c249f3c7-b947-4b6f-8a42-56e97ef7712e.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 02:57:20 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   39C    P0              76W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          On  | 00000000:3B:00.0 Off |                    0 |
| N/A   30C    P0              73W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          On  | 00000000:4C:00.0 Off |                    0 |
| N/A   30C    P0              73W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   39C    P0             119W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   40C    P0              99W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   29C    P0              70W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   40C    P0             107W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   30C    P0             114W / 700W |     22MiB / 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:31957ms step_avg:nanms
step:2/1530 train_loss:10.0815 train_time:32069ms step_avg:nanms
step:3/1530 train_loss:8.3958 train_time:32227ms step_avg:nanms
step:4/1530 train_loss:7.5350 train_time:32388ms step_avg:nanms
step:5/1530 train_loss:7.4574 train_time:32549ms step_avg:nanms
step:6/1530 train_loss:6.9802 train_time:32709ms step_avg:nanms
step:7/1530 train_loss:7.1917 train_time:32869ms step_avg:nanms
step:8/1530 train_loss:6.7305 train_time:33029ms step_avg:nanms
step:9/1530 train_loss:6.6114 train_time:33190ms step_avg:nanms
step:10/1530 train_loss:6.4987 train_time:33349ms step_avg:nanms
step:11/1530 train_loss:6.4679 train_time:115ms step_avg:nanms
step:12/1530 train_loss:6.3592 train_time:275ms step_avg:nanms
step:13/1530 train_loss:6.2551 train_time:435ms step_avg:144.84ms
step:14/1530 train_loss:6.2441 train_time:595ms step_avg:148.64ms
step:15/1530 train_loss:6.1670 train_time:755ms step_avg:150.98ms
step:16/1530 train_loss:6.1115 train_time:916ms step_avg:152.71ms
step:17/1530 train_loss:6.1569 train_time:1076ms step_avg:153.65ms
step:18/1530 train_loss:5.9711 train_time:1236ms step_avg:154.50ms
step:19/1530 train_loss:5.9678 train_time:1396ms step_avg:155.12ms
step:20/1530 train_loss:5.6896 train_time:1557ms step_avg:155.70ms
step:21/1530 train_loss:5.9484 train_time:1718ms step_avg:156.15ms
step:22/1530 train_loss:6.1642 train_time:1878ms step_avg:156.52ms
step:23/1530 train_loss:5.8312 train_time:2038ms step_avg:156.78ms
step:24/1530 train_loss:6.0114 train_time:2198ms step_avg:156.99ms
step:25/1530 train_loss:5.6715 train_time:2359ms step_avg:157.29ms
step:26/1530 train_loss:5.5841 train_time:2521ms step_avg:157.56ms
step:27/1530 train_loss:5.7579 train_time:2682ms step_avg:157.79ms
step:28/1530 train_loss:5.4027 train_time:2843ms step_avg:157.94ms
step:29/1530 train_loss:5.6503 train_time:3004ms step_avg:158.11ms
step:30/1530 train_loss:5.4607 train_time:3164ms step_avg:158.20ms
step:31/1530 train_loss:5.4275 train_time:3326ms step_avg:158.36ms
step:32/1530 train_loss:5.2874 train_time:3486ms step_avg:158.44ms
step:33/1530 train_loss:5.5878 train_time:3646ms step_avg:158.52ms
step:34/1530 train_loss:5.4981 train_time:3807ms step_avg:158.62ms
step:35/1530 train_loss:5.6116 train_time:3966ms step_avg:158.66ms
step:36/1530 train_loss:5.5434 train_time:4127ms step_avg:158.74ms
step:37/1530 train_loss:5.4454 train_time:4287ms step_avg:158.79ms
step:38/1530 train_loss:5.2909 train_time:4448ms step_avg:158.87ms
step:39/1530 train_loss:5.3169 train_time:4609ms step_avg:158.94ms
step:40/1530 train_loss:5.2414 train_time:4769ms step_avg:158.96ms
step:41/1530 train_loss:5.2235 train_time:4930ms step_avg:159.02ms
step:42/1530 train_loss:5.1795 train_time:5090ms step_avg:159.07ms
step:43/1530 train_loss:5.2721 train_time:5250ms step_avg:159.09ms
step:44/1530 train_loss:5.2368 train_time:5410ms step_avg:159.11ms
step:45/1530 train_loss:5.3654 train_time:5570ms step_avg:159.14ms
step:46/1530 train_loss:5.1522 train_time:5730ms step_avg:159.16ms
step:47/1530 train_loss:5.0392 train_time:5890ms step_avg:159.19ms
step:48/1530 train_loss:5.1914 train_time:6049ms step_avg:159.19ms
step:49/1530 train_loss:5.1223 train_time:6210ms step_avg:159.24ms
step:50/1530 train_loss:5.2349 train_time:6370ms step_avg:159.25ms
step:51/1530 train_loss:5.1305 train_time:6529ms step_avg:159.25ms
step:52/1530 train_loss:5.0325 train_time:6690ms step_avg:159.29ms
step:53/1530 train_loss:5.1791 train_time:6851ms step_avg:159.31ms
step:54/1530 train_loss:5.0111 train_time:7010ms step_avg:159.32ms
step:55/1530 train_loss:5.4065 train_time:7171ms step_avg:159.36ms
step:56/1530 train_loss:5.0345 train_time:7331ms step_avg:159.37ms
step:57/1530 train_loss:4.8768 train_time:7491ms step_avg:159.38ms
step:58/1530 train_loss:5.0351 train_time:7651ms step_avg:159.40ms
step:59/1530 train_loss:5.0076 train_time:7812ms step_avg:159.42ms
step:60/1530 train_loss:5.1369 train_time:7972ms step_avg:159.44ms
step:61/1530 train_loss:4.8518 train_time:8131ms step_avg:159.42ms
step:62/1530 train_loss:4.9630 train_time:8292ms step_avg:159.45ms
step:63/1530 train_loss:4.9557 train_time:8452ms step_avg:159.47ms
step:64/1530 train_loss:4.8739 train_time:8612ms step_avg:159.47ms
step:65/1530 train_loss:4.7943 train_time:8771ms step_avg:159.48ms
step:66/1530 train_loss:4.9225 train_time:8932ms step_avg:159.50ms
step:67/1530 train_loss:4.8280 train_time:9092ms step_avg:159.51ms
step:68/1530 train_loss:5.0967 train_time:9252ms step_avg:159.51ms
step:69/1530 train_loss:4.7220 train_time:9412ms step_avg:159.52ms
step:70/1530 train_loss:4.8663 train_time:9572ms step_avg:159.54ms
step:71/1530 train_loss:4.9755 train_time:9732ms step_avg:159.54ms
step:72/1530 train_loss:4.8709 train_time:9892ms step_avg:159.54ms
step:73/1530 train_loss:4.7541 train_time:10053ms step_avg:159.57ms
step:74/1530 train_loss:4.9088 train_time:10213ms step_avg:159.57ms
step:75/1530 train_loss:4.8382 train_time:10372ms step_avg:159.57ms
step:76/1530 train_loss:4.7932 train_time:10532ms step_avg:159.58ms
step:77/1530 train_loss:4.9084 train_time:10693ms step_avg:159.60ms
step:78/1530 train_loss:5.1351 train_time:10854ms step_avg:159.61ms
step:79/1530 train_loss:4.7955 train_time:11013ms step_avg:159.61ms
step:80/1530 train_loss:4.8296 train_time:11174ms step_avg:159.63ms
step:81/1530 train_loss:4.6318 train_time:11335ms step_avg:159.65ms
step:82/1530 train_loss:4.8133 train_time:11496ms step_avg:159.66ms
step:83/1530 train_loss:4.7589 train_time:11656ms step_avg:159.68ms
step:84/1530 train_loss:4.7352 train_time:11817ms step_avg:159.70ms
step:85/1530 train_loss:4.5912 train_time:11978ms step_avg:159.70ms
step:86/1530 train_loss:4.8106 train_time:12137ms step_avg:159.70ms
step:87/1530 train_loss:4.7302 train_time:12298ms step_avg:159.72ms
step:88/1530 train_loss:4.7312 train_time:12460ms step_avg:159.75ms
step:89/1530 train_loss:4.6993 train_time:12621ms step_avg:159.76ms
step:90/1530 train_loss:4.6467 train_time:12783ms step_avg:159.79ms
step:91/1530 train_loss:4.6391 train_time:12944ms step_avg:159.80ms
step:92/1530 train_loss:4.8136 train_time:13105ms step_avg:159.82ms
step:93/1530 train_loss:4.6221 train_time:13265ms step_avg:159.81ms
step:94/1530 train_loss:4.6246 train_time:13427ms step_avg:159.84ms
step:95/1530 train_loss:4.6737 train_time:13588ms step_avg:159.85ms
step:96/1530 train_loss:4.5706 train_time:13748ms step_avg:159.86ms
step:97/1530 train_loss:4.6262 train_time:13909ms step_avg:159.87ms
step:98/1530 train_loss:4.5521 train_time:14069ms step_avg:159.88ms
step:99/1530 train_loss:4.6397 train_time:14230ms step_avg:159.89ms
step:100/1530 train_loss:4.6693 train_time:14391ms step_avg:159.90ms
step:101/1530 train_loss:4.5254 train_time:14551ms step_avg:159.90ms
step:102/1530 train_loss:4.6877 train_time:14712ms step_avg:159.91ms
step:103/1530 train_loss:4.5643 train_time:14872ms step_avg:159.92ms
step:104/1530 train_loss:4.5296 train_time:15032ms step_avg:159.92ms
step:105/1530 train_loss:4.5483 train_time:15192ms step_avg:159.92ms
step:106/1530 train_loss:4.5872 train_time:15352ms step_avg:159.92ms
step:107/1530 train_loss:4.4974 train_time:15514ms step_avg:159.94ms
step:108/1530 train_loss:4.3517 train_time:15674ms step_avg:159.94ms
step:109/1530 train_loss:4.4693 train_time:15834ms step_avg:159.94ms
step:110/1530 train_loss:4.4773 train_time:15995ms step_avg:159.95ms
step:111/1530 train_loss:4.4121 train_time:16154ms step_avg:159.94ms
step:112/1530 train_loss:4.5729 train_time:16315ms step_avg:159.96ms
step:113/1530 train_loss:4.4817 train_time:16475ms step_avg:159.95ms
step:114/1530 train_loss:4.3527 train_time:16636ms step_avg:159.96ms
step:115/1530 train_loss:4.5015 train_time:16799ms step_avg:159.99ms
step:116/1530 train_loss:4.4531 train_time:16963ms step_avg:160.03ms
step:117/1530 train_loss:4.3590 train_time:17127ms step_avg:160.07ms
step:118/1530 train_loss:4.5822 train_time:17291ms step_avg:160.10ms
step:119/1530 train_loss:4.4320 train_time:17453ms step_avg:160.12ms
step:120/1530 train_loss:4.3196 train_time:17619ms step_avg:160.17ms
step:121/1530 train_loss:4.2945 train_time:17784ms step_avg:160.21ms
step:122/1530 train_loss:4.4399 train_time:17947ms step_avg:160.24ms
step:123/1530 train_loss:4.2731 train_time:18112ms step_avg:160.28ms
step:124/1530 train_loss:4.5758 train_time:18276ms step_avg:160.31ms
step:125/1530 train_loss:4.4382 train_time:18439ms step_avg:160.34ms
step:125/1530 val_loss:4.3977 train_time:18486ms step_avg:160.75ms
step:126/1530 train_loss:4.4104 train_time:18605ms step_avg:160.39ms
step:127/1530 train_loss:4.4252 train_time:18771ms step_avg:160.44ms
step:128/1530 train_loss:4.3701 train_time:18935ms step_avg:160.47ms
step:129/1530 train_loss:4.6806 train_time:19099ms step_avg:160.50ms
step:130/1530 train_loss:4.3578 train_time:19263ms step_avg:160.53ms
step:131/1530 train_loss:4.3979 train_time:19427ms step_avg:160.56ms
step:132/1530 train_loss:4.3473 train_time:19591ms step_avg:160.58ms
step:133/1530 train_loss:4.4452 train_time:19755ms step_avg:160.61ms
step:134/1530 train_loss:4.2639 train_time:19919ms step_avg:160.63ms
step:135/1530 train_loss:4.4483 train_time:20083ms step_avg:160.67ms
step:136/1530 train_loss:4.2142 train_time:20248ms step_avg:160.69ms
step:137/1530 train_loss:4.3786 train_time:20411ms step_avg:160.72ms
step:138/1530 train_loss:4.2902 train_time:20576ms step_avg:160.75ms
step:139/1530 train_loss:4.3760 train_time:20739ms step_avg:160.77ms
step:140/1530 train_loss:4.4678 train_time:20903ms step_avg:160.79ms
step:141/1530 train_loss:4.3052 train_time:21067ms step_avg:160.82ms
step:142/1530 train_loss:4.3010 train_time:21232ms step_avg:160.85ms
step:143/1530 train_loss:4.2561 train_time:21397ms step_avg:160.88ms
step:144/1530 train_loss:4.3501 train_time:21561ms step_avg:160.90ms
step:145/1530 train_loss:4.3035 train_time:21727ms step_avg:160.94ms
step:146/1530 train_loss:4.1695 train_time:21892ms step_avg:160.97ms
step:147/1530 train_loss:4.3257 train_time:22056ms step_avg:160.99ms
step:148/1530 train_loss:4.3633 train_time:22220ms step_avg:161.01ms
step:149/1530 train_loss:4.2967 train_time:22383ms step_avg:161.03ms
step:150/1530 train_loss:4.4348 train_time:22548ms step_avg:161.06ms
step:151/1530 train_loss:4.2691 train_time:22712ms step_avg:161.08ms
step:152/1530 train_loss:4.2714 train_time:22876ms step_avg:161.10ms
step:153/1530 train_loss:4.3686 train_time:23040ms step_avg:161.12ms
step:154/1530 train_loss:4.3725 train_time:23204ms step_avg:161.14ms
step:155/1530 train_loss:4.2652 train_time:23367ms step_avg:161.15ms
step:156/1530 train_loss:4.3452 train_time:23532ms step_avg:161.18ms
step:157/1530 train_loss:4.4000 train_time:23696ms step_avg:161.20ms
step:158/1530 train_loss:4.2382 train_time:23860ms step_avg:161.22ms
step:159/1530 train_loss:4.3035 train_time:24024ms step_avg:161.24ms
step:160/1530 train_loss:4.1246 train_time:24190ms step_avg:161.26ms
step:161/1530 train_loss:4.3492 train_time:24354ms step_avg:161.28ms
step:162/1530 train_loss:4.3643 train_time:24518ms step_avg:161.30ms
step:163/1530 train_loss:4.3479 train_time:24681ms step_avg:161.31ms
step:164/1530 train_loss:4.1851 train_time:24844ms step_avg:161.33ms
step:165/1530 train_loss:4.2768 train_time:25008ms step_avg:161.34ms
step:166/1530 train_loss:4.3388 train_time:25173ms step_avg:161.37ms
step:167/1530 train_loss:4.1948 train_time:25338ms step_avg:161.39ms
step:168/1530 train_loss:4.2858 train_time:25502ms step_avg:161.40ms
step:169/1530 train_loss:4.1628 train_time:25665ms step_avg:161.42ms
step:170/1530 train_loss:4.0267 train_time:25831ms step_avg:161.44ms
step:171/1530 train_loss:4.2031 train_time:25995ms step_avg:161.46ms
step:172/1530 train_loss:4.2125 train_time:26158ms step_avg:161.47ms
step:173/1530 train_loss:4.2673 train_time:26321ms step_avg:161.48ms
step:174/1530 train_loss:4.4194 train_time:26485ms step_avg:161.49ms
step:175/1530 train_loss:4.2447 train_time:26647ms step_avg:161.50ms
step:176/1530 train_loss:4.0892 train_time:26810ms step_avg:161.51ms
