms step_avg:169.30ms
step:856/1530 train_loss:3.5442 train_time:143230ms step_avg:169.30ms
step:857/1530 train_loss:3.6037 train_time:143403ms step_avg:169.31ms
step:858/1530 train_loss:3.4623 train_time:143579ms step_avg:169.31ms
step:859/1530 train_loss:3.5481 train_time:143753ms step_avg:169.32ms
step:860/1530 train_loss:3.5758 train_time:143925ms step_avg:169.32ms
step:861/1530 train_loss:3.6230 train_time:144102ms step_avg:169.33ms
step:862/1530 train_loss:3.5999 train_time:144279ms step_avg:169.34ms
step:863/1530 train_loss:3.5680 train_time:144455ms step_avg:169.35ms
step:864/1530 train_loss:3.3731 train_time:144630ms step_avg:169.36ms
step:865/1530 train_loss:3.5919 train_time:144802ms step_avg:169.36ms
step:866/1530 train_loss:3.8754 train_time:144977ms step_avg:169.37ms
step:867/1530 train_loss:3.4514 train_time:145150ms step_avg:169.37ms
step:868/1530 train_loss:3.6445 train_time:145323ms step_avg:169.37ms
step:869/1530 train_loss:3.6079 train_time:145496ms step_avg:169.38ms
step:870/1530 train_loss:3.4467 train_time:145670ms step_avg:169.38ms
step:871/1530 train_loss:3.3814 train_time:145845ms step_avg:169.39ms
step:872/1530 train_loss:3.6445 train_time:146023ms step_avg:169.40ms
step:873/1530 train_loss:3.4601 train_time:146195ms step_avg:169.40ms
step:874/1530 train_loss:3.2212 train_time:146373ms step_avg:169.41ms
step:875/1530 train_loss:3.6293 train_time:146546ms step_avg:169.42ms
step:875/1530 val_loss:3.5147 train_time:146596ms step_avg:169.47ms
step:876/1530 train_loss:3.4318 train_time:146719ms step_avg:169.42ms
step:877/1530 train_loss:3.6119 train_time:146894ms step_avg:169.43ms
step:878/1530 train_loss:3.4619 train_time:147068ms step_avg:169.43ms
step:879/1530 train_loss:3.6455 train_time:147241ms step_avg:169.44ms
step:880/1530 train_loss:3.3065 train_time:147414ms step_avg:169.44ms
step:881/1530 train_loss:3.4709 train_time:147587ms step_avg:169.45ms
step:882/1530 train_loss:3.6949 train_time:147759ms step_avg:169.45ms
step:883/1530 train_loss:3.8389 train_time:147933ms step_avg:169.45ms
step:884/1530 train_loss:3.5627 train_time:148110ms step_avg:169.46ms
step:885/1530 train_loss:3.4869 train_time:148284ms step_avg:169.47ms
step:886/1530 train_loss:3.5668 train_time:148456ms step_avg:169.47ms
step:887/1530 train_loss:4.0737 train_time:148631ms step_avg:169.48ms
step:888/1530 train_loss:3.8257 train_time:148811ms step_avg:169.49ms
step:889/1530 train_loss:3.5172 train_time:148985ms step_avg:169.49ms
step:890/1530 train_loss:3.5313 train_time:149156ms step_avg:169.50ms
step:891/1530 train_loss:3.3550 train_time:149332ms step_avg:169.50ms
step:892/1530 train_loss:3.7081 train_time:149506ms step_avg:169.51ms
step:893/1530 train_loss:3.4134 train_time:149678ms step_avg:169.51ms
step:894/1530 train_loss:3.6323 train_time:149855ms step_avg:169.52ms
step:895/1530 train_loss:3.6721 train_time:150031ms step_avg:169.53ms
step:896/1530 train_loss:3.4965 train_time:150206ms step_avg:169.53ms
step:897/1530 train_loss:3.5375 train_time:150380ms step_avg:169.54ms
step:898/1530 train_loss:3.5826 train_time:150556ms step_avg:169.55ms
step:899/1530 train_loss:3.4733 train_time:150730ms step_avg:169.55ms
step:900/1530 train_loss:3.4202 train_time:150903ms step_avg:169.55ms
step:901/1530 train_loss:3.6137 train_time:151075ms step_avg:169.56ms
step:902/1530 train_loss:3.6311 train_time:151249ms step_avg:169.56ms
step:903/1530 train_loss:3.5351 train_time:151424ms step_avg:169.57ms
step:904/1530 train_loss:3.4858 train_time:151597ms step_avg:169.57ms
step:905/1530 train_loss:3.4970 train_time:151769ms step_avg:169.57ms
step:906/1530 train_loss:3.6949 train_time:151945ms step_avg:169.58ms
step:907/1530 train_loss:3.5122 train_time:152119ms step_avg:169.59ms
step:908/1530 train_loss:3.5594 train_time:152292ms step_avg:169.59ms
step:909/1530 train_loss:3.4533 train_time:152468ms step_avg:169.60ms
step:910/1530 train_loss:3.5197 train_time:152648ms step_avg:169.61ms
step:911/1530 train_loss:3.6359 train_time:152822ms step_avg:169.61ms
step:912/1530 train_loss:3.5997 train_time:153000ms step_avg:169.62ms
step:913/1530 train_loss:3.4542 train_time:153177ms step_avg:169.63ms
step:914/1530 train_loss:3.7432 train_time:153356ms step_avg:169.64ms
step:915/1530 train_loss:3.5336 train_time:153537ms step_avg:169.65ms
step:916/1530 train_loss:3.6135 train_time:153714ms step_avg:169.66ms
step:917/1530 train_loss:3.5974 train_time:153889ms step_avg:169.67ms
step:918/1530 train_loss:4.8191 train_time:154070ms step_avg:169.68ms
step:919/1530 train_loss:3.4862 train_time:154249ms step_avg:169.69ms
step:920/1530 train_loss:3.5849 train_time:154425ms step_avg:169.70ms
step:921/1530 train_loss:3.5447 train_time:154601ms step_avg:169.70ms
step:922/1530 train_loss:3.5793 train_time:154777ms step_avg:169.71ms
step:923/1530 train_loss:3.6050 train_time:154953ms step_avg:169.72ms
step:924/1530 train_loss:3.6751 train_time:155132ms step_avg:169.73ms
step:925/1530 train_loss:3.6437 train_time:155307ms step_avg:169.73ms
step:926/1530 train_loss:3.5511 train_time:155480ms step_avg:169.74ms
step:927/1530 train_loss:3.5465 train_time:155656ms step_avg:169.75ms
step:928/1530 train_loss:3.7877 train_time:155834ms step_avg:169.75ms
step:929/1530 train_loss:3.6073 train_time:156011ms step_avg:169.76ms
step:930/1530 train_loss:3.3998 train_time:156186ms step_avg:169.77ms
step:931/1530 train_loss:3.4938 train_time:156361ms step_avg:169.77ms
step:932/1530 train_loss:3.6453 train_time:156538ms step_avg:169.78ms
step:933/1530 train_loss:3.3582 train_time:156715ms step_avg:169.79ms
step:934/1530 train_loss:3.5807 train_time:156894ms step_avg:169.80ms
step:935/1530 train_loss:3.4399 train_time:157071ms step_avg:169.81ms
step:936/1530 train_loss:3.5133 train_time:157249ms step_avg:169.82ms
step:937/1530 train_loss:3.6121 train_time:157429ms step_avg:169.83ms
step:938/1530 train_loss:3.5395 train_time:157602ms step_avg:169.83ms
step:939/1530 train_loss:3.6657 train_time:157781ms step_avg:169.84ms
step:940/1530 train_loss:3.4749 train_time:157956ms step_avg:169.84ms