step:177/1530 train_loss:4.0770 train_time:26974ms step_avg:161.52ms
step:178/1530 train_loss:4.1816 train_time:27136ms step_avg:161.52ms
step:179/1530 train_loss:4.1323 train_time:27299ms step_avg:161.53ms
step:180/1530 train_loss:4.1119 train_time:27462ms step_avg:161.54ms
step:181/1530 train_loss:4.2966 train_time:27626ms step_avg:161.56ms
step:182/1530 train_loss:4.1532 train_time:27790ms step_avg:161.57ms
step:183/1530 train_loss:4.1298 train_time:27953ms step_avg:161.58ms
step:184/1530 train_loss:4.1315 train_time:28116ms step_avg:161.59ms
step:185/1530 train_loss:4.2058 train_time:28280ms step_avg:161.60ms
step:186/1530 train_loss:4.1731 train_time:28442ms step_avg:161.60ms
step:187/1530 train_loss:4.2365 train_time:28604ms step_avg:161.60ms
step:188/1530 train_loss:4.1671 train_time:28903ms step_avg:162.37ms
step:189/1530 train_loss:4.1159 train_time:29242ms step_avg:163.36ms
step:190/1530 train_loss:4.2105 train_time:29403ms step_avg:163.35ms
step:191/1530 train_loss:4.0840 train_time:29567ms step_avg:163.35ms
step:192/1530 train_loss:4.0351 train_time:29730ms step_avg:163.35ms
step:193/1530 train_loss:4.2514 train_time:29895ms step_avg:163.36ms
step:194/1530 train_loss:4.1749 train_time:30057ms step_avg:163.35ms
step:195/1530 train_loss:4.3650 train_time:30219ms step_avg:163.35ms
step:196/1530 train_loss:4.1848 train_time:30383ms step_avg:163.35ms
step:197/1530 train_loss:4.0484 train_time:30546ms step_avg:163.35ms
step:198/1530 train_loss:4.1889 train_time:30707ms step_avg:163.34ms
step:199/1530 train_loss:4.0450 train_time:30870ms step_avg:163.34ms
step:200/1530 train_loss:4.1192 train_time:31034ms step_avg:163.34ms
step:201/1530 train_loss:4.0168 train_time:31197ms step_avg:163.34ms
step:202/1530 train_loss:4.2502 train_time:31360ms step_avg:163.33ms
step:203/1530 train_loss:4.0622 train_time:31523ms step_avg:163.33ms
step:204/1530 train_loss:4.1933 train_time:31685ms step_avg:163.33ms
step:205/1530 train_loss:4.2572 train_time:31849ms step_avg:163.33ms
step:206/1530 train_loss:3.9545 train_time:32011ms step_avg:163.32ms
step:207/1530 train_loss:4.0852 train_time:32175ms step_avg:163.32ms
step:208/1530 train_loss:4.1032 train_time:32337ms step_avg:163.32ms
step:209/1530 train_loss:4.2389 train_time:32501ms step_avg:163.32ms
step:210/1530 train_loss:4.1646 train_time:32664ms step_avg:163.32ms
step:211/1530 train_loss:4.0635 train_time:32826ms step_avg:163.32ms
step:212/1530 train_loss:4.1334 train_time:32990ms step_avg:163.32ms
step:213/1530 train_loss:4.0552 train_time:33153ms step_avg:163.32ms
step:214/1530 train_loss:4.1077 train_time:33316ms step_avg:163.31ms
step:215/1530 train_loss:3.9549 train_time:33478ms step_avg:163.31ms
step:216/1530 train_loss:4.0110 train_time:33641ms step_avg:163.30ms
step:217/1530 train_loss:4.0231 train_time:33804ms step_avg:163.30ms
step:218/1530 train_loss:4.0904 train_time:33967ms step_avg:163.30ms
step:219/1530 train_loss:4.0889 train_time:34130ms step_avg:163.30ms
step:220/1530 train_loss:4.0851 train_time:34294ms step_avg:163.30ms
step:221/1530 train_loss:4.0950 train_time:34456ms step_avg:163.30ms
step:222/1530 train_loss:4.0045 train_time:34619ms step_avg:163.30ms
step:223/1530 train_loss:3.9980 train_time:34783ms step_avg:163.30ms
step:224/1530 train_loss:4.2989 train_time:34946ms step_avg:163.30ms
step:225/1530 train_loss:3.9300 train_time:35110ms step_avg:163.30ms
step:226/1530 train_loss:3.9948 train_time:35274ms step_avg:163.31ms
step:227/1530 train_loss:3.9779 train_time:35437ms step_avg:163.30ms
step:228/1530 train_loss:4.1565 train_time:35602ms step_avg:163.31ms
step:229/1530 train_loss:3.9349 train_time:35769ms step_avg:163.33ms
step:230/1530 train_loss:4.0437 train_time:35935ms step_avg:163.34ms
step:231/1530 train_loss:3.9031 train_time:36101ms step_avg:163.35ms
step:232/1530 train_loss:3.9689 train_time:36267ms step_avg:163.36ms
step:233/1530 train_loss:4.0984 train_time:36434ms step_avg:163.38ms
step:234/1530 train_loss:4.0403 train_time:36599ms step_avg:163.39ms
step:235/1530 train_loss:3.9054 train_time:36766ms step_avg:163.41ms
step:236/1530 train_loss:4.0800 train_time:36933ms step_avg:163.42ms
step:237/1530 train_loss:4.0794 train_time:37099ms step_avg:163.43ms
step:238/1530 train_loss:3.9455 train_time:37266ms step_avg:163.45ms
step:239/1530 train_loss:4.0862 train_time:37431ms step_avg:163.45ms
step:240/1530 train_loss:4.1263 train_time:37598ms step_avg:163.47ms
step:241/1530 train_loss:3.9803 train_time:37763ms step_avg:163.48ms
step:242/1530 train_loss:4.1706 train_time:37930ms step_avg:163.49ms
step:243/1530 train_loss:4.0204 train_time:38096ms step_avg:163.50ms
step:244/1530 train_loss:4.0836 train_time:38262ms step_avg:163.51ms
step:245/1530 train_loss:4.1485 train_time:38428ms step_avg:163.52ms
step:246/1530 train_loss:4.0670 train_time:38595ms step_avg:163.54ms
step:247/1530 train_loss:4.0054 train_time:38761ms step_avg:163.55ms
step:248/1530 train_loss:4.1085 train_time:38929ms step_avg:163.57ms
step:249/1530 train_loss:3.9338 train_time:39095ms step_avg:163.58ms
step:250/1530 train_loss:3.9749 train_time:39261ms step_avg:163.59ms
step:250/1530 val_loss:4.0085 train_time:39309ms step_avg:163.79ms
step:251/1530 train_loss:4.0797 train_time:39430ms step_avg:163.61ms
step:252/1530 train_loss:4.1680 train_time:39598ms step_avg:163.63ms
step:253/1530 train_loss:3.9376 train_time:39765ms step_avg:163.64ms
step:254/1530 train_loss:3.8882 train_time:39931ms step_avg:163.65ms
step:255/1530 train_loss:4.0836 train_time:40096ms step_avg:163.66ms
step:256/1530 train_loss:3.9932 train_time:40262ms step_avg:163.67ms
step:257/1530 train_loss:3.9940 train_time:40428ms step_avg:163.68ms
step:258/1530 train_loss:3.9883 train_time:40594ms step_avg:163.68ms
step:259/1530 train_loss:4.0324 train_time:40760ms step_avg:163.69ms
step:260/1530 train_loss:4.0665 train_time:40927ms step_avg:163.71ms
step:261/1530 train_loss:4.0343 train_time:41093ms step_avg:163.72ms
step:262/1530 train_loss:3.9938 train_time:41259ms step_avg:163.73ms
step:263/1530 train_loss:3.8981 train_time:41426ms step_avg:163.74ms
step:264/1530 train_loss:3.9898 train_time:41591ms step_avg:163.74ms
step:265/1530 train_loss:3.8729 train_time:41757ms step_avg:163.75ms
step:266/1530 train_loss:3.9281 train_time:41924ms step_avg:163.77ms
step:267/1530 train_loss:3.9431 train_time:42089ms step_avg:163.77ms
step:268/1530 train_loss:3.9629 train_time:42255ms step_avg:163.78ms
step:269/1530 train_loss:3.8600 train_time:42421ms step_avg:163.79ms
step:270/1530 train_loss:4.1027 train_time:42587ms step_avg:163.79ms
step:271/1530 train_loss:3.9715 train_time:42753ms step_avg:163.80ms
step:272/1530 train_loss:3.9305 train_time:42920ms step_avg:163.82ms
step:273/1530 train_loss:3.9464 train_time:43085ms step_avg:163.82ms
step:274/1530 train_loss:4.0546 train_time:43252ms step_avg:163.83ms
step:275/1530 train_loss:4.0674 train_time:43418ms step_avg:163.84ms
step:276/1530 train_loss:4.2371 train_time:43585ms step_avg:163.85ms
step:277/1530 train_loss:4.0462 train_time:43750ms step_avg:163.86ms
step:278/1530 train_loss:4.0907 train_time:43917ms step_avg:163.87ms
step:279/1530 train_loss:3.9979 train_time:44084ms step_avg:163.88ms
step:280/1530 train_loss:4.2442 train_time:44251ms step_avg:163.89ms
step:281/1530 train_loss:3.9891 train_time:44418ms step_avg:163.90ms
step:282/1530 train_loss:3.9495 train_time:44584ms step_avg:163.91ms
step:283/1530 train_loss:3.9220 train_time:44749ms step_avg:163.92ms
step:284/1530 train_loss:4.0623 train_time:44917ms step_avg:163.93ms
step:285/1530 train_loss:4.0711 train_time:45083ms step_avg:163.94ms
step:286/1530 train_loss:4.0978 train_time:45248ms step_avg:163.94ms
step:287/1530 train_loss:3.9136 train_time:45412ms step_avg:163.94ms
step:288/1530 train_loss:4.0229 train_time:45578ms step_avg:163.95ms
step:289/1530 train_loss:3.8729 train_time:45743ms step_avg:163.95ms
step:290/1530 train_loss:3.8608 train_time:45908ms step_avg:163.96ms
step:291/1530 train_loss:3.9176 train_time:46073ms step_avg:163.96ms
step:292/1530 train_loss:3.8684 train_time:46239ms step_avg:163.97ms
step:293/1530 train_loss:3.9031 train_time:46404ms step_avg:163.97ms
step:294/1530 train_loss:3.9478 train_time:46569ms step_avg:163.98ms
step:295/1530 train_loss:3.8448 train_time:46735ms step_avg:163.98ms
step:296/1530 train_loss:3.8678 train_time:46899ms step_avg:163.98ms
step:297/1530 train_loss:3.8682 train_time:47064ms step_avg:163.99ms
step:298/1530 train_loss:3.9759 train_time:47230ms step_avg:163.99ms
step:299/1530 train_loss:3.8289 train_time:47395ms step_avg:164.00ms
step:300/1530 train_loss:3.9773 train_time:47560ms step_avg:164.00ms
step:301/1530 train_loss:3.9675 train_time:47726ms step_avg:164.01ms
step:302/1530 train_loss:3.9349 train_time:47890ms step_avg:164.01ms
step:303/1530 train_loss:3.9790 train_time:48056ms step_avg:164.01ms
step:304/1530 train_loss:3.9712 train_time:48223ms step_avg:164.02ms
step:305/1530 train_loss:4.4603 train_time:48388ms step_avg:164.03ms
step:306/1530 train_loss:3.9439 train_time:48553ms step_avg:164.03ms
step:307/1530 train_loss:3.8364 train_time:48717ms step_avg:164.03ms
step:308/1530 train_loss:3.9872 train_time:48883ms step_avg:164.04ms
step:309/1530 train_loss:3.8672 train_time:49048ms step_avg:164.04ms
step:310/1530 train_loss:4.0877 train_time:49213ms step_avg:164.04ms
step:311/1530 train_loss:3.9380 train_time:49378ms step_avg:164.05ms
step:312/1530 train_loss:3.8722 train_time:49544ms step_avg:164.05ms
step:313/1530 train_loss:3.9447 train_time:49709ms step_avg:164.06ms
step:314/1530 train_loss:4.0711 train_time:49875ms step_avg:164.06ms
step:315/1530 train_loss:3.9495 train_time:50040ms step_avg:164.07ms
step:316/1530 train_loss:3.8028 train_time:50206ms step_avg:164.07ms
step:317/1530 train_loss:3.8814 train_time:50371ms step_avg:164.08ms
step:318/1530 train_loss:3.9347 train_time:50536ms step_avg:164.08ms
step:319/1530 train_loss:3.9013 train_time:50701ms step_avg:164.08ms
step:320/1530 train_loss:4.0199 train_time:50867ms step_avg:164.09ms
step:321/1530 train_loss:3.9664 train_time:51032ms step_avg:164.09ms
step:322/1530 train_loss:3.9342 train_time:51199ms step_avg:164.10ms
step:323/1530 train_loss:4.0161 train_time:51365ms step_avg:164.10ms
step:324/1530 train_loss:3.9508 train_time:51530ms step_avg:164.11ms
step:325/1530 train_loss:4.0312 train_time:51695ms step_avg:164.11ms
step:326/1530 train_loss:3.9087 train_time:51860ms step_avg:164.11ms
step:327/1530 train_loss:4.4070 train_time:52026ms step_avg:164.12ms
step:328/1530 train_loss:4.0805 train_time:52191ms step_avg:164.12ms
step:329/1530 train_loss:3.8058 train_time:52355ms step_avg:164.12ms
step:330/1530 train_loss:3.7450 train_time:52522ms step_avg:164.13ms
step:331/1530 train_loss:3.9834 train_time:52687ms step_avg:164.13ms
step:332/1530 train_loss:3.9178 train_time:52852ms step_avg:164.14ms
step:333/1530 train_loss:3.8862 train_time:53017ms step_avg:164.14ms
step:334/1530 train_loss:3.8519 train_time:53183ms step_avg:164.14ms
step:335/1530 train_loss:4.0139 train_time:53349ms step_avg:164.15ms
step:336/1530 train_loss:3.9671 train_time:53512ms step_avg:164.15ms
step:337/1530 train_loss:4.4249 train_time:53680ms step_avg:164.16ms
step:338/1530 train_loss:3.9407 train_time:53846ms step_avg:164.17ms
step:339/1530 train_loss:3.8699 train_time:54011ms step_avg:164.17ms
step:340/1530 train_loss:3.9421 train_time:54175ms step_avg:164.17ms
step:341/1530 train_loss:3.8661 train_time:54343ms step_avg:164.18ms
step:342/1530 train_loss:3.8156 train_time:54510ms step_avg:164.19ms
step:343/1530 train_loss:3.8484 train_time:54679ms step_avg:164.20ms
step:344/1530 train_loss:3.9981 train_time:54847ms step_avg:164.21ms
step:345/1530 train_loss:3.8271 train_time:55017ms step_avg:164.23ms
step:346/1530 train_loss:3.7703 train_time:55185ms step_avg:164.24ms
step:347/1530 train_loss:3.8081 train_time:55353ms step_avg:164.25ms
step:348/1530 train_loss:3.8695 train_time:55521ms step_avg:164.26ms
step:349/1530 train_loss:3.8387 train_time:55689ms step_avg:164.27ms
step:350/1530 train_loss:3.5756 train_time:55857ms step_avg:164.29ms
step:351/1530 train_loss:3.8361 train_time:56026ms step_avg:164.30ms
step:352/1530 train_loss:4.1814 train_time:56195ms step_avg:164.31ms
step:353/1530 train_loss:3.6570 train_time:56363ms step_avg:164.32ms
step:354/1530 train_loss:3.9309 train_time:56529ms step_avg:164.33ms
step:355/1530 train_loss:3.7947 train_time:56698ms step_avg:164.34ms
step:356/1530 train_loss:3.8858 train_time:56866ms step_avg:164.35ms
step:357/1530 train_loss:3.7826 train_time:57035ms step_avg:164.37ms
step:358/1530 train_loss:3.8695 train_time:57204ms step_avg:164.38ms
step:359/1530 train_loss:3.8042 train_time:57372ms step_avg:164.39ms
step:360/1530 train_loss:3.4373 train_time:57540ms step_avg:164.40ms
step:361/1530 train_loss:4.0266 train_time:57709ms step_avg:164.41ms
step:362/1530 train_loss:3.9241 train_time:57876ms step_avg:164.42ms
step:363/1530 train_loss:3.8467 train_time:58044ms step_avg:164.43ms