step:941/1530 train_loss:3.5405 train_time:158131ms step_avg:169.85ms
step:942/1530 train_loss:3.3494 train_time:158309ms step_avg:169.86ms
step:943/1530 train_loss:3.7089 train_time:158489ms step_avg:169.87ms
step:944/1530 train_loss:3.4030 train_time:158810ms step_avg:170.03ms
step:945/1530 train_loss:3.4192 train_time:158993ms step_avg:170.05ms
step:946/1530 train_loss:5.0815 train_time:159174ms step_avg:170.06ms
step:947/1530 train_loss:3.5943 train_time:159351ms step_avg:170.07ms
step:948/1530 train_loss:3.4817 train_time:159527ms step_avg:170.07ms
step:949/1530 train_loss:3.3675 train_time:159856ms step_avg:170.24ms
step:950/1530 train_loss:3.4377 train_time:160031ms step_avg:170.25ms
step:951/1530 train_loss:3.4040 train_time:160211ms step_avg:170.26ms
step:952/1530 train_loss:3.4757 train_time:160386ms step_avg:170.26ms
step:953/1530 train_loss:3.5606 train_time:160563ms step_avg:170.27ms
step:954/1530 train_loss:3.4397 train_time:160741ms step_avg:170.28ms
step:955/1530 train_loss:3.4727 train_time:160915ms step_avg:170.28ms
step:956/1530 train_loss:3.4417 train_time:161091ms step_avg:170.29ms
step:957/1530 train_loss:3.4908 train_time:161271ms step_avg:170.30ms
step:958/1530 train_loss:3.4973 train_time:161450ms step_avg:170.31ms
step:959/1530 train_loss:3.5046 train_time:161625ms step_avg:170.31ms
step:960/1530 train_loss:3.4024 train_time:161802ms step_avg:170.32ms
step:961/1530 train_loss:3.6412 train_time:161977ms step_avg:170.32ms
step:962/1530 train_loss:3.5867 train_time:162152ms step_avg:170.33ms
step:963/1530 train_loss:3.6226 train_time:162331ms step_avg:170.34ms
step:964/1530 train_loss:3.4267 train_time:162510ms step_avg:170.35ms
step:965/1530 train_loss:3.4705 train_time:162684ms step_avg:170.35ms
step:966/1530 train_loss:3.7072 train_time:162859ms step_avg:170.35ms
step:967/1530 train_loss:3.5192 train_time:163034ms step_avg:170.36ms
step:968/1530 train_loss:3.5087 train_time:163212ms step_avg:170.37ms
step:969/1530 train_loss:3.5808 train_time:163385ms step_avg:170.37ms
step:970/1530 train_loss:3.3659 train_time:163558ms step_avg:170.37ms
step:971/1530 train_loss:3.5302 train_time:163733ms step_avg:170.38ms
step:972/1530 train_loss:3.4709 train_time:163908ms step_avg:170.38ms
step:973/1530 train_loss:3.5365 train_time:164082ms step_avg:170.39ms
step:974/1530 train_loss:3.5834 train_time:164259ms step_avg:170.39ms
step:975/1530 train_loss:3.4583 train_time:164434ms step_avg:170.40ms
step:976/1530 train_loss:3.6632 train_time:164609ms step_avg:170.40ms
step:977/1530 train_loss:3.5665 train_time:164783ms step_avg:170.41ms
step:978/1530 train_loss:3.3483 train_time:164957ms step_avg:170.41ms
step:979/1530 train_loss:3.6209 train_time:165133ms step_avg:170.42ms
step:980/1530 train_loss:3.4123 train_time:165311ms step_avg:170.42ms
step:981/1530 train_loss:3.5655 train_time:165488ms step_avg:170.43ms
step:982/1530 train_loss:3.5376 train_time:165660ms step_avg:170.43ms
step:983/1530 train_loss:3.5051 train_time:165837ms step_avg:170.44ms
step:984/1530 train_loss:3.4939 train_time:166012ms step_avg:170.44ms
step:985/1530 train_loss:3.5618 train_time:166190ms step_avg:170.45ms
step:986/1530 train_loss:3.4108 train_time:166366ms step_avg:170.46ms
step:987/1530 train_loss:3.4812 train_time:166538ms step_avg:170.46ms
step:988/1530 train_loss:3.4916 train_time:166714ms step_avg:170.46ms
step:989/1530 train_loss:3.4124 train_time:166887ms step_avg:170.47ms
step:990/1530 train_loss:3.6525 train_time:167063ms step_avg:170.47ms
step:991/1530 train_loss:3.4638 train_time:167238ms step_avg:170.48ms
step:992/1530 train_loss:3.4408 train_time:167419ms step_avg:170.49ms
step:993/1530 train_loss:3.4971 train_time:167599ms step_avg:170.50ms
step:994/1530 train_loss:3.5950 train_time:167773ms step_avg:170.50ms
step:995/1530 train_loss:3.5235 train_time:167945ms step_avg:170.50ms
step:996/1530 train_loss:3.4557 train_time:168118ms step_avg:170.50ms
step:997/1530 train_loss:3.7527 train_time:168292ms step_avg:170.51ms
step:998/1530 train_loss:3.4312 train_time:168465ms step_avg:170.51ms
step:999/1530 train_loss:3.5845 train_time:168638ms step_avg:170.51ms
step:1000/1530 train_loss:3.4388 train_time:168815ms step_avg:170.52ms
step:1000/1530 val_loss:3.4617 train_time:168866ms step_avg:170.57ms
step:1001/1530 train_loss:3.4942 train_time:168991ms step_avg:170.53ms
step:1002/1530 train_loss:3.3720 train_time:169166ms step_avg:170.53ms
step:1003/1530 train_loss:3.5525 train_time:169342ms step_avg:170.54ms
step:1004/1530 train_loss:3.5987 train_time:169518ms step_avg:170.54ms
step:1005/1530 train_loss:3.3886 train_time:169693ms step_avg:170.55ms
step:1006/1530 train_loss:3.4599 train_time:169871ms step_avg:170.55ms
step:1007/1530 train_loss:3.4291 train_time:170050ms step_avg:170.56ms
step:1008/1530 train_loss:3.5525 train_time:170227ms step_avg:170.57ms
step:1009/1530 train_loss:3.6595 train_time:170406ms step_avg:170.58ms
step:1010/1530 train_loss:3.5612 train_time:170579ms step_avg:170.58ms
step:1011/1530 train_loss:3.5325 train_time:170752ms step_avg:170.58ms
step:1012/1530 train_loss:3.3840 train_time:170928ms step_avg:170.59ms
step:1013/1530 train_loss:3.5297 train_time:171101ms step_avg:170.59ms
step:1014/1530 train_loss:3.6150 train_time:171277ms step_avg:170.59ms
step:1015/1530 train_loss:3.3236 train_time:171454ms step_avg:170.60ms
step:1016/1530 train_loss:3.4018 train_time:171630ms step_avg:170.61ms
step:1017/1530 train_loss:3.3832 train_time:171805ms step_avg:170.61ms
step:1018/1530 train_loss:3.3896 train_time:171979ms step_avg:170.61ms
step:1019/1530 train_loss:3.5168 train_time:172155ms step_avg:170.62ms
step:1020/1530 train_loss:3.3711 train_time:172334ms step_avg:170.63ms
step:1021/1530 train_loss:3.3422 train_time:172510ms step_avg:170.63ms
step:1022/1530 train_loss:3.4760 train_time:172687ms step_avg:170.64ms
step:1023/1530 train_loss:3.4970 train_time:172861ms step_avg:170.64ms
step:1024/1530 train_loss:3.4722 train_time:173037ms step_avg:170.65ms
step:1025/1530 train_loss:3.4686 train_time:173217ms step_avg:170.66ms