step:364/1530 train_loss:3.7578 train_time:58212ms step_avg:164.44ms
step:365/1530 train_loss:3.9255 train_time:58381ms step_avg:164.45ms
step:366/1530 train_loss:3.8748 train_time:58550ms step_avg:164.47ms
step:367/1530 train_loss:3.8644 train_time:58718ms step_avg:164.48ms
step:368/1530 train_loss:3.8595 train_time:58886ms step_avg:164.49ms
step:369/1530 train_loss:3.7530 train_time:59054ms step_avg:164.50ms
step:370/1530 train_loss:3.8927 train_time:59221ms step_avg:164.50ms
step:371/1530 train_loss:3.7369 train_time:59388ms step_avg:164.51ms
step:372/1530 train_loss:3.7044 train_time:59556ms step_avg:164.52ms
step:373/1530 train_loss:3.9239 train_time:59724ms step_avg:164.53ms
step:374/1530 train_loss:3.8386 train_time:59891ms step_avg:164.54ms
step:375/1530 train_loss:3.8094 train_time:60059ms step_avg:164.54ms
step:375/1530 val_loss:3.8331 train_time:60108ms step_avg:164.68ms
step:376/1530 train_loss:3.8702 train_time:60230ms step_avg:164.56ms
step:377/1530 train_loss:3.7978 train_time:60533ms step_avg:164.94ms
step:378/1530 train_loss:3.8506 train_time:60711ms step_avg:164.98ms
step:379/1530 train_loss:3.8769 train_time:61043ms step_avg:165.43ms
step:380/1530 train_loss:3.9583 train_time:61211ms step_avg:165.44ms
step:381/1530 train_loss:3.8511 train_time:61378ms step_avg:165.44ms
step:382/1530 train_loss:3.8162 train_time:61549ms step_avg:165.45ms
step:383/1530 train_loss:3.8053 train_time:61717ms step_avg:165.46ms
step:384/1530 train_loss:3.8797 train_time:61883ms step_avg:165.46ms
step:385/1530 train_loss:3.7990 train_time:62052ms step_avg:165.47ms
step:386/1530 train_loss:3.8958 train_time:62220ms step_avg:165.48ms
step:387/1530 train_loss:4.0690 train_time:62389ms step_avg:165.49ms
step:388/1530 train_loss:3.7978 train_time:62556ms step_avg:165.49ms
step:389/1530 train_loss:3.7974 train_time:62724ms step_avg:165.50ms
step:390/1530 train_loss:3.8997 train_time:62893ms step_avg:165.51ms
step:391/1530 train_loss:3.8157 train_time:63061ms step_avg:165.52ms
step:392/1530 train_loss:3.9307 train_time:63228ms step_avg:165.52ms
step:393/1530 train_loss:3.7701 train_time:63396ms step_avg:165.52ms
step:394/1530 train_loss:3.8859 train_time:63565ms step_avg:165.53ms
step:395/1530 train_loss:3.6401 train_time:63732ms step_avg:165.54ms
step:396/1530 train_loss:3.8438 train_time:63900ms step_avg:165.54ms
step:397/1530 train_loss:3.8668 train_time:64069ms step_avg:165.55ms
step:398/1530 train_loss:3.8935 train_time:64235ms step_avg:165.56ms
step:399/1530 train_loss:3.7801 train_time:64401ms step_avg:165.56ms
step:400/1530 train_loss:3.8299 train_time:64571ms step_avg:165.57ms
step:401/1530 train_loss:3.9254 train_time:64739ms step_avg:165.57ms
step:402/1530 train_loss:3.8526 train_time:64907ms step_avg:165.58ms
step:403/1530 train_loss:3.9704 train_time:65074ms step_avg:165.58ms
step:404/1530 train_loss:3.6868 train_time:65240ms step_avg:165.58ms
step:405/1530 train_loss:3.7926 train_time:65408ms step_avg:165.59ms
step:406/1530 train_loss:4.1026 train_time:65577ms step_avg:165.60ms
step:407/1530 train_loss:3.7842 train_time:65744ms step_avg:165.60ms
step:408/1530 train_loss:3.8247 train_time:65910ms step_avg:165.60ms
step:409/1530 train_loss:3.8635 train_time:66078ms step_avg:165.61ms
step:410/1530 train_loss:3.7638 train_time:66246ms step_avg:165.61ms
step:411/1530 train_loss:3.7690 train_time:66414ms step_avg:165.62ms
step:412/1530 train_loss:4.1850 train_time:66582ms step_avg:165.63ms
step:413/1530 train_loss:3.6524 train_time:66750ms step_avg:165.63ms
step:414/1530 train_loss:4.0229 train_time:66917ms step_avg:165.64ms
step:415/1530 train_loss:3.7614 train_time:67085ms step_avg:165.64ms
step:416/1530 train_loss:3.7732 train_time:67253ms step_avg:165.65ms
step:417/1530 train_loss:3.9627 train_time:67420ms step_avg:165.65ms
step:418/1530 train_loss:3.6946 train_time:67589ms step_avg:165.66ms
step:419/1530 train_loss:3.8044 train_time:67756ms step_avg:165.66ms
step:420/1530 train_loss:3.7105 train_time:67923ms step_avg:165.67ms
step:421/1530 train_loss:3.6569 train_time:68090ms step_avg:165.67ms
step:422/1530 train_loss:3.7900 train_time:68256ms step_avg:165.67ms
step:423/1530 train_loss:3.8809 train_time:68424ms step_avg:165.67ms
step:424/1530 train_loss:3.6204 train_time:68592ms step_avg:165.68ms
step:425/1530 train_loss:3.8063 train_time:68759ms step_avg:165.68ms
step:426/1530 train_loss:3.6581 train_time:68927ms step_avg:165.69ms
step:427/1530 train_loss:3.8970 train_time:69093ms step_avg:165.69ms
step:428/1530 train_loss:3.8120 train_time:69261ms step_avg:165.70ms
step:429/1530 train_loss:3.7674 train_time:69428ms step_avg:165.70ms
step:430/1530 train_loss:3.7123 train_time:69596ms step_avg:165.70ms
step:431/1530 train_loss:3.6397 train_time:69764ms step_avg:165.71ms
step:432/1530 train_loss:3.7676 train_time:69931ms step_avg:165.71ms
step:433/1530 train_loss:3.8200 train_time:70099ms step_avg:165.72ms
step:434/1530 train_loss:3.7807 train_time:70268ms step_avg:165.73ms
step:435/1530 train_loss:3.8186 train_time:70434ms step_avg:165.73ms
step:436/1530 train_loss:3.8299 train_time:70601ms step_avg:165.73ms
step:437/1530 train_loss:3.7428 train_time:70769ms step_avg:165.74ms
step:438/1530 train_loss:3.7105 train_time:70936ms step_avg:165.74ms
step:439/1530 train_loss:3.7174 train_time:71103ms step_avg:165.74ms
step:440/1530 train_loss:3.8943 train_time:71271ms step_avg:165.75ms
step:441/1530 train_loss:3.7682 train_time:71438ms step_avg:165.75ms
step:442/1530 train_loss:3.7471 train_time:71605ms step_avg:165.75ms
step:443/1530 train_loss:3.6317 train_time:71773ms step_avg:165.76ms
step:444/1530 train_loss:3.9235 train_time:71939ms step_avg:165.76ms
step:445/1530 train_loss:3.8535 train_time:72106ms step_avg:165.76ms
step:446/1530 train_loss:3.8438 train_time:72273ms step_avg:165.76ms
step:447/1530 train_loss:3.7573 train_time:72440ms step_avg:165.77ms
step:448/1530 train_loss:3.8588 train_time:72607ms step_avg:165.77ms
step:449/1530 train_loss:3.6997 train_time:72774ms step_avg:165.77ms
step:450/1530 train_loss:3.7280 train_time:72940ms step_avg:165.77ms
step:451/1530 train_loss:3.5920 train_time:73109ms step_avg:165.78ms
step:452/1530 train_loss:3.7163 train_time:73276ms step_avg:165.78ms
step:453/1530 train_loss:3.6714 train_time:73443ms step_avg:165.79ms
step:454/1530 train_loss:3.6490 train_time:73611ms step_avg:165.79ms
step:455/1530 train_loss:3.8469 train_time:73778ms step_avg:165.79ms
step:456/1530 train_loss:3.7339 train_time:73949ms step_avg:165.80ms
step:457/1530 train_loss:3.7886 train_time:74119ms step_avg:165.81ms
step:458/1530 train_loss:3.8347 train_time:74289ms step_avg:165.82ms
step:459/1530 train_loss:3.6422 train_time:74459ms step_avg:165.83ms
step:460/1530 train_loss:3.7997 train_time:74628ms step_avg:165.84ms
step:461/1530 train_loss:3.7022 train_time:74800ms step_avg:165.85ms
step:462/1530 train_loss:3.7453 train_time:74972ms step_avg:165.87ms
step:463/1530 train_loss:3.7836 train_time:75142ms step_avg:165.88ms
step:464/1530 train_loss:3.7216 train_time:75311ms step_avg:165.88ms
step:465/1530 train_loss:3.7189 train_time:75480ms step_avg:165.89ms
step:466/1530 train_loss:3.8064 train_time:75650ms step_avg:165.90ms
step:467/1530 train_loss:3.8313 train_time:75822ms step_avg:165.91ms
step:468/1530 train_loss:3.7968 train_time:75991ms step_avg:165.92ms
step:469/1530 train_loss:3.6981 train_time:76160ms step_avg:165.93ms
step:470/1530 train_loss:3.7785 train_time:76329ms step_avg:165.93ms
step:471/1530 train_loss:3.8152 train_time:76499ms step_avg:165.94ms
step:472/1530 train_loss:3.7982 train_time:76671ms step_avg:165.95ms
step:473/1530 train_loss:3.7244 train_time:76839ms step_avg:165.96ms
step:474/1530 train_loss:3.5979 train_time:77009ms step_avg:165.97ms
step:475/1530 train_loss:4.0271 train_time:77178ms step_avg:165.97ms
step:476/1530 train_loss:3.7608 train_time:77348ms step_avg:165.98ms
step:477/1530 train_loss:3.6014 train_time:77518ms step_avg:165.99ms
step:478/1530 train_loss:3.8306 train_time:77687ms step_avg:166.00ms
step:479/1530 train_loss:3.7765 train_time:77857ms step_avg:166.01ms
step:480/1530 train_loss:3.9232 train_time:78027ms step_avg:166.01ms
step:481/1530 train_loss:3.7348 train_time:78196ms step_avg:166.02ms
step:482/1530 train_loss:3.5322 train_time:78366ms step_avg:166.03ms
step:483/1530 train_loss:3.8110 train_time:78534ms step_avg:166.03ms
step:484/1530 train_loss:3.6679 train_time:78704ms step_avg:166.04ms
step:485/1530 train_loss:3.6667 train_time:78874ms step_avg:166.05ms
step:486/1530 train_loss:3.5741 train_time:79045ms step_avg:166.06ms
step:487/1530 train_loss:3.6928 train_time:79214ms step_avg:166.07ms
step:488/1530 train_loss:3.8858 train_time:79385ms step_avg:166.08ms
step:489/1530 train_loss:3.7163 train_time:79554ms step_avg:166.08ms
step:490/1530 train_loss:3.6000 train_time:79723ms step_avg:166.09ms
step:491/1530 train_loss:3.6218 train_time:79893ms step_avg:166.10ms
step:492/1530 train_loss:3.7407 train_time:80064ms step_avg:166.11ms
step:493/1530 train_loss:3.5791 train_time:80233ms step_avg:166.11ms
step:494/1530 train_loss:3.7033 train_time:80404ms step_avg:166.12ms
step:495/1530 train_loss:3.6667 train_time:80575ms step_avg:166.13ms
step:496/1530 train_loss:3.5186 train_time:80746ms step_avg:166.14ms
step:497/1530 train_loss:3.7418 train_time:80914ms step_avg:166.15ms
step:498/1530 train_loss:3.7957 train_time:81084ms step_avg:166.16ms
step:499/1530 train_loss:3.8295 train_time:81254ms step_avg:166.16ms
step:500/1530 train_loss:3.7440 train_time:81425ms step_avg:166.17ms
step:500/1530 val_loss:3.7113 train_time:81474ms step_avg:166.27ms
step:501/1530 train_loss:3.8103 train_time:81595ms step_avg:166.18ms
step:502/1530 train_loss:3.7583 train_time:81766ms step_avg:166.19ms
step:503/1530 train_loss:3.7803 train_time:81937ms step_avg:166.20ms
step:504/1530 train_loss:3.7258 train_time:82106ms step_avg:166.21ms
step:505/1530 train_loss:3.8117 train_time:82275ms step_avg:166.21ms
step:506/1530 train_loss:3.6615 train_time:82445ms step_avg:166.22ms
step:507/1530 train_loss:3.7704 train_time:82615ms step_avg:166.23ms
step:508/1530 train_loss:3.8295 train_time:82787ms step_avg:166.24ms
step:509/1530 train_loss:3.7809 train_time:82957ms step_avg:166.25ms
step:510/1530 train_loss:3.5918 train_time:83126ms step_avg:166.25ms
step:511/1530 train_loss:3.7840 train_time:83295ms step_avg:166.26ms
step:512/1530 train_loss:3.7248 train_time:83466ms step_avg:166.27ms
step:513/1530 train_loss:3.6734 train_time:83635ms step_avg:166.27ms
step:514/1530 train_loss:3.8269 train_time:83807ms step_avg:166.28ms
step:515/1530 train_loss:3.7443 train_time:83977ms step_avg:166.29ms
step:516/1530 train_loss:4.0829 train_time:84148ms step_avg:166.30ms
step:517/1530 train_loss:3.6955 train_time:84316ms step_avg:166.30ms
step:518/1530 train_loss:3.7772 train_time:84485ms step_avg:166.31ms
step:519/1530 train_loss:3.6633 train_time:84654ms step_avg:166.31ms
step:520/1530 train_loss:3.6862 train_time:84824ms step_avg:166.32ms
step:521/1530 train_loss:3.6693 train_time:84993ms step_avg:166.33ms
step:522/1530 train_loss:3.6667 train_time:85163ms step_avg:166.33ms
step:523/1530 train_loss:4.2961 train_time:85332ms step_avg:166.34ms
step:524/1530 train_loss:3.7424 train_time:85501ms step_avg:166.34ms
step:525/1530 train_loss:3.6930 train_time:85669ms step_avg:166.35ms
step:526/1530 train_loss:3.7040 train_time:85838ms step_avg:166.35ms
step:527/1530 train_loss:3.6617 train_time:86007ms step_avg:166.36ms
step:528/1530 train_loss:3.6308 train_time:86175ms step_avg:166.36ms
step:529/1530 train_loss:3.8558 train_time:86345ms step_avg:166.37ms
step:530/1530 train_loss:3.6572 train_time:86514ms step_avg:166.37ms
step:531/1530 train_loss:3.9212 train_time:86684ms step_avg:166.38ms
step:532/1530 train_loss:3.7397 train_time:86853ms step_avg:166.38ms
step:533/1530 train_loss:3.6632 train_time:87024ms step_avg:166.39ms
step:534/1530 train_loss:3.6779 train_time:87192ms step_avg:166.40ms
step:535/1530 train_loss:3.6158 train_time:87362ms step_avg:166.40ms
step:536/1530 train_loss:3.7567 train_time:87530ms step_avg:166.41ms
step:537/1530 train_loss:3.7308 train_time:87700ms step_avg:166.41ms
step:538/1530 train_loss:3.6314 train_time:87870ms step_avg:166.42ms
step:539/1530 train_loss:4.1172 train_time:88041ms step_avg:166.43ms
step:540/1530 train_loss:3.6820 train_time:88210ms step_avg:166.43ms
step:541/1530 train_loss:3.7946 train_time:88378ms step_avg:166.44ms
step:542/1530 train_loss:3.5974 train_time:88547ms step_avg:166.44ms
step:543/1530 train_loss:3.5909 train_time:88716ms step_avg:166.45ms
step:544/1530 train_loss:3.6428 train_time:88885ms step_avg:166.45ms
step:545/1530 train_loss:3.6013 train_time:89054ms step_avg:166.46ms
step:546/1530 train_loss:3.6362 train_time:89225ms step_avg:166.46ms
step:547/1530 train_loss:3.6453 train_time:89393ms step_avg:166.47ms
step:548/1530 train_loss:3.6110 train_time:89563ms step_avg:166.47ms
step:549/1530 train_loss:3.7308 train_time:89731ms step_avg:166.48ms
step:550/1530 train_loss:3.6274 train_time:89902ms step_avg:166.48ms
step:551/1530 train_loss:3.6347 train_time:90069ms step_avg:166.49ms
step:552/1530 train_loss:3.9343 train_time:90239ms step_avg:166.49ms
step:553/1530 train_loss:3.7644 train_time:90408ms step_avg:166.50ms
step:554/1530 train_loss:3.7131 train_time:90576ms step_avg:166.50ms
step:555/1530 train_loss:3.6353 train_time:90746ms step_avg:166.51ms
step:556/1530 train_loss:3.7047 train_time:90915ms step_avg:166.51ms
step:557/1530 train_loss:3.3306 train_time:91085ms step_avg:166.52ms
step:558/1530 train_loss:3.6165 train_time:91253ms step_avg:166.52ms
step:559/1530 train_loss:3.6578 train_time:91422ms step_avg:166.52ms
step:560/1530 train_loss:3.6949 train_time:91590ms step_avg:166.53ms
step:561/1530 train_loss:3.6150 train_time:91758ms step_avg:166.53ms
step:562/1530 train_loss:3.5571 train_time:91927ms step_avg:166.53ms
step:563/1530 train_loss:3.7664 train_time:92096ms step_avg:166.54ms
step:564/1530 train_loss:3.5805 train_time:92265ms step_avg:166.54ms
step:565/1530 train_loss:3.6853 train_time:92433ms step_avg:166.55ms
step:566/1530 train_loss:3.6307 train_time:92738ms step_avg:166.79ms
step:567/1530 train_loss:3.6075 train_time:92915ms step_avg:166.81ms
step:568/1530 train_loss:3.6879 train_time:93085ms step_avg:166.82ms
step:569/1530 train_loss:3.6526 train_time:93417ms step_avg:167.12ms
step:570/1530 train_loss:3.6989 train_time:93588ms step_avg:167.12ms
step:571/1530 train_loss:3.7672 train_time:93758ms step_avg:167.13ms
step:572/1530 train_loss:3.7297 train_time:93930ms step_avg:167.14ms
step:573/1530 train_loss:3.7409 train_time:94103ms step_avg:167.15ms
step:574/1530 train_loss:3.7818 train_time:94276ms step_avg:167.16ms
step:575/1530 train_loss:3.7382 train_time:94448ms step_avg:167.16ms
step:576/1530 train_loss:3.7652 train_time:94620ms step_avg:167.17ms
step:577/1530 train_loss:3.6732 train_time:94791ms step_avg:167.18ms
step:578/1530 train_loss:3.6853 train_time:94964ms step_avg:167.19ms
step:579/1530 train_loss:3.6780 train_time:95134ms step_avg:167.20ms
step:580/1530 train_loss:3.5925 train_time:95305ms step_avg:167.20ms
step:581/1530 train_loss:3.6475 train_time:95475ms step_avg:167.21ms
step:582/1530 train_loss:3.8557 train_time:95647ms step_avg:167.21ms
step:583/1530 train_loss:3.6345 train_time:95820ms step_avg:167.22ms
step:584/1530 train_loss:3.6000 train_time:95991ms step_avg:167.23ms
step:585/1530 train_loss:3.7922 train_time:96162ms step_avg:167.24ms
step:586/1530 train_loss:3.5268 train_time:96333ms step_avg:167.24ms
step:587/1530 train_loss:3.6790 train_time:96504ms step_avg:167.25ms
step:588/1530 train_loss:3.6474 train_time:96674ms step_avg:167.26ms
step:589/1530 train_loss:4.0022 train_time:96846ms step_avg:167.26ms
step:590/1530 train_loss:3.7867 train_time:97018ms step_avg:167.27ms
step:591/1530 train_loss:3.5171 train_time:97189ms step_avg:167.28ms
step:592/1530 train_loss:3.5444 train_time:97362ms step_avg:167.29ms
step:593/1530 train_loss:3.5130 train_time:97534ms step_avg:167.30ms
step:594/1530 train_loss:3.5658 train_time:97706ms step_avg:167.31ms
step:595/1530 train_loss:3.9162 train_time:97880ms step_avg:167.32ms
step:596/1530 train_loss:3.6563 train_time:98052ms step_avg:167.32ms
step:597/1530 train_loss:3.5981 train_time:98224ms step_avg:167.33ms
step:598/1530 train_loss:3.6641 train_time:98393ms step_avg:167.34ms
step:599/1530 train_loss:3.4840 train_time:98564ms step_avg:167.34ms
step:600/1530 train_loss:3.6014 train_time:98735ms step_avg:167.35ms
step:601/1530 train_loss:3.6546 train_time:98908ms step_avg:167.36ms
step:602/1530 train_loss:3.6711 train_time:99081ms step_avg:167.37ms
step:603/1530 train_loss:3.7846 train_time:99252ms step_avg:167.37ms
step:604/1530 train_loss:3.6210 train_time:99424ms step_avg:167.38ms
step:605/1530 train_loss:3.6199 train_time:99596ms step_avg:167.39ms
step:606/1530 train_loss:3.5802 train_time:99769ms step_avg:167.40ms
step:607/1530 train_loss:3.8469 train_time:99941ms step_avg:167.40ms
step:608/1530 train_loss:3.6390 train_time:100112ms step_avg:167.41ms
step:609/1530 train_loss:3.6226 train_time:100285ms step_avg:167.42ms
step:610/1530 train_loss:3.7084 train_time:100453ms step_avg:167.42ms
step:611/1530 train_loss:3.6048 train_time:100626ms step_avg:167.43ms
step:612/1530 train_loss:3.5753 train_time:100796ms step_avg:167.43ms
step:613/1530 train_loss:3.7689 train_time:100967ms step_avg:167.44ms
step:614/1530 train_loss:3.7124 train_time:101137ms step_avg:167.45ms
step:615/1530 train_loss:3.7146 train_time:101308ms step_avg:167.45ms
step:616/1530 train_loss:3.6425 train_time:101478ms step_avg:167.45ms
step:617/1530 train_loss:3.5666 train_time:101650ms step_avg:167.46ms
step:618/1530 train_loss:3.6981 train_time:101822ms step_avg:167.47ms
step:619/1530 train_loss:3.5631 train_time:101992ms step_avg:167.47ms
step:620/1530 train_loss:3.5969 train_time:102163ms step_avg:167.48ms
step:621/1530 train_loss:3.9269 train_time:102334ms step_avg:167.49ms
step:622/1530 train_loss:3.5812 train_time:102507ms step_avg:167.50ms
step:623/1530 train_loss:3.6083 train_time:102680ms step_avg:167.50ms
step:624/1530 train_loss:3.7042 train_time:102850ms step_avg:167.51ms
step:625/1530 train_loss:3.7065 train_time:103022ms step_avg:167.51ms
step:625/1530 val_loss:3.6289 train_time:103071ms step_avg:167.59ms
step:626/1530 train_loss:3.7428 train_time:103195ms step_avg:167.52ms
step:627/1530 train_loss:3.7205 train_time:103367ms step_avg:167.53ms
step:628/1530 train_loss:3.7713 train_time:103537ms step_avg:167.54ms
step:629/1530 train_loss:3.6017 train_time:103708ms step_avg:167.54ms
step:630/1530 train_loss:3.7308 train_time:103878ms step_avg:167.54ms
step:631/1530 train_loss:3.7486 train_time:104048ms step_avg:167.55ms
step:632/1530 train_loss:3.6591 train_time:104219ms step_avg:167.55ms
step:633/1530 train_loss:3.6128 train_time:104391ms step_avg:167.56ms
step:634/1530 train_loss:3.7014 train_time:104561ms step_avg:167.57ms
step:635/1530 train_loss:3.9506 train_time:104731ms step_avg:167.57ms
step:636/1530 train_loss:3.5573 train_time:104901ms step_avg:167.57ms
step:637/1530 train_loss:3.3595 train_time:105074ms step_avg:167.58ms
step:638/1530 train_loss:3.5965 train_time:105244ms step_avg:167.59ms
step:639/1530 train_loss:3.6392 train_time:105415ms step_avg:167.59ms
step:640/1530 train_loss:3.5852 train_time:105584ms step_avg:167.59ms
step:641/1530 train_loss:3.5987 train_time:105755ms step_avg:167.60ms
step:642/1530 train_loss:3.6363 train_time:105926ms step_avg:167.60ms
step:643/1530 train_loss:3.6062 train_time:106097ms step_avg:167.61ms
step:644/1530 train_loss:3.5731 train_time:106266ms step_avg:167.61ms
step:645/1530 train_loss:3.7860 train_time:106437ms step_avg:167.62ms
step:646/1530 train_loss:3.6821 train_time:106608ms step_avg:167.62ms
step:647/1530 train_loss:3.6719 train_time:106778ms step_avg:167.63ms
step:648/1530 train_loss:3.7167 train_time:106950ms step_avg:167.63ms
step:649/1530 train_loss:3.7731 train_time:107120ms step_avg:167.64ms
step:650/1530 train_loss:3.6297 train_time:107291ms step_avg:167.64ms
step:651/1530 train_loss:3.7698 train_time:107462ms step_avg:167.65ms
step:652/1530 train_loss:3.5907 train_time:107633ms step_avg:167.65ms
step:653/1530 train_loss:3.6688 train_time:107801ms step_avg:167.65ms
step:654/1530 train_loss:3.4294 train_time:107975ms step_avg:167.66ms
step:655/1530 train_loss:3.5905 train_time:108145ms step_avg:167.67ms
step:656/1530 train_loss:3.5808 train_time:108315ms step_avg:167.67ms
step:657/1530 train_loss:3.5035 train_time:108485ms step_avg:167.67ms
step:658/1530 train_loss:3.6949 train_time:108657ms step_avg:167.68ms
step:659/1530 train_loss:3.5912 train_time:108829ms step_avg:167.69ms
step:660/1530 train_loss:3.6920 train_time:109000ms step_avg:167.69ms
step:661/1530 train_loss:3.7617 train_time:109172ms step_avg:167.70ms
step:662/1530 train_loss:3.6770 train_time:109342ms step_avg:167.70ms
step:663/1530 train_loss:3.5636 train_time:109512ms step_avg:167.71ms
step:664/1530 train_loss:3.6207 train_time:109681ms step_avg:167.71ms
step:665/1530 train_loss:3.4958 train_time:109854ms step_avg:167.72ms
step:666/1530 train_loss:3.7882 train_time:110024ms step_avg:167.72ms
step:667/1530 train_loss:3.6113 train_time:110196ms step_avg:167.73ms
step:668/1530 train_loss:3.6492 train_time:110366ms step_avg:167.73ms
step:669/1530 train_loss:3.5031 train_time:110538ms step_avg:167.74ms
step:670/1530 train_loss:3.6050 train_time:110708ms step_avg:167.74ms
step:671/1530 train_loss:3.5693 train_time:110878ms step_avg:167.74ms
step:672/1530 train_loss:3.5715 train_time:111050ms step_avg:167.75ms
step:673/1530 train_loss:3.8599 train_time:111220ms step_avg:167.75ms
step:674/1530 train_loss:3.6362 train_time:111392ms step_avg:167.76ms
step:675/1530 train_loss:3.7168 train_time:111563ms step_avg:167.76ms
step:676/1530 train_loss:3.4967 train_time:111734ms step_avg:167.77ms
step:677/1530 train_loss:3.6078 train_time:111905ms step_avg:167.77ms
step:678/1530 train_loss:3.5608 train_time:112076ms step_avg:167.78ms
step:679/1530 train_loss:3.6869 train_time:112248ms step_avg:167.78ms
step:680/1530 train_loss:3.5904 train_time:112418ms step_avg:167.79ms
step:681/1530 train_loss:3.6219 train_time:112590ms step_avg:167.79ms
step:682/1530 train_loss:3.6703 train_time:112764ms step_avg:167.80ms
step:683/1530 train_loss:3.7462 train_time:112938ms step_avg:167.81ms
step:684/1530 train_loss:3.6556 train_time:113108ms step_avg:167.82ms
step:685/1530 train_loss:3.6946 train_time:113284ms step_avg:167.83ms
step:686/1530 train_loss:3.6461 train_time:113459ms step_avg:167.84ms
step:687/1530 train_loss:3.6759 train_time:113631ms step_avg:167.84ms
step:688/1530 train_loss:3.2298 train_time:113806ms step_avg:167.86ms
step:689/1530 train_loss:3.4101 train_time:113980ms step_avg:167.86ms
step:690/1530 train_loss:3.5477 train_time:114155ms step_avg:167.88ms
step:691/1530 train_loss:3.4161 train_time:114328ms step_avg:167.88ms
step:692/1530 train_loss:3.6351 train_time:114499ms step_avg:167.89ms
step:693/1530 train_loss:3.6555 train_time:114672ms step_avg:167.89ms
step:694/1530 train_loss:3.5603 train_time:114845ms step_avg:167.90ms
step:695/1530 train_loss:3.5402 train_time:115017ms step_avg:167.91ms
step:696/1530 train_loss:3.8579 train_time:115190ms step_avg:167.92ms
step:697/1530 train_loss:3.5951 train_time:115363ms step_avg:167.92ms
step:698/1530 train_loss:3.6469 train_time:115535ms step_avg:167.93ms
step:699/1530 train_loss:3.7718 train_time:115708ms step_avg:167.94ms
step:700/1530 train_loss:3.5746 train_time:115881ms step_avg:167.94ms
step:701/1530 train_loss:3.5513 train_time:116055ms step_avg:167.95ms
step:702/1530 train_loss:3.5202 train_time:116226ms step_avg:167.96ms
step:703/1530 train_loss:3.5030 train_time:116397ms step_avg:167.96ms
step:704/1530 train_loss:3.5763 train_time:116570ms step_avg:167.97ms
step:705/1530 train_loss:3.5698 train_time:116746ms step_avg:167.98ms
step:706/1530 train_loss:3.5876 train_time:116922ms step_avg:167.99ms
step:707/1530 train_loss:3.6480 train_time:117097ms step_avg:168.00ms
step:708/1530 train_loss:3.6031 train_time:117269ms step_avg:168.01ms
step:709/1530 train_loss:3.5893 train_time:117442ms step_avg:168.01ms
step:710/1530 train_loss:3.5490 train_time:117615ms step_avg:168.02ms
step:711/1530 train_loss:3.6002 train_time:117788ms step_avg:168.03ms
step:712/1530 train_loss:3.6578 train_time:117963ms step_avg:168.04ms
step:713/1530 train_loss:3.6578 train_time:118140ms step_avg:168.05ms
step:714/1530 train_loss:3.5668 train_time:118312ms step_avg:168.06ms
step:715/1530 train_loss:3.5761 train_time:118485ms step_avg:168.06ms
step:716/1530 train_loss:3.5952 train_time:118658ms step_avg:168.07ms
step:717/1530 train_loss:3.7053 train_time:118833ms step_avg:168.08ms
step:718/1530 train_loss:3.6059 train_time:119004ms step_avg:168.08ms
step:719/1530 train_loss:3.6835 train_time:119176ms step_avg:168.09ms
step:720/1530 train_loss:3.8525 train_time:119353ms step_avg:168.10ms
step:721/1530 train_loss:3.4752 train_time:119526ms step_avg:168.11ms
step:722/1530 train_loss:3.7476 train_time:119699ms step_avg:168.12ms
step:723/1530 train_loss:3.7743 train_time:119870ms step_avg:168.12ms
step:724/1530 train_loss:3.5714 train_time:120043ms step_avg:168.13ms
step:725/1530 train_loss:3.6584 train_time:120217ms step_avg:168.14ms
step:726/1530 train_loss:3.5334 train_time:120389ms step_avg:168.14ms
step:727/1530 train_loss:3.5865 train_time:120564ms step_avg:168.15ms
step:728/1530 train_loss:3.7332 train_time:120738ms step_avg:168.16ms
step:729/1530 train_loss:3.6730 train_time:120911ms step_avg:168.17ms
step:730/1530 train_loss:3.6681 train_time:121084ms step_avg:168.17ms
step:731/1530 train_loss:3.5658 train_time:121257ms step_avg:168.18ms
step:732/1530 train_loss:3.5995 train_time:121430ms step_avg:168.19ms
step:733/1530 train_loss:3.8351 train_time:121603ms step_avg:168.19ms
step:734/1530 train_loss:3.5703 train_time:121778ms step_avg:168.20ms