step:1026/1530 train_loss:3.6143 train_time:173392ms step_avg:170.66ms
step:1027/1530 train_loss:3.3159 train_time:173569ms step_avg:170.67ms
step:1028/1530 train_loss:3.3901 train_time:173751ms step_avg:170.68ms
step:1029/1530 train_loss:3.3063 train_time:173931ms step_avg:170.69ms
step:1030/1530 train_loss:3.5360 train_time:174108ms step_avg:170.69ms
step:1031/1530 train_loss:3.5009 train_time:174283ms step_avg:170.70ms
step:1032/1530 train_loss:3.6863 train_time:174464ms step_avg:170.71ms
step:1033/1530 train_loss:3.4850 train_time:174640ms step_avg:170.71ms
step:1034/1530 train_loss:3.3893 train_time:174818ms step_avg:170.72ms
step:1035/1530 train_loss:3.4382 train_time:174996ms step_avg:170.73ms
step:1036/1530 train_loss:3.4778 train_time:175173ms step_avg:170.73ms
step:1037/1530 train_loss:3.7813 train_time:175352ms step_avg:170.74ms
step:1038/1530 train_loss:3.6059 train_time:175533ms step_avg:170.75ms
step:1039/1530 train_loss:3.5080 train_time:175715ms step_avg:170.76ms
step:1040/1530 train_loss:3.4115 train_time:175892ms step_avg:170.77ms
step:1041/1530 train_loss:3.4811 train_time:176070ms step_avg:170.78ms
step:1042/1530 train_loss:3.5217 train_time:176244ms step_avg:170.78ms
step:1043/1530 train_loss:3.4420 train_time:176419ms step_avg:170.78ms
step:1044/1530 train_loss:3.4540 train_time:176597ms step_avg:170.79ms
step:1045/1530 train_loss:3.5134 train_time:176775ms step_avg:170.80ms
step:1046/1530 train_loss:3.4194 train_time:176950ms step_avg:170.80ms
step:1047/1530 train_loss:3.6265 train_time:177127ms step_avg:170.81ms
step:1048/1530 train_loss:3.4950 train_time:177303ms step_avg:170.81ms
step:1049/1530 train_loss:3.3942 train_time:177478ms step_avg:170.82ms
step:1050/1530 train_loss:3.3889 train_time:177656ms step_avg:170.82ms
step:1051/1530 train_loss:3.4904 train_time:177834ms step_avg:170.83ms
step:1052/1530 train_loss:3.3498 train_time:178014ms step_avg:170.84ms
step:1053/1530 train_loss:3.6875 train_time:178192ms step_avg:170.85ms
step:1054/1530 train_loss:3.5317 train_time:178372ms step_avg:170.85ms
step:1055/1530 train_loss:3.3808 train_time:178547ms step_avg:170.86ms
step:1056/1530 train_loss:3.4929 train_time:178721ms step_avg:170.86ms
step:1057/1530 train_loss:3.5724 train_time:178897ms step_avg:170.87ms
step:1058/1530 train_loss:3.3019 train_time:179075ms step_avg:170.87ms
step:1059/1530 train_loss:3.3643 train_time:179257ms step_avg:170.88ms
step:1060/1530 train_loss:3.4334 train_time:179434ms step_avg:170.89ms
step:1061/1530 train_loss:3.4065 train_time:179609ms step_avg:170.89ms
step:1062/1530 train_loss:3.3805 train_time:179784ms step_avg:170.90ms
step:1063/1530 train_loss:3.4517 train_time:179958ms step_avg:170.90ms
step:1064/1530 train_loss:3.3778 train_time:180133ms step_avg:170.90ms
step:1065/1530 train_loss:3.3520 train_time:180313ms step_avg:170.91ms
step:1066/1530 train_loss:3.4124 train_time:180489ms step_avg:170.92ms
step:1067/1530 train_loss:3.2741 train_time:180668ms step_avg:170.93ms
step:1068/1530 train_loss:3.4303 train_time:180844ms step_avg:170.93ms
step:1069/1530 train_loss:3.2891 train_time:181026ms step_avg:170.94ms
step:1070/1530 train_loss:3.5650 train_time:181200ms step_avg:170.94ms
step:1071/1530 train_loss:3.5069 train_time:181382ms step_avg:170.95ms
step:1072/1530 train_loss:3.4319 train_time:181557ms step_avg:170.96ms
step:1073/1530 train_loss:3.5175 train_time:181732ms step_avg:170.96ms
step:1074/1530 train_loss:3.4275 train_time:181908ms step_avg:170.97ms
step:1075/1530 train_loss:3.3908 train_time:182085ms step_avg:170.97ms
step:1076/1530 train_loss:3.7907 train_time:182261ms step_avg:170.98ms
step:1077/1530 train_loss:3.4202 train_time:182435ms step_avg:170.98ms
step:1078/1530 train_loss:3.0807 train_time:182621ms step_avg:170.99ms
step:1079/1530 train_loss:3.5277 train_time:182798ms step_avg:171.00ms
step:1080/1530 train_loss:3.4181 train_time:182975ms step_avg:171.00ms
step:1081/1530 train_loss:3.4957 train_time:183150ms step_avg:171.01ms
step:1082/1530 train_loss:3.5884 train_time:183328ms step_avg:171.02ms
step:1083/1530 train_loss:3.4885 train_time:183503ms step_avg:171.02ms
step:1084/1530 train_loss:3.4593 train_time:183677ms step_avg:171.02ms
step:1085/1530 train_loss:3.4294 train_time:183853ms step_avg:171.03ms
step:1086/1530 train_loss:3.6213 train_time:184031ms step_avg:171.03ms
step:1087/1530 train_loss:3.4957 train_time:184205ms step_avg:171.04ms
step:1088/1530 train_loss:3.3621 train_time:184381ms step_avg:171.04ms
step:1089/1530 train_loss:3.3674 train_time:184560ms step_avg:171.05ms
step:1090/1530 train_loss:3.4779 train_time:184739ms step_avg:171.05ms
step:1091/1530 train_loss:3.2791 train_time:184917ms step_avg:171.06ms
step:1092/1530 train_loss:3.4830 train_time:185094ms step_avg:171.07ms
step:1093/1530 train_loss:3.5986 train_time:185271ms step_avg:171.07ms
step:1094/1530 train_loss:3.4406 train_time:185446ms step_avg:171.08ms
step:1095/1530 train_loss:3.4150 train_time:185622ms step_avg:171.08ms
step:1096/1530 train_loss:3.4242 train_time:185798ms step_avg:171.09ms
step:1097/1530 train_loss:3.4885 train_time:185976ms step_avg:171.09ms
step:1098/1530 train_loss:3.5573 train_time:186155ms step_avg:171.10ms
step:1099/1530 train_loss:3.5237 train_time:186331ms step_avg:171.10ms
step:1100/1530 train_loss:3.4204 train_time:186509ms step_avg:171.11ms
step:1101/1530 train_loss:3.2825 train_time:186685ms step_avg:171.11ms
step:1102/1530 train_loss:3.3003 train_time:186863ms step_avg:171.12ms
step:1103/1530 train_loss:3.4392 train_time:187045ms step_avg:171.13ms
step:1104/1530 train_loss:3.3126 train_time:187219ms step_avg:171.13ms
step:1105/1530 train_loss:4.0558 train_time:187398ms step_avg:171.14ms
step:1106/1530 train_loss:3.2211 train_time:187573ms step_avg:171.14ms
step:1107/1530 train_loss:3.5653 train_time:187749ms step_avg:171.15ms
step:1108/1530 train_loss:3.3424 train_time:187923ms step_avg:171.15ms
step:1109/1530 train_loss:3.4974 train_time:188097ms step_avg:171.15ms
step:1110/1530 train_loss:3.4227 train_time:188272ms step_avg:171.16ms