step:735/1530 train_loss:3.6295 train_time:121951ms step_avg:168.21ms
step:736/1530 train_loss:3.7400 train_time:122124ms step_avg:168.22ms
step:737/1530 train_loss:3.6825 train_time:122298ms step_avg:168.22ms
step:738/1530 train_loss:3.6145 train_time:122470ms step_avg:168.23ms
step:739/1530 train_loss:3.5053 train_time:122641ms step_avg:168.23ms
step:740/1530 train_loss:4.1205 train_time:122820ms step_avg:168.25ms
step:741/1530 train_loss:3.4951 train_time:122993ms step_avg:168.25ms
step:742/1530 train_loss:3.5531 train_time:123164ms step_avg:168.26ms
step:743/1530 train_loss:3.5905 train_time:123337ms step_avg:168.26ms
step:744/1530 train_loss:3.6468 train_time:123510ms step_avg:168.27ms
step:745/1530 train_loss:3.5866 train_time:123682ms step_avg:168.27ms
step:746/1530 train_loss:3.6009 train_time:123857ms step_avg:168.28ms
step:747/1530 train_loss:3.6489 train_time:124028ms step_avg:168.29ms
step:748/1530 train_loss:3.5720 train_time:124206ms step_avg:168.30ms
step:749/1530 train_loss:3.5687 train_time:124379ms step_avg:168.31ms
step:750/1530 train_loss:3.6037 train_time:124552ms step_avg:168.31ms
step:750/1530 val_loss:3.5730 train_time:124601ms step_avg:168.38ms
step:751/1530 train_loss:3.5793 train_time:124725ms step_avg:168.32ms
step:752/1530 train_loss:3.6193 train_time:124899ms step_avg:168.33ms
step:753/1530 train_loss:3.6293 train_time:125074ms step_avg:168.34ms
step:754/1530 train_loss:3.5954 train_time:125246ms step_avg:168.34ms
step:755/1530 train_loss:3.6931 train_time:125553ms step_avg:168.53ms
step:756/1530 train_loss:3.4716 train_time:125739ms step_avg:168.55ms
step:757/1530 train_loss:3.7389 train_time:125909ms step_avg:168.55ms
step:758/1530 train_loss:3.6572 train_time:126082ms step_avg:168.56ms
step:759/1530 train_loss:3.5925 train_time:126416ms step_avg:168.78ms
step:760/1530 train_loss:3.7139 train_time:126586ms step_avg:168.78ms
step:761/1530 train_loss:3.4076 train_time:126759ms step_avg:168.79ms
step:762/1530 train_loss:3.5538 train_time:126931ms step_avg:168.79ms
step:763/1530 train_loss:3.6675 train_time:127104ms step_avg:168.80ms
step:764/1530 train_loss:3.3283 train_time:127278ms step_avg:168.80ms
step:765/1530 train_loss:3.7383 train_time:127449ms step_avg:168.81ms
step:766/1530 train_loss:3.5776 train_time:127622ms step_avg:168.81ms
step:767/1530 train_loss:3.5733 train_time:127793ms step_avg:168.81ms
step:768/1530 train_loss:3.5780 train_time:127966ms step_avg:168.82ms
step:769/1530 train_loss:3.5887 train_time:128141ms step_avg:168.83ms
step:770/1530 train_loss:3.6455 train_time:128312ms step_avg:168.83ms
step:771/1530 train_loss:3.8947 train_time:128484ms step_avg:168.84ms
step:772/1530 train_loss:3.4601 train_time:128656ms step_avg:168.84ms
step:773/1530 train_loss:3.6384 train_time:128826ms step_avg:168.84ms
step:774/1530 train_loss:3.6441 train_time:128998ms step_avg:168.85ms
step:775/1530 train_loss:3.6149 train_time:129170ms step_avg:168.85ms
step:776/1530 train_loss:3.4078 train_time:129344ms step_avg:168.86ms
step:777/1530 train_loss:3.3933 train_time:129518ms step_avg:168.86ms
step:778/1530 train_loss:3.4961 train_time:129689ms step_avg:168.87ms
step:779/1530 train_loss:3.5844 train_time:129861ms step_avg:168.87ms
step:780/1530 train_loss:3.5927 train_time:130036ms step_avg:168.88ms
step:781/1530 train_loss:3.6782 train_time:130208ms step_avg:168.88ms
step:782/1530 train_loss:3.5894 train_time:130380ms step_avg:168.89ms
step:783/1530 train_loss:3.5734 train_time:130552ms step_avg:168.89ms
step:784/1530 train_loss:3.6117 train_time:130723ms step_avg:168.89ms
step:785/1530 train_loss:3.5653 train_time:130894ms step_avg:168.90ms
step:786/1530 train_loss:3.4444 train_time:131066ms step_avg:168.90ms
step:787/1530 train_loss:3.7154 train_time:131238ms step_avg:168.90ms
step:788/1530 train_loss:3.5056 train_time:131411ms step_avg:168.91ms
step:789/1530 train_loss:3.5554 train_time:131583ms step_avg:168.91ms
step:790/1530 train_loss:3.6347 train_time:131757ms step_avg:168.92ms
step:791/1530 train_loss:3.7794 train_time:131933ms step_avg:168.93ms
step:792/1530 train_loss:3.7739 train_time:132104ms step_avg:168.93ms
step:793/1530 train_loss:3.4523 train_time:132277ms step_avg:168.94ms
step:794/1530 train_loss:3.6008 train_time:132450ms step_avg:168.94ms
step:795/1530 train_loss:3.6792 train_time:132624ms step_avg:168.95ms
step:796/1530 train_loss:3.7666 train_time:132801ms step_avg:168.96ms
step:797/1530 train_loss:3.5324 train_time:132976ms step_avg:168.97ms
step:798/1530 train_loss:3.6509 train_time:133149ms step_avg:168.97ms
step:799/1530 train_loss:3.5388 train_time:133326ms step_avg:168.98ms
step:800/1530 train_loss:3.5336 train_time:133499ms step_avg:168.99ms
step:801/1530 train_loss:3.6336 train_time:133674ms step_avg:168.99ms
step:802/1530 train_loss:3.5012 train_time:133850ms step_avg:169.00ms
step:803/1530 train_loss:3.4938 train_time:134023ms step_avg:169.01ms
step:804/1530 train_loss:3.6233 train_time:134198ms step_avg:169.01ms
step:805/1530 train_loss:3.5259 train_time:134372ms step_avg:169.02ms
step:806/1530 train_loss:3.5668 train_time:134546ms step_avg:169.03ms
step:807/1530 train_loss:3.6453 train_time:134720ms step_avg:169.03ms
step:808/1530 train_loss:3.5486 train_time:134897ms step_avg:169.04ms
step:809/1530 train_loss:3.5003 train_time:135069ms step_avg:169.05ms
step:810/1530 train_loss:3.5646 train_time:135244ms step_avg:169.05ms
step:811/1530 train_loss:3.5816 train_time:135419ms step_avg:169.06ms
step:812/1530 train_loss:3.6050 train_time:135591ms step_avg:169.07ms
step:813/1530 train_loss:3.6297 train_time:135762ms step_avg:169.07ms
step:814/1530 train_loss:3.5693 train_time:135938ms step_avg:169.08ms
step:815/1530 train_loss:3.5693 train_time:136110ms step_avg:169.08ms
step:816/1530 train_loss:3.6860 train_time:136285ms step_avg:169.09ms
step:817/1530 train_loss:3.7680 train_time:136460ms step_avg:169.10ms
step:818/1530 train_loss:3.5278 train_time:136632ms step_avg:169.10ms
step:819/1530 train_loss:3.7239 train_time:136807ms step_avg:169.11ms
step:820/1530 train_loss:3.4974 train_time:136983ms step_avg:169.12ms
step:821/1530 train_loss:3.5685 train_time:137157ms step_avg:169.12ms
step:822/1530 train_loss:3.7051 train_time:137333ms step_avg:169.13ms
step:823/1530 train_loss:3.5863 train_time:137507ms step_avg:169.13ms
step:824/1530 train_loss:3.5182 train_time:137681ms step_avg:169.14ms
step:825/1530 train_loss:3.6237 train_time:137858ms step_avg:169.15ms
step:826/1530 train_loss:3.4865 train_time:138034ms step_avg:169.16ms
step:827/1530 train_loss:3.7381 train_time:138208ms step_avg:169.17ms
step:828/1530 train_loss:3.6203 train_time:138381ms step_avg:169.17ms
step:829/1530 train_loss:3.6395 train_time:138558ms step_avg:169.18ms
step:830/1530 train_loss:3.5424 train_time:138733ms step_avg:169.19ms
step:831/1530 train_loss:3.6096 train_time:138907ms step_avg:169.19ms
step:832/1530 train_loss:3.5216 train_time:139082ms step_avg:169.20ms
step:833/1530 train_loss:3.6587 train_time:139259ms step_avg:169.21ms
step:834/1530 train_loss:3.4788 train_time:139431ms step_avg:169.21ms
step:835/1530 train_loss:3.4634 train_time:139605ms step_avg:169.22ms
step:836/1530 train_loss:3.7159 train_time:139781ms step_avg:169.23ms
step:837/1530 train_loss:3.4064 train_time:139955ms step_avg:169.23ms
step:838/1530 train_loss:3.6029 train_time:140128ms step_avg:169.24ms
step:839/1530 train_loss:3.4290 train_time:140303ms step_avg:169.24ms
step:840/1530 train_loss:3.4757 train_time:140477ms step_avg:169.25ms
step:841/1530 train_loss:3.5788 train_time:140649ms step_avg:169.25ms
step:842/1530 train_loss:3.5914 train_time:140824ms step_avg:169.26ms
step:843/1530 train_loss:3.5660 train_time:140996ms step_avg:169.26ms
step:844/1530 train_loss:3.4348 train_time:141168ms step_avg:169.27ms
step:845/1530 train_loss:3.6698 train_time:141341ms step_avg:169.27ms
step:846/1530 train_loss:3.5252 train_time:141517ms step_avg:169.28ms
step:847/1530 train_loss:3.5020 train_time:141691ms step_avg:169.28ms
step:848/1530 train_loss:3.6502 train_time:141865ms step_avg:169.29ms
step:849/1530 train_loss:3.5004 train_time:142040ms step_avg:169.30ms
step:850/1530 train_loss:3.4482 train_time:142212ms step_avg:169.30ms
step:851/1530 train_loss:3.7335 train_time:142385ms step_avg:169.30ms
step:852/1530 train_loss:3.4440 train_time:142559ms step_avg:169.31ms
step:853/1530 train_loss:3.5681 train_time:142731ms step_avg:169.31ms
step:854/1530 train_loss:3.6664 train_time:142905ms step_avg:169.32ms
step:855/1530 train_loss:3.5250 train_time:143079ms step_avg:169.32ms
step:856/1530 train_loss:3.5526 train_time:143254ms step_avg:169.33ms
step:857/1530 train_loss:3.6094 train_time:143427ms step_avg:169.34ms
step:858/1530 train_loss:3.4767 train_time:143603ms step_avg:169.34ms
step:859/1530 train_loss:3.5701 train_time:143778ms step_avg:169.35ms
step:860/1530 train_loss:3.5920 train_time:143950ms step_avg:169.35ms
step:861/1530 train_loss:3.6351 train_time:144126ms step_avg:169.36ms
step:862/1530 train_loss:3.6081 train_time:144304ms step_avg:169.37ms
step:863/1530 train_loss:3.5769 train_time:144481ms step_avg:169.38ms
step:864/1530 train_loss:3.3917 train_time:144654ms step_avg:169.38ms
step:865/1530 train_loss:3.6029 train_time:144825ms step_avg:169.39ms
step:866/1530 train_loss:3.8894 train_time:145003ms step_avg:169.40ms
step:867/1530 train_loss:3.4629 train_time:145177ms step_avg:169.40ms
step:868/1530 train_loss:3.6537 train_time:145348ms step_avg:169.40ms
step:869/1530 train_loss:3.6228 train_time:145521ms step_avg:169.41ms
step:870/1530 train_loss:3.4529 train_time:145697ms step_avg:169.42ms
step:871/1530 train_loss:3.4018 train_time:145870ms step_avg:169.42ms
step:872/1530 train_loss:3.6553 train_time:146045ms step_avg:169.43ms
step:873/1530 train_loss:3.4671 train_time:146220ms step_avg:169.43ms
step:874/1530 train_loss:3.2228 train_time:146399ms step_avg:169.44ms
step:875/1530 train_loss:3.6382 train_time:146574ms step_avg:169.45ms
step:875/1530 val_loss:3.5265 train_time:146622ms step_avg:169.51ms
step:876/1530 train_loss:3.4462 train_time:146747ms step_avg:169.45ms
step:877/1530 train_loss:3.6286 train_time:146923ms step_avg:169.46ms
step:878/1530 train_loss:3.4718 train_time:147098ms step_avg:169.47ms
step:879/1530 train_loss:3.6533 train_time:147272ms step_avg:169.47ms
step:880/1530 train_loss:3.3203 train_time:147443ms step_avg:169.47ms
step:881/1530 train_loss:3.4831 train_time:147615ms step_avg:169.48ms
step:882/1530 train_loss:3.7043 train_time:147788ms step_avg:169.48ms
step:883/1530 train_loss:3.8464 train_time:147961ms step_avg:169.49ms
step:884/1530 train_loss:3.5745 train_time:148135ms step_avg:169.49ms
step:885/1530 train_loss:3.5022 train_time:148308ms step_avg:169.50ms
step:886/1530 train_loss:3.5745 train_time:148481ms step_avg:169.50ms
step:887/1530 train_loss:4.0940 train_time:148656ms step_avg:169.51ms
step:888/1530 train_loss:3.8466 train_time:148836ms step_avg:169.52ms
step:889/1530 train_loss:3.5299 train_time:149009ms step_avg:169.52ms
step:890/1530 train_loss:3.5399 train_time:149181ms step_avg:169.52ms
step:891/1530 train_loss:3.3671 train_time:149355ms step_avg:169.53ms
step:892/1530 train_loss:3.7198 train_time:149528ms step_avg:169.53ms
step:893/1530 train_loss:3.4251 train_time:149700ms step_avg:169.54ms
step:894/1530 train_loss:3.6388 train_time:149877ms step_avg:169.54ms
step:895/1530 train_loss:3.6834 train_time:150052ms step_avg:169.55ms
step:896/1530 train_loss:3.5005 train_time:150224ms step_avg:169.55ms
step:897/1530 train_loss:3.5474 train_time:150400ms step_avg:169.56ms
step:898/1530 train_loss:3.5988 train_time:150576ms step_avg:169.57ms
step:899/1530 train_loss:3.4817 train_time:150748ms step_avg:169.57ms
step:900/1530 train_loss:3.4343 train_time:150921ms step_avg:169.57ms
step:901/1530 train_loss:3.6250 train_time:151095ms step_avg:169.58ms
step:902/1530 train_loss:3.6390 train_time:151268ms step_avg:169.58ms
step:903/1530 train_loss:3.5441 train_time:151443ms step_avg:169.59ms
step:904/1530 train_loss:3.4944 train_time:151615ms step_avg:169.59ms
step:905/1530 train_loss:3.5032 train_time:151787ms step_avg:169.59ms
step:906/1530 train_loss:3.7090 train_time:151961ms step_avg:169.60ms
step:907/1530 train_loss:3.5203 train_time:152135ms step_avg:169.60ms
step:908/1530 train_loss:3.5693 train_time:152310ms step_avg:169.61ms
step:909/1530 train_loss:3.4613 train_time:152486ms step_avg:169.62ms
step:910/1530 train_loss:3.5360 train_time:152667ms step_avg:169.63ms
step:911/1530 train_loss:3.6526 train_time:152843ms step_avg:169.64ms
step:912/1530 train_loss:3.6025 train_time:153021ms step_avg:169.65ms
step:913/1530 train_loss:3.4703 train_time:153199ms step_avg:169.66ms
step:914/1530 train_loss:3.7456 train_time:153378ms step_avg:169.67ms
step:915/1530 train_loss:3.5354 train_time:153557ms step_avg:169.68ms
step:916/1530 train_loss:3.6272 train_time:153732ms step_avg:169.68ms
step:917/1530 train_loss:3.6072 train_time:153906ms step_avg:169.69ms
step:918/1530 train_loss:4.8334 train_time:154084ms step_avg:169.70ms
step:919/1530 train_loss:3.5036 train_time:154263ms step_avg:169.71ms
step:920/1530 train_loss:3.5944 train_time:154440ms step_avg:169.71ms
step:921/1530 train_loss:3.5524 train_time:154618ms step_avg:169.72ms
step:922/1530 train_loss:3.5842 train_time:154795ms step_avg:169.73ms
step:923/1530 train_loss:3.6176 train_time:154971ms step_avg:169.74ms
step:924/1530 train_loss:3.6806 train_time:155149ms step_avg:169.75ms
step:925/1530 train_loss:3.6561 train_time:155322ms step_avg:169.75ms
step:926/1530 train_loss:3.5605 train_time:155496ms step_avg:169.76ms
step:927/1530 train_loss:3.5616 train_time:155673ms step_avg:169.76ms
step:928/1530 train_loss:3.7863 train_time:155850ms step_avg:169.77ms
step:929/1530 train_loss:3.6153 train_time:156023ms step_avg:169.78ms
step:930/1530 train_loss:3.4080 train_time:156199ms step_avg:169.78ms
step:931/1530 train_loss:3.5013 train_time:156374ms step_avg:169.79ms
step:932/1530 train_loss:3.6564 train_time:156553ms step_avg:169.80ms
step:933/1530 train_loss:3.3680 train_time:156729ms step_avg:169.80ms
step:934/1530 train_loss:3.5914 train_time:156909ms step_avg:169.82ms
step:935/1530 train_loss:3.4458 train_time:157087ms step_avg:169.82ms
step:936/1530 train_loss:3.5238 train_time:157264ms step_avg:169.83ms
step:937/1530 train_loss:3.6279 train_time:157441ms step_avg:169.84ms
step:938/1530 train_loss:3.5456 train_time:157617ms step_avg:169.85ms
step:939/1530 train_loss:3.6782 train_time:157798ms step_avg:169.86ms
step:940/1530 train_loss:3.4859 train_time:157974ms step_avg:169.86ms
step:941/1530 train_loss:3.5594 train_time:158152ms step_avg:169.87ms
step:942/1530 train_loss:3.3644 train_time:158328ms step_avg:169.88ms
step:943/1530 train_loss:3.7175 train_time:158507ms step_avg:169.89ms
step:944/1530 train_loss:3.4057 train_time:158822ms step_avg:170.05ms
step:945/1530 train_loss:3.4344 train_time:159005ms step_avg:170.06ms
step:946/1530 train_loss:5.0746 train_time:159186ms step_avg:170.07ms
step:947/1530 train_loss:3.6041 train_time:159363ms step_avg:170.08ms
step:948/1530 train_loss:3.4900 train_time:159538ms step_avg:170.08ms
step:949/1530 train_loss:3.3847 train_time:159870ms step_avg:170.26ms
step:950/1530 train_loss:3.4514 train_time:160044ms step_avg:170.26ms
step:951/1530 train_loss:3.4186 train_time:160222ms step_avg:170.27ms
step:952/1530 train_loss:3.4862 train_time:160398ms step_avg:170.27ms
step:953/1530 train_loss:3.5698 train_time:160576ms step_avg:170.28ms
step:954/1530 train_loss:3.4505 train_time:160755ms step_avg:170.29ms
step:955/1530 train_loss:3.4844 train_time:160930ms step_avg:170.30ms
step:956/1530 train_loss:3.4492 train_time:161105ms step_avg:170.30ms
step:957/1530 train_loss:3.5049 train_time:161285ms step_avg:170.31ms
step:958/1530 train_loss:3.5076 train_time:161465ms step_avg:170.32ms
step:959/1530 train_loss:3.5154 train_time:161640ms step_avg:170.33ms
step:960/1530 train_loss:3.4120 train_time:161819ms step_avg:170.34ms
step:961/1530 train_loss:3.6538 train_time:161994ms step_avg:170.34ms
step:962/1530 train_loss:3.5966 train_time:162170ms step_avg:170.35ms
step:963/1530 train_loss:3.6354 train_time:162346ms step_avg:170.35ms
step:964/1530 train_loss:3.4330 train_time:162525ms step_avg:170.36ms
step:965/1530 train_loss:3.4870 train_time:162698ms step_avg:170.36ms
step:966/1530 train_loss:3.7141 train_time:162874ms step_avg:170.37ms
step:967/1530 train_loss:3.5294 train_time:163049ms step_avg:170.37ms
step:968/1530 train_loss:3.5177 train_time:163224ms step_avg:170.38ms
step:969/1530 train_loss:3.5875 train_time:163398ms step_avg:170.38ms
step:970/1530 train_loss:3.3778 train_time:163571ms step_avg:170.39ms
step:971/1530 train_loss:3.5393 train_time:163744ms step_avg:170.39ms
step:972/1530 train_loss:3.4758 train_time:163918ms step_avg:170.39ms
step:973/1530 train_loss:3.5430 train_time:164093ms step_avg:170.40ms
step:974/1530 train_loss:3.5939 train_time:164271ms step_avg:170.41ms
step:975/1530 train_loss:3.4692 train_time:164446ms step_avg:170.41ms
step:976/1530 train_loss:3.6809 train_time:164621ms step_avg:170.41ms
step:977/1530 train_loss:3.5715 train_time:164795ms step_avg:170.42ms
step:978/1530 train_loss:3.3629 train_time:164971ms step_avg:170.42ms
step:979/1530 train_loss:3.6358 train_time:165145ms step_avg:170.43ms
step:980/1530 train_loss:3.4196 train_time:165321ms step_avg:170.43ms
step:981/1530 train_loss:3.5813 train_time:165498ms step_avg:170.44ms
step:982/1530 train_loss:3.5507 train_time:165673ms step_avg:170.45ms
step:983/1530 train_loss:3.5184 train_time:165848ms step_avg:170.45ms
step:984/1530 train_loss:3.4984 train_time:166022ms step_avg:170.45ms
step:985/1530 train_loss:3.5803 train_time:166199ms step_avg:170.46ms
step:986/1530 train_loss:3.4177 train_time:166376ms step_avg:170.47ms
step:987/1530 train_loss:3.4922 train_time:166548ms step_avg:170.47ms
step:988/1530 train_loss:3.4913 train_time:166722ms step_avg:170.47ms
step:989/1530 train_loss:3.4217 train_time:166896ms step_avg:170.48ms
step:990/1530 train_loss:3.6680 train_time:167073ms step_avg:170.48ms
step:991/1530 train_loss:3.4731 train_time:167246ms step_avg:170.49ms
step:992/1530 train_loss:3.4541 train_time:167427ms step_avg:170.50ms
step:993/1530 train_loss:3.5099 train_time:167606ms step_avg:170.50ms
step:994/1530 train_loss:3.6019 train_time:167779ms step_avg:170.51ms
step:995/1530 train_loss:3.5397 train_time:167952ms step_avg:170.51ms
step:996/1530 train_loss:3.4614 train_time:168126ms step_avg:170.51ms
step:997/1530 train_loss:3.7606 train_time:168300ms step_avg:170.52ms
step:998/1530 train_loss:3.4446 train_time:168473ms step_avg:170.52ms
step:999/1530 train_loss:3.5932 train_time:168648ms step_avg:170.52ms
step:1000/1530 train_loss:3.4410 train_time:168826ms step_avg:170.53ms
step:1000/1530 val_loss:3.4721 train_time:168877ms step_avg:170.58ms
step:1001/1530 train_loss:3.5071 train_time:169001ms step_avg:170.54ms
step:1002/1530 train_loss:3.3789 train_time:169176ms step_avg:170.54ms
step:1003/1530 train_loss:3.5611 train_time:169354ms step_avg:170.55ms
step:1004/1530 train_loss:3.6107 train_time:169527ms step_avg:170.55ms
step:1005/1530 train_loss:3.3991 train_time:169704ms step_avg:170.56ms
step:1006/1530 train_loss:3.4673 train_time:169882ms step_avg:170.56ms
step:1007/1530 train_loss:3.4464 train_time:170057ms step_avg:170.57ms
step:1008/1530 train_loss:3.5671 train_time:170233ms step_avg:170.57ms
step:1009/1530 train_loss:3.6667 train_time:170412ms step_avg:170.58ms
step:1010/1530 train_loss:3.5601 train_time:170585ms step_avg:170.59ms
step:1011/1530 train_loss:3.5365 train_time:170759ms step_avg:170.59ms
step:1012/1530 train_loss:3.3925 train_time:170935ms step_avg:170.59ms
step:1013/1530 train_loss:3.5358 train_time:171110ms step_avg:170.60ms
step:1014/1530 train_loss:3.6314 train_time:171286ms step_avg:170.60ms
step:1015/1530 train_loss:3.3308 train_time:171462ms step_avg:170.61ms
step:1016/1530 train_loss:3.4153 train_time:171637ms step_avg:170.61ms
step:1017/1530 train_loss:3.4043 train_time:171813ms step_avg:170.62ms
step:1018/1530 train_loss:3.3994 train_time:171988ms step_avg:170.62ms
step:1019/1530 train_loss:3.5213 train_time:172163ms step_avg:170.63ms
step:1020/1530 train_loss:3.3843 train_time:172340ms step_avg:170.63ms
step:1021/1530 train_loss:3.3600 train_time:172516ms step_avg:170.64ms
step:1022/1530 train_loss:3.4835 train_time:172693ms step_avg:170.65ms
step:1023/1530 train_loss:3.5083 train_time:172867ms step_avg:170.65ms
step:1024/1530 train_loss:3.4808 train_time:173043ms step_avg:170.65ms
step:1025/1530 train_loss:3.4872 train_time:173222ms step_avg:170.66ms
step:1026/1530 train_loss:3.6247 train_time:173398ms step_avg:170.67ms
step:1027/1530 train_loss:3.3265 train_time:173572ms step_avg:170.67ms
step:1028/1530 train_loss:3.4034 train_time:173753ms step_avg:170.68ms
step:1029/1530 train_loss:3.3162 train_time:173934ms step_avg:170.69ms
step:1030/1530 train_loss:3.5426 train_time:174111ms step_avg:170.70ms
step:1031/1530 train_loss:3.5120 train_time:174287ms step_avg:170.70ms
step:1032/1530 train_loss:3.7007 train_time:174468ms step_avg:170.71ms
step:1033/1530 train_loss:3.4959 train_time:174644ms step_avg:170.72ms
step:1034/1530 train_loss:3.4071 train_time:174821ms step_avg:170.72ms
step:1035/1530 train_loss:3.4514 train_time:175001ms step_avg:170.73ms
step:1036/1530 train_loss:3.4839 train_time:175178ms step_avg:170.74ms
step:1037/1530 train_loss:3.7902 train_time:175356ms step_avg:170.75ms
step:1038/1530 train_loss:3.6221 train_time:175535ms step_avg:170.75ms
step:1039/1530 train_loss:3.5146 train_time:175717ms step_avg:170.76ms
step:1040/1530 train_loss:3.4215 train_time:175894ms step_avg:170.77ms
step:1041/1530 train_loss:3.4922 train_time:176073ms step_avg:170.78ms
step:1042/1530 train_loss:3.5259 train_time:176246ms step_avg:170.78ms
step:1043/1530 train_loss:3.4477 train_time:176421ms step_avg:170.79ms
step:1044/1530 train_loss:3.4615 train_time:176599ms step_avg:170.79ms
step:1045/1530 train_loss:3.5219 train_time:176778ms step_avg:170.80ms
step:1046/1530 train_loss:3.4294 train_time:176952ms step_avg:170.80ms
step:1047/1530 train_loss:3.6374 train_time:177127ms step_avg:170.81ms
step:1048/1530 train_loss:3.5005 train_time:177303ms step_avg:170.81ms
step:1049/1530 train_loss:3.4113 train_time:177479ms step_avg:170.82ms
step:1050/1530 train_loss:3.3988 train_time:177659ms step_avg:170.83ms
step:1051/1530 train_loss:3.5006 train_time:177836ms step_avg:170.83ms
step:1052/1530 train_loss:3.3717 train_time:178014ms step_avg:170.84ms
step:1053/1530 train_loss:3.6972 train_time:178191ms step_avg:170.84ms
step:1054/1530 train_loss:3.5418 train_time:178370ms step_avg:170.85ms
step:1055/1530 train_loss:3.3877 train_time:178544ms step_avg:170.86ms
step:1056/1530 train_loss:3.5016 train_time:178720ms step_avg:170.86ms
step:1057/1530 train_loss:3.5826 train_time:178898ms step_avg:170.87ms
step:1058/1530 train_loss:3.3041 train_time:179077ms step_avg:170.87ms
step:1059/1530 train_loss:3.3763 train_time:179259ms step_avg:170.89ms
step:1060/1530 train_loss:3.4400 train_time:179435ms step_avg:170.89ms
step:1061/1530 train_loss:3.4219 train_time:179609ms step_avg:170.89ms
step:1062/1530 train_loss:3.3857 train_time:179784ms step_avg:170.90ms
step:1063/1530 train_loss:3.4640 train_time:179960ms step_avg:170.90ms
step:1064/1530 train_loss:3.3900 train_time:180135ms step_avg:170.91ms
step:1065/1530 train_loss:3.3655 train_time:180314ms step_avg:170.91ms
step:1066/1530 train_loss:3.4183 train_time:180491ms step_avg:170.92ms
step:1067/1530 train_loss:3.3040 train_time:180669ms step_avg:170.93ms
step:1068/1530 train_loss:3.4364 train_time:180845ms step_avg:170.93ms
step:1069/1530 train_loss:3.2982 train_time:181025ms step_avg:170.94ms
step:1070/1530 train_loss:3.5706 train_time:181200ms step_avg:170.94ms
step:1071/1530 train_loss:3.5154 train_time:181379ms step_avg:170.95ms
step:1072/1530 train_loss:3.4422 train_time:181554ms step_avg:170.96ms
step:1073/1530 train_loss:3.5286 train_time:181728ms step_avg:170.96ms
step:1074/1530 train_loss:3.4358 train_time:181904ms step_avg:170.96ms
step:1075/1530 train_loss:3.4061 train_time:182082ms step_avg:170.97ms
step:1076/1530 train_loss:3.8049 train_time:182259ms step_avg:170.97ms
step:1077/1530 train_loss:3.4421 train_time:182435ms step_avg:170.98ms
step:1078/1530 train_loss:3.1037 train_time:182619ms step_avg:170.99ms
step:1079/1530 train_loss:3.5375 train_time:182797ms step_avg:171.00ms
step:1080/1530 train_loss:3.4308 train_time:182976ms step_avg:171.01ms
step:1081/1530 train_loss:3.5078 train_time:183151ms step_avg:171.01ms
step:1082/1530 train_loss:3.5946 train_time:183326ms step_avg:171.01ms
step:1083/1530 train_loss:3.5021 train_time:183502ms step_avg:171.02ms
step:1084/1530 train_loss:3.4656 train_time:183679ms step_avg:171.02ms
step:1085/1530 train_loss:3.4332 train_time:183855ms step_avg:171.03ms
step:1086/1530 train_loss:3.6347 train_time:184031ms step_avg:171.03ms
step:1087/1530 train_loss:3.5085 train_time:184205ms step_avg:171.03ms
step:1088/1530 train_loss:3.3750 train_time:184382ms step_avg:171.04ms
step:1089/1530 train_loss:3.3811 train_time:184562ms step_avg:171.05ms
step:1090/1530 train_loss:3.4867 train_time:184741ms step_avg:171.06ms
step:1091/1530 train_loss:3.2866 train_time:184917ms step_avg:171.06ms
step:1092/1530 train_loss:3.4908 train_time:185095ms step_avg:171.07ms
step:1093/1530 train_loss:3.6108 train_time:185274ms step_avg:171.07ms
step:1094/1530 train_loss:3.4557 train_time:185449ms step_avg:171.08ms
step:1095/1530 train_loss:3.4208 train_time:185623ms step_avg:171.08ms
step:1096/1530 train_loss:3.4313 train_time:185803ms step_avg:171.09ms
step:1097/1530 train_loss:3.4983 train_time:185981ms step_avg:171.10ms
step:1098/1530 train_loss:3.5664 train_time:186159ms step_avg:171.10ms
step:1099/1530 train_loss:3.5343 train_time:186335ms step_avg:171.11ms