step:1111/1530 train_loss:3.4799 train_time:188448ms step_avg:171.16ms
step:1112/1530 train_loss:3.5570 train_time:188627ms step_avg:171.17ms
step:1113/1530 train_loss:3.4243 train_time:188810ms step_avg:171.18ms
step:1114/1530 train_loss:3.3685 train_time:188990ms step_avg:171.19ms
step:1115/1530 train_loss:3.2344 train_time:189169ms step_avg:171.19ms
step:1116/1530 train_loss:3.4229 train_time:189342ms step_avg:171.20ms
step:1117/1530 train_loss:3.5877 train_time:189521ms step_avg:171.20ms
step:1118/1530 train_loss:3.6166 train_time:189698ms step_avg:171.21ms
step:1119/1530 train_loss:3.4761 train_time:189874ms step_avg:171.21ms
step:1120/1530 train_loss:3.4864 train_time:190051ms step_avg:171.22ms
step:1121/1530 train_loss:3.3851 train_time:190231ms step_avg:171.23ms
step:1122/1530 train_loss:3.4555 train_time:190406ms step_avg:171.23ms
step:1123/1530 train_loss:3.5734 train_time:190580ms step_avg:171.23ms
step:1124/1530 train_loss:3.3361 train_time:190756ms step_avg:171.24ms
step:1125/1530 train_loss:3.2155 train_time:190934ms step_avg:171.24ms
step:1125/1530 val_loss:3.4040 train_time:190984ms step_avg:171.29ms
step:1126/1530 train_loss:3.4717 train_time:191110ms step_avg:171.25ms
step:1127/1530 train_loss:3.6691 train_time:191288ms step_avg:171.25ms
step:1128/1530 train_loss:3.2236 train_time:191467ms step_avg:171.26ms
step:1129/1530 train_loss:3.5522 train_time:191649ms step_avg:171.27ms
step:1130/1530 train_loss:3.3741 train_time:191828ms step_avg:171.27ms
step:1131/1530 train_loss:3.3938 train_time:192010ms step_avg:171.28ms
step:1132/1530 train_loss:3.3653 train_time:192183ms step_avg:171.29ms
step:1133/1530 train_loss:3.4848 train_time:192494ms step_avg:171.41ms
step:1134/1530 train_loss:3.4426 train_time:192680ms step_avg:171.42ms
step:1135/1530 train_loss:3.5138 train_time:192856ms step_avg:171.43ms
step:1136/1530 train_loss:3.5613 train_time:193033ms step_avg:171.43ms
step:1137/1530 train_loss:3.4531 train_time:193210ms step_avg:171.44ms
step:1138/1530 train_loss:3.3480 train_time:193389ms step_avg:171.44ms
step:1139/1530 train_loss:3.6469 train_time:193717ms step_avg:171.58ms
step:1140/1530 train_loss:3.4518 train_time:193892ms step_avg:171.59ms
step:1141/1530 train_loss:3.5910 train_time:194073ms step_avg:171.59ms
step:1142/1530 train_loss:3.4372 train_time:194249ms step_avg:171.60ms
step:1143/1530 train_loss:3.3569 train_time:194430ms step_avg:171.61ms
step:1144/1530 train_loss:3.4348 train_time:194609ms step_avg:171.61ms
step:1145/1530 train_loss:3.5850 train_time:194783ms step_avg:171.62ms
step:1146/1530 train_loss:3.5541 train_time:194964ms step_avg:171.62ms
step:1147/1530 train_loss:3.4856 train_time:195142ms step_avg:171.63ms
step:1148/1530 train_loss:3.4961 train_time:195320ms step_avg:171.63ms
step:1149/1530 train_loss:3.3177 train_time:195500ms step_avg:171.64ms
step:1150/1530 train_loss:3.3731 train_time:195676ms step_avg:171.65ms
step:1151/1530 train_loss:3.3178 train_time:195854ms step_avg:171.65ms
step:1152/1530 train_loss:3.3920 train_time:196035ms step_avg:171.66ms
step:1153/1530 train_loss:3.4247 train_time:196216ms step_avg:171.67ms
step:1154/1530 train_loss:3.5126 train_time:196392ms step_avg:171.67ms
step:1155/1530 train_loss:3.3094 train_time:196572ms step_avg:171.68ms
step:1156/1530 train_loss:3.5329 train_time:196757ms step_avg:171.69ms
step:1157/1530 train_loss:3.4905 train_time:196934ms step_avg:171.69ms
step:1158/1530 train_loss:3.2437 train_time:197111ms step_avg:171.70ms
step:1159/1530 train_loss:3.3430 train_time:197287ms step_avg:171.70ms
step:1160/1530 train_loss:3.3324 train_time:197462ms step_avg:171.71ms
step:1161/1530 train_loss:3.0710 train_time:197641ms step_avg:171.71ms
step:1162/1530 train_loss:3.4157 train_time:197820ms step_avg:171.72ms
step:1163/1530 train_loss:3.3842 train_time:197998ms step_avg:171.72ms
step:1164/1530 train_loss:3.2887 train_time:198175ms step_avg:171.73ms
step:1165/1530 train_loss:3.2411 train_time:198351ms step_avg:171.73ms
step:1166/1530 train_loss:3.3857 train_time:198530ms step_avg:171.74ms
step:1167/1530 train_loss:3.4066 train_time:198707ms step_avg:171.74ms
step:1168/1530 train_loss:3.7233 train_time:198883ms step_avg:171.75ms
step:1169/1530 train_loss:3.3729 train_time:199062ms step_avg:171.75ms
step:1170/1530 train_loss:3.3848 train_time:199238ms step_avg:171.76ms
step:1171/1530 train_loss:3.2957 train_time:199414ms step_avg:171.76ms
step:1172/1530 train_loss:3.4214 train_time:199589ms step_avg:171.76ms
step:1173/1530 train_loss:3.5318 train_time:199771ms step_avg:171.77ms
step:1174/1530 train_loss:3.3746 train_time:199956ms step_avg:171.78ms
step:1175/1530 train_loss:3.3587 train_time:200136ms step_avg:171.79ms
step:1176/1530 train_loss:3.4201 train_time:200318ms step_avg:171.80ms
step:1177/1530 train_loss:3.4486 train_time:200501ms step_avg:171.81ms
step:1178/1530 train_loss:3.4922 train_time:200677ms step_avg:171.81ms
step:1179/1530 train_loss:3.3934 train_time:200852ms step_avg:171.82ms
step:1180/1530 train_loss:3.3518 train_time:201040ms step_avg:171.83ms
step:1181/1530 train_loss:3.3307 train_time:201218ms step_avg:171.83ms
step:1182/1530 train_loss:3.3703 train_time:201394ms step_avg:171.84ms
step:1183/1530 train_loss:3.3299 train_time:201572ms step_avg:171.84ms
step:1184/1530 train_loss:3.5099 train_time:201749ms step_avg:171.85ms
step:1185/1530 train_loss:3.5377 train_time:201931ms step_avg:171.86ms
step:1186/1530 train_loss:3.3610 train_time:202111ms step_avg:171.86ms
step:1187/1530 train_loss:3.4163 train_time:202297ms step_avg:171.88ms
step:1188/1530 train_loss:3.4392 train_time:202473ms step_avg:171.88ms
step:1189/1530 train_loss:3.2710 train_time:202653ms step_avg:171.89ms
step:1190/1530 train_loss:3.4381 train_time:202832ms step_avg:171.89ms
step:1191/1530 train_loss:3.5745 train_time:203013ms step_avg:171.90ms
step:1192/1530 train_loss:3.3881 train_time:203188ms step_avg:171.90ms
step:1193/1530 train_loss:3.2696 train_time:203364ms step_avg:171.90ms
step:1194/1530 train_loss:3.5498 train_time:203542ms step_avg:171.91ms
step:1195/1530 train_loss:3.3698 train_time:203724ms step_avg:171.92ms
step:1196/1530 train_loss:3.3800 train_time:203910ms step_avg:171.93ms
step:1197/1530 train_loss:3.2909 train_time:204090ms step_avg:171.94ms
step:1198/1530 train_loss:3.2971 train_time:204275ms step_avg:171.95ms
step:1199/1530 train_loss:3.3346 train_time:204454ms step_avg:171.95ms
step:1200/1530 train_loss:3.4433 train_time:204632ms step_avg:171.96ms
step:1201/1530 train_loss:3.4726 train_time:204810ms step_avg:171.96ms
step:1202/1530 train_loss:3.6890 train_time:205001ms step_avg:171.98ms
step:1203/1530 train_loss:3.4050 train_time:205181ms step_avg:171.99ms
step:1204/1530 train_loss:3.3057 train_time:205362ms step_avg:172.00ms
step:1205/1530 train_loss:3.4350 train_time:205538ms step_avg:172.00ms
step:1206/1530 train_loss:3.4665 train_time:205714ms step_avg:172.00ms
step:1207/1530 train_loss:3.5053 train_time:205892ms step_avg:172.01ms
step:1208/1530 train_loss:3.3902 train_time:206067ms step_avg:172.01ms
step:1209/1530 train_loss:3.2403 train_time:206249ms step_avg:172.02ms
step:1210/1530 train_loss:3.3012 train_time:206429ms step_avg:172.02ms
step:1211/1530 train_loss:3.3915 train_time:206607ms step_avg:172.03ms
step:1212/1530 train_loss:3.3939 train_time:206785ms step_avg:172.03ms
step:1213/1530 train_loss:3.4085 train_time:206965ms step_avg:172.04ms
step:1214/1530 train_loss:3.2467 train_time:207144ms step_avg:172.05ms
step:1215/1530 train_loss:3.3898 train_time:207322ms step_avg:172.05ms
step:1216/1530 train_loss:3.3245 train_time:207501ms step_avg:172.06ms
step:1217/1530 train_loss:3.3145 train_time:207677ms step_avg:172.06ms
step:1218/1530 train_loss:3.4059 train_time:207855ms step_avg:172.07ms
step:1219/1530 train_loss:3.2464 train_time:208037ms step_avg:172.07ms
step:1220/1530 train_loss:3.4762 train_time:208214ms step_avg:172.08ms
step:1221/1530 train_loss:3.5010 train_time:208390ms step_avg:172.08ms
step:1222/1530 train_loss:3.4302 train_time:208564ms step_avg:172.08ms
step:1223/1530 train_loss:3.2917 train_time:208743ms step_avg:172.09ms
step:1224/1530 train_loss:3.2514 train_time:208927ms step_avg:172.10ms
step:1225/1530 train_loss:3.3644 train_time:209104ms step_avg:172.10ms
step:1226/1530 train_loss:3.3299 train_time:209284ms step_avg:172.11ms
step:1227/1530 train_loss:3.2691 train_time:209464ms step_avg:172.11ms
step:1228/1530 train_loss:3.4436 train_time:209639ms step_avg:172.12ms
step:1229/1530 train_loss:3.3667 train_time:209819ms step_avg:172.12ms
step:1230/1530 train_loss:3.3955 train_time:210000ms step_avg:172.13ms
step:1231/1530 train_loss:3.5729 train_time:210179ms step_avg:172.14ms
step:1232/1530 train_loss:3.4910 train_time:210361ms step_avg:172.14ms
step:1233/1530 train_loss:3.4266 train_time:210535ms step_avg:172.15ms
step:1234/1530 train_loss:3.5858 train_time:210712ms step_avg:172.15ms
step:1235/1530 train_loss:3.3153 train_time:210891ms step_avg:172.16ms
step:1236/1530 train_loss:3.2865 train_time:211068ms step_avg:172.16ms
step:1237/1530 train_loss:3.2720 train_time:211247ms step_avg:172.17ms
step:1238/1530 train_loss:3.2712 train_time:211430ms step_avg:172.17ms
step:1239/1530 train_loss:3.3237 train_time:211608ms step_avg:172.18ms
step:1240/1530 train_loss:3.3771 train_time:211785ms step_avg:172.18ms
step:1241/1530 train_loss:3.4205 train_time:211965ms step_avg:172.19ms
step:1242/1530 train_loss:3.2916 train_time:212142ms step_avg:172.19ms
step:1243/1530 train_loss:3.4055 train_time:212321ms step_avg:172.20ms
step:1244/1530 train_loss:3.4030 train_time:212494ms step_avg:172.20ms
step:1245/1530 train_loss:3.4061 train_time:212671ms step_avg:172.20ms
step:1246/1530 train_loss:3.2348 train_time:212851ms step_avg:172.21ms
step:1247/1530 train_loss:3.3652 train_time:213028ms step_avg:172.21ms
step:1248/1530 train_loss:3.4230 train_time:213205ms step_avg:172.22ms
step:1249/1530 train_loss:3.4200 train_time:213384ms step_avg:172.22ms
step:1250/1530 train_loss:3.2999 train_time:213562ms step_avg:172.23ms
step:1250/1530 val_loss:3.3516 train_time:213617ms step_avg:172.27ms
step:1251/1530 train_loss:3.4889 train_time:213749ms step_avg:172.24ms
step:1252/1530 train_loss:3.3588 train_time:213925ms step_avg:172.24ms
step:1253/1530 train_loss:3.3067 train_time:214104ms step_avg:172.25ms
step:1254/1530 train_loss:3.4072 train_time:214284ms step_avg:172.25ms
step:1255/1530 train_loss:3.5143 train_time:214472ms step_avg:172.27ms
step:1256/1530 train_loss:3.3002 train_time:214653ms step_avg:172.27ms
step:1257/1530 train_loss:3.3691 train_time:214831ms step_avg:172.28ms
step:1258/1530 train_loss:3.3615 train_time:215016ms step_avg:172.29ms
step:1259/1530 train_loss:3.3249 train_time:215194ms step_avg:172.29ms
step:1260/1530 train_loss:3.2057 train_time:215371ms step_avg:172.30ms
step:1261/1530 train_loss:3.3023 train_time:215551ms step_avg:172.30ms
step:1262/1530 train_loss:3.3218 train_time:215735ms step_avg:172.31ms
step:1263/1530 train_loss:3.2387 train_time:215917ms step_avg:172.32ms
step:1264/1530 train_loss:3.4405 train_time:216092ms step_avg:172.32ms
step:1265/1530 train_loss:3.4193 train_time:216268ms step_avg:172.33ms
step:1266/1530 train_loss:3.4408 train_time:216448ms step_avg:172.33ms
step:1267/1530 train_loss:3.3661 train_time:216628ms step_avg:172.34ms
step:1268/1530 train_loss:3.4017 train_time:216809ms step_avg:172.34ms
step:1269/1530 train_loss:3.2512 train_time:216993ms step_avg:172.35ms
step:1270/1530 train_loss:3.0997 train_time:217171ms step_avg:172.36ms
step:1271/1530 train_loss:3.4009 train_time:217350ms step_avg:172.36ms
step:1272/1530 train_loss:3.3465 train_time:217527ms step_avg:172.37ms
step:1273/1530 train_loss:3.3758 train_time:217710ms step_avg:172.38ms
step:1274/1530 train_loss:3.3595 train_time:217892ms step_avg:172.38ms
step:1275/1530 train_loss:3.4265 train_time:218069ms step_avg:172.39ms
step:1276/1530 train_loss:3.4645 train_time:218244ms step_avg:172.39ms
step:1277/1530 train_loss:3.4084 train_time:218422ms step_avg:172.39ms
step:1278/1530 train_loss:3.4019 train_time:218598ms step_avg:172.40ms