step:1100/1530 train_loss:3.4333 train_time:186513ms step_avg:171.11ms
step:1101/1530 train_loss:3.2970 train_time:186690ms step_avg:171.12ms
step:1102/1530 train_loss:3.3199 train_time:186867ms step_avg:171.12ms
step:1103/1530 train_loss:3.4471 train_time:187047ms step_avg:171.13ms
step:1104/1530 train_loss:3.3234 train_time:187223ms step_avg:171.14ms
step:1105/1530 train_loss:4.0705 train_time:187402ms step_avg:171.14ms
step:1106/1530 train_loss:3.2328 train_time:187578ms step_avg:171.15ms
step:1107/1530 train_loss:3.5722 train_time:187753ms step_avg:171.15ms
step:1108/1530 train_loss:3.3515 train_time:187927ms step_avg:171.15ms
step:1109/1530 train_loss:3.5047 train_time:188102ms step_avg:171.16ms
step:1110/1530 train_loss:3.4311 train_time:188276ms step_avg:171.16ms
step:1111/1530 train_loss:3.4937 train_time:188450ms step_avg:171.16ms
step:1112/1530 train_loss:3.5630 train_time:188629ms step_avg:171.17ms
step:1113/1530 train_loss:3.4400 train_time:188813ms step_avg:171.18ms
step:1114/1530 train_loss:3.3842 train_time:188993ms step_avg:171.19ms
step:1115/1530 train_loss:3.2439 train_time:189172ms step_avg:171.20ms
step:1116/1530 train_loss:3.4284 train_time:189345ms step_avg:171.20ms
step:1117/1530 train_loss:3.5983 train_time:189524ms step_avg:171.20ms
step:1118/1530 train_loss:3.6279 train_time:189702ms step_avg:171.21ms
step:1119/1530 train_loss:3.4796 train_time:189877ms step_avg:171.21ms
step:1120/1530 train_loss:3.4945 train_time:190055ms step_avg:171.22ms
step:1121/1530 train_loss:3.3939 train_time:190233ms step_avg:171.23ms
step:1122/1530 train_loss:3.4673 train_time:190407ms step_avg:171.23ms
step:1123/1530 train_loss:3.5851 train_time:190583ms step_avg:171.23ms
step:1124/1530 train_loss:3.3435 train_time:190760ms step_avg:171.24ms
step:1125/1530 train_loss:3.2346 train_time:190937ms step_avg:171.24ms
step:1125/1530 val_loss:3.4137 train_time:190988ms step_avg:171.29ms
step:1126/1530 train_loss:3.4844 train_time:191112ms step_avg:171.25ms
step:1127/1530 train_loss:3.6723 train_time:191290ms step_avg:171.25ms
step:1128/1530 train_loss:3.2378 train_time:191469ms step_avg:171.26ms
step:1129/1530 train_loss:3.5572 train_time:191648ms step_avg:171.27ms
step:1130/1530 train_loss:3.3780 train_time:191827ms step_avg:171.27ms
step:1131/1530 train_loss:3.4065 train_time:192007ms step_avg:171.28ms
step:1132/1530 train_loss:3.3739 train_time:192183ms step_avg:171.29ms
step:1133/1530 train_loss:3.4971 train_time:192497ms step_avg:171.41ms
step:1134/1530 train_loss:3.4521 train_time:192683ms step_avg:171.43ms
step:1135/1530 train_loss:3.5238 train_time:192861ms step_avg:171.43ms
step:1136/1530 train_loss:3.5685 train_time:193038ms step_avg:171.44ms
step:1137/1530 train_loss:3.4608 train_time:193212ms step_avg:171.44ms
step:1138/1530 train_loss:3.3624 train_time:193391ms step_avg:171.45ms
step:1139/1530 train_loss:3.6606 train_time:193727ms step_avg:171.59ms
step:1140/1530 train_loss:3.4639 train_time:193904ms step_avg:171.60ms
step:1141/1530 train_loss:3.6007 train_time:194086ms step_avg:171.61ms
step:1142/1530 train_loss:3.4512 train_time:194263ms step_avg:171.61ms
step:1143/1530 train_loss:3.3678 train_time:194442ms step_avg:171.62ms
step:1144/1530 train_loss:3.4499 train_time:194619ms step_avg:171.62ms
step:1145/1530 train_loss:3.5959 train_time:194793ms step_avg:171.62ms
step:1146/1530 train_loss:3.5702 train_time:194974ms step_avg:171.63ms
step:1147/1530 train_loss:3.5024 train_time:195151ms step_avg:171.64ms
step:1148/1530 train_loss:3.5024 train_time:195328ms step_avg:171.64ms
step:1149/1530 train_loss:3.3345 train_time:195507ms step_avg:171.65ms
step:1150/1530 train_loss:3.3821 train_time:195683ms step_avg:171.65ms
step:1151/1530 train_loss:3.3228 train_time:195862ms step_avg:171.66ms
step:1152/1530 train_loss:3.4088 train_time:196045ms step_avg:171.67ms
step:1153/1530 train_loss:3.4406 train_time:196223ms step_avg:171.67ms
step:1154/1530 train_loss:3.5215 train_time:196400ms step_avg:171.68ms
step:1155/1530 train_loss:3.3259 train_time:196584ms step_avg:171.69ms
step:1156/1530 train_loss:3.5367 train_time:196766ms step_avg:171.70ms
step:1157/1530 train_loss:3.4998 train_time:196945ms step_avg:171.70ms
step:1158/1530 train_loss:3.2521 train_time:197121ms step_avg:171.71ms
step:1159/1530 train_loss:3.3546 train_time:197298ms step_avg:171.71ms
step:1160/1530 train_loss:3.3430 train_time:197473ms step_avg:171.72ms
step:1161/1530 train_loss:3.0900 train_time:197651ms step_avg:171.72ms
step:1162/1530 train_loss:3.4272 train_time:197828ms step_avg:171.73ms
step:1163/1530 train_loss:3.3929 train_time:198007ms step_avg:171.73ms
step:1164/1530 train_loss:3.2985 train_time:198185ms step_avg:171.74ms
step:1165/1530 train_loss:3.2548 train_time:198362ms step_avg:171.74ms
step:1166/1530 train_loss:3.3938 train_time:198541ms step_avg:171.75ms
step:1167/1530 train_loss:3.4164 train_time:198716ms step_avg:171.75ms
step:1168/1530 train_loss:3.7271 train_time:198890ms step_avg:171.75ms
step:1169/1530 train_loss:3.3817 train_time:199067ms step_avg:171.76ms
step:1170/1530 train_loss:3.3966 train_time:199244ms step_avg:171.76ms
step:1171/1530 train_loss:3.2971 train_time:199420ms step_avg:171.77ms
step:1172/1530 train_loss:3.4287 train_time:199595ms step_avg:171.77ms
step:1173/1530 train_loss:3.5495 train_time:199773ms step_avg:171.77ms
step:1174/1530 train_loss:3.3866 train_time:199959ms step_avg:171.79ms
step:1175/1530 train_loss:3.3669 train_time:200138ms step_avg:171.79ms
step:1176/1530 train_loss:3.4340 train_time:200317ms step_avg:171.80ms
step:1177/1530 train_loss:3.4488 train_time:200500ms step_avg:171.81ms
step:1178/1530 train_loss:3.5019 train_time:200677ms step_avg:171.81ms
step:1179/1530 train_loss:3.4078 train_time:200851ms step_avg:171.81ms
step:1180/1530 train_loss:3.3630 train_time:201038ms step_avg:171.83ms
step:1181/1530 train_loss:3.3433 train_time:201216ms step_avg:171.83ms
step:1182/1530 train_loss:3.3818 train_time:201394ms step_avg:171.84ms
step:1183/1530 train_loss:3.3423 train_time:201573ms step_avg:171.84ms
step:1184/1530 train_loss:3.5149 train_time:201750ms step_avg:171.85ms
step:1185/1530 train_loss:3.5513 train_time:201931ms step_avg:171.86ms
step:1186/1530 train_loss:3.3680 train_time:202110ms step_avg:171.86ms
step:1187/1530 train_loss:3.4224 train_time:202294ms step_avg:171.87ms
step:1188/1530 train_loss:3.4442 train_time:202470ms step_avg:171.88ms
step:1189/1530 train_loss:3.2805 train_time:202650ms step_avg:171.88ms
step:1190/1530 train_loss:3.4480 train_time:202828ms step_avg:171.89ms
step:1191/1530 train_loss:3.5873 train_time:203009ms step_avg:171.90ms
step:1192/1530 train_loss:3.3945 train_time:203185ms step_avg:171.90ms
step:1193/1530 train_loss:3.2827 train_time:203362ms step_avg:171.90ms
step:1194/1530 train_loss:3.5589 train_time:203540ms step_avg:171.91ms
step:1195/1530 train_loss:3.3753 train_time:203720ms step_avg:171.92ms
step:1196/1530 train_loss:3.3910 train_time:203906ms step_avg:171.93ms
step:1197/1530 train_loss:3.2976 train_time:204086ms step_avg:171.93ms
step:1198/1530 train_loss:3.3019 train_time:204271ms step_avg:171.94ms
step:1199/1530 train_loss:3.3463 train_time:204450ms step_avg:171.95ms
step:1200/1530 train_loss:3.4522 train_time:204626ms step_avg:171.95ms
step:1201/1530 train_loss:3.4859 train_time:204804ms step_avg:171.96ms
step:1202/1530 train_loss:3.6012 train_time:204993ms step_avg:171.97ms
step:1203/1530 train_loss:3.4117 train_time:205172ms step_avg:171.98ms
step:1204/1530 train_loss:3.3095 train_time:205353ms step_avg:171.99ms
step:1205/1530 train_loss:3.4383 train_time:205529ms step_avg:171.99ms
step:1206/1530 train_loss:3.4755 train_time:205706ms step_avg:171.99ms
step:1207/1530 train_loss:3.5236 train_time:205885ms step_avg:172.00ms
step:1208/1530 train_loss:3.4004 train_time:206061ms step_avg:172.00ms
step:1209/1530 train_loss:3.2526 train_time:206239ms step_avg:172.01ms
step:1210/1530 train_loss:3.3090 train_time:206418ms step_avg:172.01ms
step:1211/1530 train_loss:3.3995 train_time:206595ms step_avg:172.02ms
step:1212/1530 train_loss:3.4023 train_time:206771ms step_avg:172.02ms
step:1213/1530 train_loss:3.4141 train_time:206950ms step_avg:172.03ms
step:1214/1530 train_loss:3.2608 train_time:207132ms step_avg:172.04ms
step:1215/1530 train_loss:3.4047 train_time:207308ms step_avg:172.04ms
step:1216/1530 train_loss:3.3348 train_time:207485ms step_avg:172.04ms
step:1217/1530 train_loss:3.3266 train_time:207662ms step_avg:172.05ms
step:1218/1530 train_loss:3.4105 train_time:207842ms step_avg:172.05ms
step:1219/1530 train_loss:3.2580 train_time:208025ms step_avg:172.06ms
step:1220/1530 train_loss:3.4825 train_time:208201ms step_avg:172.07ms
step:1221/1530 train_loss:3.5089 train_time:208378ms step_avg:172.07ms
step:1222/1530 train_loss:3.4323 train_time:208552ms step_avg:172.07ms
step:1223/1530 train_loss:3.3034 train_time:208729ms step_avg:172.08ms
step:1224/1530 train_loss:3.2556 train_time:208910ms step_avg:172.08ms
step:1225/1530 train_loss:3.3684 train_time:209088ms step_avg:172.09ms
step:1226/1530 train_loss:3.3358 train_time:209268ms step_avg:172.10ms
step:1227/1530 train_loss:3.2812 train_time:209447ms step_avg:172.10ms
step:1228/1530 train_loss:3.4501 train_time:209623ms step_avg:172.10ms
step:1229/1530 train_loss:3.3729 train_time:209804ms step_avg:172.11ms
step:1230/1530 train_loss:3.4040 train_time:209988ms step_avg:172.12ms
step:1231/1530 train_loss:3.5801 train_time:210168ms step_avg:172.13ms
step:1232/1530 train_loss:3.5033 train_time:210349ms step_avg:172.13ms
step:1233/1530 train_loss:3.4345 train_time:210527ms step_avg:172.14ms
step:1234/1530 train_loss:3.5883 train_time:210706ms step_avg:172.15ms
step:1235/1530 train_loss:3.3268 train_time:210887ms step_avg:172.15ms
step:1236/1530 train_loss:3.2987 train_time:211065ms step_avg:172.16ms
step:1237/1530 train_loss:3.2780 train_time:211242ms step_avg:172.16ms
step:1238/1530 train_loss:3.2854 train_time:211425ms step_avg:172.17ms
step:1239/1530 train_loss:3.3357 train_time:211604ms step_avg:172.18ms
step:1240/1530 train_loss:3.3885 train_time:211782ms step_avg:172.18ms
step:1241/1530 train_loss:3.4343 train_time:211961ms step_avg:172.19ms
step:1242/1530 train_loss:3.3042 train_time:212138ms step_avg:172.19ms
step:1243/1530 train_loss:3.4119 train_time:212317ms step_avg:172.20ms
step:1244/1530 train_loss:3.4108 train_time:212491ms step_avg:172.20ms
step:1245/1530 train_loss:3.4177 train_time:212668ms step_avg:172.20ms
step:1246/1530 train_loss:3.2477 train_time:212845ms step_avg:172.20ms
step:1247/1530 train_loss:3.3766 train_time:213020ms step_avg:172.21ms
step:1248/1530 train_loss:3.4322 train_time:213197ms step_avg:172.21ms
step:1249/1530 train_loss:3.4295 train_time:213377ms step_avg:172.22ms
step:1250/1530 train_loss:3.3073 train_time:213555ms step_avg:172.22ms
step:1250/1530 val_loss:3.3605 train_time:213608ms step_avg:172.26ms
step:1251/1530 train_loss:3.4993 train_time:213738ms step_avg:172.23ms
step:1252/1530 train_loss:3.3689 train_time:213914ms step_avg:172.23ms
step:1253/1530 train_loss:3.3129 train_time:214091ms step_avg:172.24ms
step:1254/1530 train_loss:3.4254 train_time:214273ms step_avg:172.25ms
step:1255/1530 train_loss:3.5258 train_time:214461ms step_avg:172.26ms
step:1256/1530 train_loss:3.3099 train_time:214643ms step_avg:172.27ms
step:1257/1530 train_loss:3.3828 train_time:214820ms step_avg:172.27ms
step:1258/1530 train_loss:3.3725 train_time:215004ms step_avg:172.28ms
step:1259/1530 train_loss:3.3307 train_time:215182ms step_avg:172.28ms
step:1260/1530 train_loss:3.2140 train_time:215359ms step_avg:172.29ms
step:1261/1530 train_loss:3.3049 train_time:215538ms step_avg:172.29ms
step:1262/1530 train_loss:3.3329 train_time:215721ms step_avg:172.30ms
step:1263/1530 train_loss:3.2454 train_time:215901ms step_avg:172.31ms
step:1264/1530 train_loss:3.4463 train_time:216078ms step_avg:172.31ms
step:1265/1530 train_loss:3.4296 train_time:216255ms step_avg:172.31ms
step:1266/1530 train_loss:3.4426 train_time:216435ms step_avg:172.32ms
step:1267/1530 train_loss:3.3751 train_time:216615ms step_avg:172.33ms
step:1268/1530 train_loss:3.4142 train_time:216796ms step_avg:172.33ms
step:1269/1530 train_loss:3.2561 train_time:216980ms step_avg:172.34ms
step:1270/1530 train_loss:3.1127 train_time:217157ms step_avg:172.35ms
step:1271/1530 train_loss:3.4113 train_time:217336ms step_avg:172.35ms
step:1272/1530 train_loss:3.3565 train_time:217512ms step_avg:172.36ms
step:1273/1530 train_loss:3.3792 train_time:217694ms step_avg:172.36ms
step:1274/1530 train_loss:3.3640 train_time:217875ms step_avg:172.37ms
step:1275/1530 train_loss:3.4399 train_time:218053ms step_avg:172.37ms
step:1276/1530 train_loss:3.4717 train_time:218227ms step_avg:172.38ms
step:1277/1530 train_loss:3.4183 train_time:218406ms step_avg:172.38ms
step:1278/1530 train_loss:3.4132 train_time:218581ms step_avg:172.38ms
step:1279/1530 train_loss:3.2697 train_time:218763ms step_avg:172.39ms
step:1280/1530 train_loss:3.3725 train_time:218949ms step_avg:172.40ms