step:1279/1530 train_loss:3.2638 train_time:218778ms step_avg:172.40ms
step:1280/1530 train_loss:3.3582 train_time:218964ms step_avg:172.41ms
step:1281/1530 train_loss:3.4205 train_time:219140ms step_avg:172.42ms
step:1282/1530 train_loss:3.4649 train_time:219314ms step_avg:172.42ms
step:1283/1530 train_loss:3.3269 train_time:219495ms step_avg:172.42ms
step:1284/1530 train_loss:3.3634 train_time:219674ms step_avg:172.43ms
step:1285/1530 train_loss:3.3558 train_time:219853ms step_avg:172.43ms
step:1286/1530 train_loss:3.3336 train_time:220030ms step_avg:172.44ms
step:1287/1530 train_loss:3.4849 train_time:220209ms step_avg:172.44ms
step:1288/1530 train_loss:3.2923 train_time:220390ms step_avg:172.45ms
step:1289/1530 train_loss:3.3794 train_time:220575ms step_avg:172.46ms
step:1290/1530 train_loss:3.4572 train_time:220759ms step_avg:172.47ms
step:1291/1530 train_loss:3.3756 train_time:220937ms step_avg:172.47ms
step:1292/1530 train_loss:3.4746 train_time:221120ms step_avg:172.48ms
step:1293/1530 train_loss:3.5139 train_time:221301ms step_avg:172.49ms
step:1294/1530 train_loss:3.4533 train_time:221483ms step_avg:172.49ms
step:1295/1530 train_loss:3.2786 train_time:221661ms step_avg:172.50ms
step:1296/1530 train_loss:3.3704 train_time:221843ms step_avg:172.51ms
step:1297/1530 train_loss:3.2735 train_time:222024ms step_avg:172.51ms
step:1298/1530 train_loss:3.2669 train_time:222207ms step_avg:172.52ms
step:1299/1530 train_loss:3.3942 train_time:222386ms step_avg:172.53ms
step:1300/1530 train_loss:3.4012 train_time:222561ms step_avg:172.53ms
step:1301/1530 train_loss:3.4008 train_time:222737ms step_avg:172.53ms
step:1302/1530 train_loss:3.5743 train_time:222919ms step_avg:172.54ms
step:1303/1530 train_loss:3.3049 train_time:223099ms step_avg:172.54ms
step:1304/1530 train_loss:3.5027 train_time:223280ms step_avg:172.55ms
step:1305/1530 train_loss:3.2572 train_time:223457ms step_avg:172.55ms
step:1306/1530 train_loss:3.4475 train_time:223638ms step_avg:172.56ms
step:1307/1530 train_loss:3.4536 train_time:223813ms step_avg:172.56ms
step:1308/1530 train_loss:3.2774 train_time:223993ms step_avg:172.57ms
step:1309/1530 train_loss:3.3056 train_time:224173ms step_avg:172.57ms
step:1310/1530 train_loss:3.2799 train_time:224350ms step_avg:172.58ms
step:1311/1530 train_loss:3.2912 train_time:224530ms step_avg:172.58ms
step:1312/1530 train_loss:3.3725 train_time:224710ms step_avg:172.59ms
step:1313/1530 train_loss:3.3379 train_time:224888ms step_avg:172.59ms
step:1314/1530 train_loss:3.0384 train_time:225072ms step_avg:172.60ms
step:1315/1530 train_loss:3.2716 train_time:225249ms step_avg:172.60ms
step:1316/1530 train_loss:3.3925 train_time:225425ms step_avg:172.61ms
step:1317/1530 train_loss:3.4176 train_time:225602ms step_avg:172.61ms
step:1318/1530 train_loss:3.3011 train_time:225789ms step_avg:172.62ms
step:1319/1530 train_loss:3.4203 train_time:225969ms step_avg:172.63ms
step:1320/1530 train_loss:3.4592 train_time:226150ms step_avg:172.63ms
step:1321/1530 train_loss:3.3623 train_time:226330ms step_avg:172.64ms
step:1322/1530 train_loss:3.3213 train_time:226641ms step_avg:172.74ms
step:1323/1530 train_loss:3.3161 train_time:226829ms step_avg:172.76ms
step:1324/1530 train_loss:3.4288 train_time:227012ms step_avg:172.76ms
step:1325/1530 train_loss:3.4875 train_time:227197ms step_avg:172.77ms
step:1326/1530 train_loss:3.2058 train_time:227376ms step_avg:172.78ms
step:1327/1530 train_loss:3.1647 train_time:227553ms step_avg:172.78ms
step:1328/1530 train_loss:3.4931 train_time:227733ms step_avg:172.79ms
step:1329/1530 train_loss:3.2928 train_time:228078ms step_avg:172.92ms
step:1330/1530 train_loss:3.4215 train_time:228259ms step_avg:172.92ms
step:1331/1530 train_loss:3.3252 train_time:228434ms step_avg:172.93ms
step:1332/1530 train_loss:3.7390 train_time:228617ms step_avg:172.93ms
step:1333/1530 train_loss:3.4812 train_time:228798ms step_avg:172.94ms
step:1334/1530 train_loss:3.3659 train_time:228974ms step_avg:172.94ms
step:1335/1530 train_loss:3.2868 train_time:229153ms step_avg:172.95ms
step:1336/1530 train_loss:3.2955 train_time:229339ms step_avg:172.96ms
step:1337/1530 train_loss:3.5460 train_time:229518ms step_avg:172.96ms
step:1338/1530 train_loss:3.5210 train_time:229697ms step_avg:172.96ms
step:1339/1530 train_loss:3.3371 train_time:229877ms step_avg:172.97ms
step:1340/1530 train_loss:3.2816 train_time:230053ms step_avg:172.97ms
step:1341/1530 train_loss:3.5909 train_time:230231ms step_avg:172.98ms
step:1342/1530 train_loss:3.3530 train_time:230412ms step_avg:172.98ms
step:1343/1530 train_loss:3.3606 train_time:230592ms step_avg:172.99ms
step:1344/1530 train_loss:3.4121 train_time:230772ms step_avg:172.99ms
step:1345/1530 train_loss:3.3779 train_time:230954ms step_avg:173.00ms
step:1346/1530 train_loss:3.2957 train_time:231130ms step_avg:173.00ms
step:1347/1530 train_loss:3.2739 train_time:231309ms step_avg:173.01ms
step:1348/1530 train_loss:3.3450 train_time:231488ms step_avg:173.01ms
step:1349/1530 train_loss:3.2719 train_time:231664ms step_avg:173.01ms
step:1350/1530 train_loss:3.3901 train_time:231845ms step_avg:173.02ms
step:1351/1530 train_loss:3.2407 train_time:232021ms step_avg:173.02ms
step:1352/1530 train_loss:3.3038 train_time:232197ms step_avg:173.02ms
step:1353/1530 train_loss:3.3997 train_time:232378ms step_avg:173.03ms
step:1354/1530 train_loss:3.2584 train_time:232554ms step_avg:173.03ms
step:1355/1530 train_loss:3.1888 train_time:232731ms step_avg:173.03ms
step:1356/1530 train_loss:3.5097 train_time:232913ms step_avg:173.04ms
step:1357/1530 train_loss:3.4209 train_time:233094ms step_avg:173.05ms
step:1358/1530 train_loss:3.1813 train_time:233271ms step_avg:173.05ms
step:1359/1530 train_loss:3.4356 train_time:233450ms step_avg:173.05ms
step:1360/1530 train_loss:3.3510 train_time:233630ms step_avg:173.06ms
step:1361/1530 train_loss:3.1157 train_time:233815ms step_avg:173.07ms
step:1362/1530 train_loss:3.3921 train_time:233997ms step_avg:173.07ms
step:1363/1530 train_loss:3.2777 train_time:234183ms step_avg:173.08ms
step:1364/1530 train_loss:3.2974 train_time:234359ms step_avg:173.09ms
step:1365/1530 train_loss:3.3141 train_time:234536ms step_avg:173.09ms
step:1366/1530 train_loss:3.4186 train_time:234715ms step_avg:173.09ms
step:1367/1530 train_loss:3.3926 train_time:234895ms step_avg:173.10ms
step:1368/1530 train_loss:3.3427 train_time:235075ms step_avg:173.10ms
step:1369/1530 train_loss:3.2702 train_time:235264ms step_avg:173.12ms
step:1370/1530 train_loss:3.6018 train_time:235444ms step_avg:173.12ms
step:1371/1530 train_loss:3.3091 train_time:235623ms step_avg:173.13ms
step:1372/1530 train_loss:3.3652 train_time:235805ms step_avg:173.13ms
step:1373/1530 train_loss:3.3648 train_time:235983ms step_avg:173.14ms
step:1374/1530 train_loss:3.1523 train_time:236165ms step_avg:173.14ms
step:1375/1530 train_loss:3.5313 train_time:236345ms step_avg:173.15ms
step:1375/1530 val_loss:3.3085 train_time:236396ms step_avg:173.18ms
step:1376/1530 train_loss:3.3448 train_time:236526ms step_avg:173.15ms
step:1377/1530 train_loss:3.4771 train_time:236704ms step_avg:173.16ms
step:1378/1530 train_loss:3.4609 train_time:236880ms step_avg:173.16ms
step:1379/1530 train_loss:3.1157 train_time:237062ms step_avg:173.16ms
step:1380/1530 train_loss:3.3067 train_time:237240ms step_avg:173.17ms
step:1381/1530 train_loss:3.6923 train_time:237427ms step_avg:173.18ms
step:1382/1530 train_loss:3.2097 train_time:237606ms step_avg:173.18ms
step:1383/1530 train_loss:3.3950 train_time:237787ms step_avg:173.19ms
step:1384/1530 train_loss:3.4702 train_time:237972ms step_avg:173.20ms
step:1385/1530 train_loss:3.4075 train_time:238145ms step_avg:173.20ms
step:1386/1530 train_loss:3.3339 train_time:238325ms step_avg:173.20ms
step:1387/1530 train_loss:3.1943 train_time:238502ms step_avg:173.20ms
step:1388/1530 train_loss:3.3458 train_time:238681ms step_avg:173.21ms
step:1389/1530 train_loss:3.3135 train_time:238865ms step_avg:173.22ms
step:1390/1530 train_loss:3.5658 train_time:239041ms step_avg:173.22ms
step:1391/1530 train_loss:3.2903 train_time:239219ms step_avg:173.22ms
step:1392/1530 train_loss:3.2875 train_time:239398ms step_avg:173.23ms
step:1393/1530 train_loss:3.2332 train_time:239578ms step_avg:173.23ms
step:1394/1530 train_loss:3.4966 train_time:239756ms step_avg:173.23ms
step:1395/1530 train_loss:3.3888 train_time:239935ms step_avg:173.24ms
step:1396/1530 train_loss:3.4034 train_time:240113ms step_avg:173.24ms
step:1397/1530 train_loss:3.3079 train_time:240289ms step_avg:173.24ms
step:1398/1530 train_loss:3.2525 train_time:240465ms step_avg:173.25ms
step:1399/1530 train_loss:3.3123 train_time:240644ms step_avg:173.25ms
step:1400/1530 train_loss:3.3126 train_time:240827ms step_avg:173.26ms
step:1401/1530 train_loss:3.3427 train_time:241003ms step_avg:173.26ms
step:1402/1530 train_loss:3.2966 train_time:241182ms step_avg:173.26ms
step:1403/1530 train_loss:3.4922 train_time:241365ms step_avg:173.27ms
step:1404/1530 train_loss:3.2793 train_time:241541ms step_avg:173.27ms
step:1405/1530 train_loss:3.3129 train_time:241723ms step_avg:173.28ms
step:1406/1530 train_loss:3.3119 train_time:241904ms step_avg:173.28ms
step:1407/1530 train_loss:3.1751 train_time:242080ms step_avg:173.29ms
step:1408/1530 train_loss:3.3099 train_time:242261ms step_avg:173.29ms
step:1409/1530 train_loss:3.2995 train_time:242449ms step_avg:173.30ms
step:1410/1530 train_loss:3.2882 train_time:242626ms step_avg:173.30ms
step:1411/1530 train_loss:3.3643 train_time:242801ms step_avg:173.31ms
step:1412/1530 train_loss:3.3312 train_time:242979ms step_avg:173.31ms
step:1413/1530 train_loss:3.3539 train_time:243159ms step_avg:173.31ms
step:1414/1530 train_loss:3.3242 train_time:243340ms step_avg:173.32ms
step:1415/1530 train_loss:3.4055 train_time:243525ms step_avg:173.33ms
step:1416/1530 train_loss:3.2259 train_time:243713ms step_avg:173.34ms
step:1417/1530 train_loss:3.2789 train_time:243897ms step_avg:173.35ms
step:1418/1530 train_loss:3.3888 train_time:244078ms step_avg:173.35ms
step:1419/1530 train_loss:3.3361 train_time:244261ms step_avg:173.36ms
step:1420/1530 train_loss:3.3628 train_time:244441ms step_avg:173.36ms
step:1421/1530 train_loss:3.3707 train_time:244621ms step_avg:173.37ms
step:1422/1530 train_loss:3.3329 train_time:244798ms step_avg:173.37ms
step:1423/1530 train_loss:3.3179 train_time:244979ms step_avg:173.37ms
step:1424/1530 train_loss:3.3304 train_time:245165ms step_avg:173.38ms
step:1425/1530 train_loss:3.1842 train_time:245354ms step_avg:173.40ms
step:1426/1530 train_loss:3.3172 train_time:245532ms step_avg:173.40ms
step:1427/1530 train_loss:3.2841 train_time:245716ms step_avg:173.41ms
step:1428/1530 train_loss:3.3771 train_time:245896ms step_avg:173.41ms
step:1429/1530 train_loss:3.3538 train_time:246074ms step_avg:173.41ms
step:1430/1530 train_loss:3.2555 train_time:246257ms step_avg:173.42ms
step:1431/1530 train_loss:3.3239 train_time:246439ms step_avg:173.43ms
step:1432/1530 train_loss:3.3333 train_time:246620ms step_avg:173.43ms
step:1433/1530 train_loss:3.1303 train_time:246804ms step_avg:173.44ms
step:1434/1530 train_loss:3.2813 train_time:246988ms step_avg:173.45ms
step:1435/1530 train_loss:3.1190 train_time:247169ms step_avg:173.45ms
step:1436/1530 train_loss:3.2242 train_time:247350ms step_avg:173.46ms
step:1437/1530 train_loss:3.4067 train_time:247527ms step_avg:173.46ms
step:1438/1530 train_loss:3.3813 train_time:247703ms step_avg:173.46ms
step:1439/1530 train_loss:3.3127 train_time:247883ms step_avg:173.47ms
step:1440/1530 train_loss:3.1895 train_time:248058ms step_avg:173.47ms
step:1441/1530 train_loss:3.3338 train_time:248239ms step_avg:173.47ms
step:1442/1530 train_loss:3.3865 train_time:248421ms step_avg:173.48ms
step:1443/1530 train_loss:3.4869 train_time:248608ms step_avg:173.49ms
step:1444/1530 train_loss:3.4482 train_time:248785ms step_avg:173.49ms
step:1445/1530 train_loss:3.3336 train_time:248963ms step_avg:173.49ms
step:1446/1530 train_loss:3.1967 train_time:249143ms step_avg:173.50ms
step:1447/1530 train_loss:3.2941 train_time:249323ms step_avg:173.50ms
step:1448/1530 train_loss:3.2923 train_time:249502ms step_avg:173.51ms
step:1449/1530 train_loss:3.3896 train_time:249681ms step_avg:173.51ms
step:1450/1530 train_loss:3.3847 train_time:249862ms step_avg:173.52ms
step:1451/1530 train_loss:3.2025 train_time:250039ms step_avg:173.52ms
step:1452/1530 train_loss:3.3251 train_time:250219ms step_avg:173.52ms
step:1453/1530 train_loss:3.2585 train_time:250394ms step_avg:173.52ms
step:1454/1530 train_loss:3.2834 train_time:250573ms step_avg:173.53ms
step:1455/1530 train_loss:3.3278 train_time:250755ms step_avg:173.53ms
step:1456/1530 train_loss:3.2816 train_time:250932ms step_avg:173.53ms
step:1457/1530 train_loss:3.1509 train_time:251109ms step_avg:173.54ms
step:1458/1530 train_loss:3.4193 train_time:251285ms step_avg:173.54ms
step:1459/1530 train_loss:3.2705 train_time:251467ms step_avg:173.55ms
step:1460/1530 train_loss:3.3114 train_time:251645ms step_avg:173.55ms
step:1461/1530 train_loss:3.4265 train_time:251825ms step_avg:173.55ms
step:1462/1530 train_loss:3.2617 train_time:252000ms step_avg:173.55ms
step:1463/1530 train_loss:3.4664 train_time:252184ms step_avg:173.56ms
step:1464/1530 train_loss:3.3602 train_time:252362ms step_avg:173.56ms
step:1465/1530 train_loss:3.3606 train_time:252542ms step_avg:173.57ms
step:1466/1530 train_loss:3.2845 train_time:252719ms step_avg:173.57ms
step:1467/1530 train_loss:3.3894 train_time:252900ms step_avg:173.58ms
step:1468/1530 train_loss:3.2842 train_time:253078ms step_avg:173.58ms
step:1469/1530 train_loss:3.2740 train_time:253259ms step_avg:173.58ms
step:1470/1530 train_loss:3.3314 train_time:253442ms step_avg:173.59ms
step:1471/1530 train_loss:3.2550 train_time:253628ms step_avg:173.60ms
step:1472/1530 train_loss:3.2457 train_time:253813ms step_avg:173.61ms
step:1473/1530 train_loss:3.4384 train_time:253991ms step_avg:173.61ms
step:1474/1530 train_loss:3.3097 train_time:254176ms step_avg:173.62ms
step:1475/1530 train_loss:3.1522 train_time:254361ms step_avg:173.63ms
step:1476/1530 train_loss:3.2655 train_time:254539ms step_avg:173.63ms
step:1477/1530 train_loss:3.2355 train_time:254727ms step_avg:173.64ms
step:1478/1530 train_loss:3.3072 train_time:254911ms step_avg:173.64ms
step:1479/1530 train_loss:3.3950 train_time:255093ms step_avg:173.65ms
step:1480/1530 train_loss:3.2681 train_time:255272ms step_avg:173.65ms
step:1481/1530 train_loss:3.4441 train_time:255456ms step_avg:173.66ms
step:1482/1530 train_loss:3.3642 train_time:255643ms step_avg:173.67ms
step:1483/1530 train_loss:3.2768 train_time:255834ms step_avg:173.68ms
step:1484/1530 train_loss:3.2630 train_time:256022ms step_avg:173.69ms
step:1485/1530 train_loss:3.2777 train_time:256201ms step_avg:173.70ms
step:1486/1530 train_loss:3.2251 train_time:256385ms step_avg:173.70ms
step:1487/1530 train_loss:3.3390 train_time:256568ms step_avg:173.71ms
step:1488/1530 train_loss:3.2434 train_time:256754ms step_avg:173.72ms
step:1489/1530 train_loss:3.3095 train_time:256935ms step_avg:173.72ms
step:1490/1530 train_loss:3.2480 train_time:257116ms step_avg:173.73ms
step:1491/1530 train_loss:3.1559 train_time:257298ms step_avg:173.73ms
step:1492/1530 train_loss:3.2648 train_time:257479ms step_avg:173.74ms
step:1493/1530 train_loss:3.4309 train_time:257659ms step_avg:173.74ms
step:1494/1530 train_loss:3.2983 train_time:257837ms step_avg:173.74ms
step:1495/1530 train_loss:3.0293 train_time:258020ms step_avg:173.75ms
step:1496/1530 train_loss:3.3605 train_time:258204ms step_avg:173.76ms
step:1497/1530 train_loss:3.3058 train_time:258388ms step_avg:173.76ms
step:1498/1530 train_loss:3.3426 train_time:258572ms step_avg:173.77ms
step:1499/1530 train_loss:3.3124 train_time:258758ms step_avg:173.78ms
step:1500/1530 train_loss:3.2959 train_time:258949ms step_avg:173.79ms
step:1500/1530 val_loss:3.2775 train_time:259003ms step_avg:173.83ms
step:1501/1530 train_loss:3.0883 train_time:259139ms step_avg:173.80ms
step:1502/1530 train_loss:3.3557 train_time:259331ms step_avg:173.81ms
step:1503/1530 train_loss:3.2400 train_time:259509ms step_avg:173.82ms
step:1504/1530 train_loss:3.2440 train_time:259691ms step_avg:173.82ms
step:1505/1530 train_loss:3.2099 train_time:259871ms step_avg:173.83ms
step:1506/1530 train_loss:3.2776 train_time:260053ms step_avg:173.83ms
step:1507/1530 train_loss:3.1799 train_time:260248ms step_avg:173.85ms
step:1508/1530 train_loss:3.4815 train_time:260431ms step_avg:173.85ms
step:1509/1530 train_loss:3.2774 train_time:260608ms step_avg:173.85ms
step:1510/1530 train_loss:3.2725 train_time:260790ms step_avg:173.86ms
step:1511/1530 train_loss:3.4145 train_time:261100ms step_avg:173.95ms
step:1512/1530 train_loss:3.4215 train_time:261288ms step_avg:173.96ms
step:1513/1530 train_loss:3.2680 train_time:261472ms step_avg:173.97ms
step:1514/1530 train_loss:3.0808 train_time:261654ms step_avg:173.97ms
step:1515/1530 train_loss:3.2400 train_time:261835ms step_avg:173.98ms
step:1516/1530 train_loss:3.2544 train_time:262021ms step_avg:173.98ms
step:1517/1530 train_loss:3.2965 train_time:262202ms step_avg:173.99ms
step:1518/1530 train_loss:3.2013 train_time:262385ms step_avg:174.00ms
step:1519/1530 train_loss:3.4991 train_time:262714ms step_avg:174.10ms
step:1520/1530 train_loss:3.1263 train_time:262896ms step_avg:174.10ms
step:1521/1530 train_loss:3.2046 train_time:263072ms step_avg:174.10ms
step:1522/1530 train_loss:3.3558 train_time:263255ms step_avg:174.11ms
step:1523/1530 train_loss:3.2251 train_time:263434ms step_avg:174.11ms
step:1524/1530 train_loss:3.3454 train_time:263614ms step_avg:174.12ms
step:1525/1530 train_loss:3.3317 train_time:263802ms step_avg:174.13ms
step:1526/1530 train_loss:3.2760 train_time:263991ms step_avg:174.14ms
step:1527/1530 train_loss:3.2897 train_time:264173ms step_avg:174.14ms
step:1528/1530 train_loss:3.4041 train_time:264352ms step_avg:174.14ms
step:1529/1530 train_loss:3.4110 train_time:264529ms step_avg:174.15ms
step:1530/1530 train_loss:3.2364 train_time:264707ms step_avg:174.15ms
step:1530/1530 val_loss:3.2750 train_time:264761ms step_avg:174.18ms