step:1281/1530 train_loss:3.4302 train_time:219127ms step_avg:172.41ms
step:1282/1530 train_loss:3.4789 train_time:219302ms step_avg:172.41ms
step:1283/1530 train_loss:3.3416 train_time:219480ms step_avg:172.41ms
step:1284/1530 train_loss:3.3770 train_time:219660ms step_avg:172.42ms
step:1285/1530 train_loss:3.3660 train_time:219838ms step_avg:172.42ms
step:1286/1530 train_loss:3.3390 train_time:220015ms step_avg:172.43ms
step:1287/1530 train_loss:3.4895 train_time:220194ms step_avg:172.43ms
step:1288/1530 train_loss:3.3015 train_time:220375ms step_avg:172.44ms
step:1289/1530 train_loss:3.3921 train_time:220561ms step_avg:172.45ms
step:1290/1530 train_loss:3.4656 train_time:220746ms step_avg:172.46ms
step:1291/1530 train_loss:3.3856 train_time:220925ms step_avg:172.46ms
step:1292/1530 train_loss:3.4859 train_time:221107ms step_avg:172.47ms
step:1293/1530 train_loss:3.5230 train_time:221287ms step_avg:172.48ms
step:1294/1530 train_loss:3.4575 train_time:221468ms step_avg:172.48ms
step:1295/1530 train_loss:3.2897 train_time:221648ms step_avg:172.49ms
step:1296/1530 train_loss:3.3820 train_time:221827ms step_avg:172.49ms
step:1297/1530 train_loss:3.2802 train_time:222007ms step_avg:172.50ms
step:1298/1530 train_loss:3.2804 train_time:222190ms step_avg:172.51ms
step:1299/1530 train_loss:3.4030 train_time:222370ms step_avg:172.51ms
step:1300/1530 train_loss:3.4078 train_time:222547ms step_avg:172.52ms
step:1301/1530 train_loss:3.4112 train_time:222723ms step_avg:172.52ms
step:1302/1530 train_loss:3.5806 train_time:222905ms step_avg:172.53ms
step:1303/1530 train_loss:3.3109 train_time:223087ms step_avg:172.53ms
step:1304/1530 train_loss:3.5176 train_time:223271ms step_avg:172.54ms
step:1305/1530 train_loss:3.2653 train_time:223448ms step_avg:172.55ms
step:1306/1530 train_loss:3.4626 train_time:223631ms step_avg:172.55ms
step:1307/1530 train_loss:3.4589 train_time:223807ms step_avg:172.56ms
step:1308/1530 train_loss:3.2905 train_time:223987ms step_avg:172.56ms
step:1309/1530 train_loss:3.3137 train_time:224168ms step_avg:172.57ms
step:1310/1530 train_loss:3.2938 train_time:224346ms step_avg:172.57ms
step:1311/1530 train_loss:3.3020 train_time:224524ms step_avg:172.58ms
step:1312/1530 train_loss:3.3786 train_time:224705ms step_avg:172.58ms
step:1313/1530 train_loss:3.3427 train_time:224880ms step_avg:172.59ms
step:1314/1530 train_loss:3.0532 train_time:225061ms step_avg:172.59ms
step:1315/1530 train_loss:3.2852 train_time:225239ms step_avg:172.60ms
step:1316/1530 train_loss:3.4050 train_time:225415ms step_avg:172.60ms
step:1317/1530 train_loss:3.4277 train_time:225594ms step_avg:172.60ms
step:1318/1530 train_loss:3.3088 train_time:225780ms step_avg:172.61ms
step:1319/1530 train_loss:3.4290 train_time:225960ms step_avg:172.62ms
step:1320/1530 train_loss:3.4687 train_time:226143ms step_avg:172.63ms
step:1321/1530 train_loss:3.3675 train_time:226320ms step_avg:172.63ms
step:1322/1530 train_loss:3.3271 train_time:226637ms step_avg:172.74ms
step:1323/1530 train_loss:3.3295 train_time:226826ms step_avg:172.75ms
step:1324/1530 train_loss:3.4431 train_time:227008ms step_avg:172.76ms
step:1325/1530 train_loss:3.4973 train_time:227195ms step_avg:172.77ms
step:1326/1530 train_loss:3.2154 train_time:227375ms step_avg:172.78ms
step:1327/1530 train_loss:3.1729 train_time:227551ms step_avg:172.78ms
step:1328/1530 train_loss:3.4965 train_time:227730ms step_avg:172.78ms
step:1329/1530 train_loss:3.3043 train_time:228080ms step_avg:172.92ms
step:1330/1530 train_loss:3.4375 train_time:228262ms step_avg:172.93ms
step:1331/1530 train_loss:3.3373 train_time:228439ms step_avg:172.93ms
step:1332/1530 train_loss:3.7417 train_time:228619ms step_avg:172.93ms
step:1333/1530 train_loss:3.4802 train_time:228799ms step_avg:172.94ms
step:1334/1530 train_loss:3.3779 train_time:228976ms step_avg:172.94ms
step:1335/1530 train_loss:3.2941 train_time:229156ms step_avg:172.95ms
step:1336/1530 train_loss:3.3055 train_time:229340ms step_avg:172.96ms
step:1337/1530 train_loss:3.5593 train_time:229519ms step_avg:172.96ms
step:1338/1530 train_loss:3.5320 train_time:229698ms step_avg:172.97ms
step:1339/1530 train_loss:3.3425 train_time:229876ms step_avg:172.97ms
step:1340/1530 train_loss:3.2892 train_time:230056ms step_avg:172.97ms
step:1341/1530 train_loss:3.6032 train_time:230233ms step_avg:172.98ms
step:1342/1530 train_loss:3.3650 train_time:230413ms step_avg:172.98ms
step:1343/1530 train_loss:3.3687 train_time:230592ms step_avg:172.99ms
step:1344/1530 train_loss:3.4176 train_time:230775ms step_avg:172.99ms
step:1345/1530 train_loss:3.3883 train_time:230957ms step_avg:173.00ms
step:1346/1530 train_loss:3.2988 train_time:231133ms step_avg:173.00ms
step:1347/1530 train_loss:3.2841 train_time:231311ms step_avg:173.01ms
step:1348/1530 train_loss:3.3563 train_time:231489ms step_avg:173.01ms
step:1349/1530 train_loss:3.2804 train_time:231666ms step_avg:173.01ms
step:1350/1530 train_loss:3.3981 train_time:231848ms step_avg:173.02ms
step:1351/1530 train_loss:3.2493 train_time:232024ms step_avg:173.02ms
step:1352/1530 train_loss:3.3157 train_time:232200ms step_avg:173.03ms
step:1353/1530 train_loss:3.4068 train_time:232379ms step_avg:173.03ms
step:1354/1530 train_loss:3.2690 train_time:232557ms step_avg:173.03ms
step:1355/1530 train_loss:3.1969 train_time:232735ms step_avg:173.04ms
step:1356/1530 train_loss:3.5190 train_time:232915ms step_avg:173.04ms
step:1357/1530 train_loss:3.4318 train_time:233096ms step_avg:173.05ms
step:1358/1530 train_loss:3.1946 train_time:233275ms step_avg:173.05ms
step:1359/1530 train_loss:3.4443 train_time:233456ms step_avg:173.06ms
step:1360/1530 train_loss:3.3571 train_time:233636ms step_avg:173.06ms
step:1361/1530 train_loss:3.1390 train_time:233822ms step_avg:173.07ms
step:1362/1530 train_loss:3.3997 train_time:234004ms step_avg:173.08ms
step:1363/1530 train_loss:3.2865 train_time:234192ms step_avg:173.09ms
step:1364/1530 train_loss:3.3044 train_time:234372ms step_avg:173.10ms
step:1365/1530 train_loss:3.3202 train_time:234552ms step_avg:173.10ms
step:1366/1530 train_loss:3.4248 train_time:234734ms step_avg:173.11ms
step:1367/1530 train_loss:3.4030 train_time:234912ms step_avg:173.11ms
step:1368/1530 train_loss:3.3514 train_time:235094ms step_avg:173.12ms
step:1369/1530 train_loss:3.2823 train_time:235282ms step_avg:173.13ms
step:1370/1530 train_loss:3.6099 train_time:235465ms step_avg:173.14ms
step:1371/1530 train_loss:3.3186 train_time:235646ms step_avg:173.14ms
step:1372/1530 train_loss:3.3756 train_time:235829ms step_avg:173.15ms
step:1373/1530 train_loss:3.3743 train_time:236010ms step_avg:173.15ms
step:1374/1530 train_loss:3.1578 train_time:236192ms step_avg:173.16ms
step:1375/1530 train_loss:3.5440 train_time:236373ms step_avg:173.17ms
step:1375/1530 val_loss:3.3182 train_time:236424ms step_avg:173.20ms
step:1376/1530 train_loss:3.3573 train_time:236554ms step_avg:173.17ms
step:1377/1530 train_loss:3.4897 train_time:236733ms step_avg:173.18ms
step:1378/1530 train_loss:3.4721 train_time:236908ms step_avg:173.18ms
step:1379/1530 train_loss:3.1175 train_time:237091ms step_avg:173.19ms
step:1380/1530 train_loss:3.3268 train_time:237271ms step_avg:173.19ms
step:1381/1530 train_loss:3.7099 train_time:237458ms step_avg:173.20ms
step:1382/1530 train_loss:3.2140 train_time:237638ms step_avg:173.21ms
step:1383/1530 train_loss:3.3983 train_time:237819ms step_avg:173.21ms
step:1384/1530 train_loss:3.4823 train_time:238004ms step_avg:173.22ms
step:1385/1530 train_loss:3.4095 train_time:238179ms step_avg:173.22ms
step:1386/1530 train_loss:3.3512 train_time:238361ms step_avg:173.23ms
step:1387/1530 train_loss:3.2080 train_time:238540ms step_avg:173.23ms
step:1388/1530 train_loss:3.3467 train_time:238719ms step_avg:173.24ms
step:1389/1530 train_loss:3.3211 train_time:238901ms step_avg:173.24ms
step:1390/1530 train_loss:3.5742 train_time:239078ms step_avg:173.24ms
step:1391/1530 train_loss:3.2978 train_time:239257ms step_avg:173.25ms
step:1392/1530 train_loss:3.2911 train_time:239437ms step_avg:173.25ms
step:1393/1530 train_loss:3.2432 train_time:239617ms step_avg:173.26ms
step:1394/1530 train_loss:3.5051 train_time:239792ms step_avg:173.26ms
step:1395/1530 train_loss:3.3956 train_time:239972ms step_avg:173.26ms
step:1396/1530 train_loss:3.4122 train_time:240150ms step_avg:173.27ms
step:1397/1530 train_loss:3.3190 train_time:240326ms step_avg:173.27ms
step:1398/1530 train_loss:3.2614 train_time:240501ms step_avg:173.27ms
step:1399/1530 train_loss:3.3215 train_time:240680ms step_avg:173.28ms
step:1400/1530 train_loss:3.3228 train_time:240863ms step_avg:173.28ms
step:1401/1530 train_loss:3.3517 train_time:241040ms step_avg:173.29ms
step:1402/1530 train_loss:3.3031 train_time:241221ms step_avg:173.29ms
step:1403/1530 train_loss:3.5004 train_time:241405ms step_avg:173.30ms
step:1404/1530 train_loss:3.2859 train_time:241582ms step_avg:173.30ms
step:1405/1530 train_loss:3.3201 train_time:241762ms step_avg:173.31ms
step:1406/1530 train_loss:3.3180 train_time:241942ms step_avg:173.31ms
step:1407/1530 train_loss:3.1806 train_time:242119ms step_avg:173.31ms
step:1408/1530 train_loss:3.3182 train_time:242299ms step_avg:173.32ms
step:1409/1530 train_loss:3.3064 train_time:242487ms step_avg:173.33ms
step:1410/1530 train_loss:3.2966 train_time:242664ms step_avg:173.33ms
step:1411/1530 train_loss:3.3726 train_time:242841ms step_avg:173.33ms
step:1412/1530 train_loss:3.3396 train_time:243018ms step_avg:173.34ms
step:1413/1530 train_loss:3.3663 train_time:243197ms step_avg:173.34ms
step:1414/1530 train_loss:3.3336 train_time:243377ms step_avg:173.35ms
step:1415/1530 train_loss:3.4107 train_time:243562ms step_avg:173.35ms
step:1416/1530 train_loss:3.2411 train_time:243750ms step_avg:173.36ms
step:1417/1530 train_loss:3.2890 train_time:243932ms step_avg:173.37ms
step:1418/1530 train_loss:3.3963 train_time:244113ms step_avg:173.38ms
step:1419/1530 train_loss:3.3550 train_time:244296ms step_avg:173.38ms
step:1420/1530 train_loss:3.3751 train_time:244476ms step_avg:173.39ms
step:1421/1530 train_loss:3.3737 train_time:244656ms step_avg:173.39ms
step:1422/1530 train_loss:3.3366 train_time:244835ms step_avg:173.40ms
step:1423/1530 train_loss:3.3261 train_time:245014ms step_avg:173.40ms
step:1424/1530 train_loss:3.3425 train_time:245198ms step_avg:173.41ms
step:1425/1530 train_loss:3.2042 train_time:245383ms step_avg:173.42ms
step:1426/1530 train_loss:3.3288 train_time:245561ms step_avg:173.42ms
step:1427/1530 train_loss:3.2946 train_time:245744ms step_avg:173.43ms
step:1428/1530 train_loss:3.3894 train_time:245922ms step_avg:173.43ms
step:1429/1530 train_loss:3.3566 train_time:246098ms step_avg:173.43ms
step:1430/1530 train_loss:3.2694 train_time:246281ms step_avg:173.44ms
step:1431/1530 train_loss:3.3298 train_time:246463ms step_avg:173.44ms
step:1432/1530 train_loss:3.3473 train_time:246646ms step_avg:173.45ms
step:1433/1530 train_loss:3.1377 train_time:246830ms step_avg:173.46ms
step:1434/1530 train_loss:3.2969 train_time:247014ms step_avg:173.46ms
step:1435/1530 train_loss:3.1237 train_time:247194ms step_avg:173.47ms
step:1436/1530 train_loss:3.2356 train_time:247375ms step_avg:173.48ms
step:1437/1530 train_loss:3.4110 train_time:247553ms step_avg:173.48ms
step:1438/1530 train_loss:3.3861 train_time:247731ms step_avg:173.48ms
step:1439/1530 train_loss:3.3199 train_time:247912ms step_avg:173.49ms
step:1440/1530 train_loss:3.2037 train_time:248087ms step_avg:173.49ms
step:1441/1530 train_loss:3.3481 train_time:248266ms step_avg:173.49ms
step:1442/1530 train_loss:3.3948 train_time:248450ms step_avg:173.50ms
step:1443/1530 train_loss:3.5014 train_time:248638ms step_avg:173.51ms
step:1444/1530 train_loss:3.4577 train_time:248815ms step_avg:173.51ms
step:1445/1530 train_loss:3.3443 train_time:248992ms step_avg:173.51ms
step:1446/1530 train_loss:3.2009 train_time:249171ms step_avg:173.52ms
step:1447/1530 train_loss:3.3015 train_time:249353ms step_avg:173.52ms
step:1448/1530 train_loss:3.3036 train_time:249531ms step_avg:173.53ms
step:1449/1530 train_loss:3.4031 train_time:249708ms step_avg:173.53ms
step:1450/1530 train_loss:3.3949 train_time:249888ms step_avg:173.53ms
step:1451/1530 train_loss:3.2084 train_time:250066ms step_avg:173.54ms
step:1452/1530 train_loss:3.3352 train_time:250247ms step_avg:173.54ms
step:1453/1530 train_loss:3.2688 train_time:250422ms step_avg:173.54ms
step:1454/1530 train_loss:3.2962 train_time:250600ms step_avg:173.55ms
step:1455/1530 train_loss:3.3384 train_time:250783ms step_avg:173.55ms
step:1456/1530 train_loss:3.2890 train_time:250960ms step_avg:173.55ms
step:1457/1530 train_loss:3.1613 train_time:251138ms step_avg:173.56ms
step:1458/1530 train_loss:3.4328 train_time:251317ms step_avg:173.56ms
step:1459/1530 train_loss:3.2771 train_time:251499ms step_avg:173.57ms
step:1460/1530 train_loss:3.3212 train_time:251679ms step_avg:173.57ms
step:1461/1530 train_loss:3.4390 train_time:251860ms step_avg:173.58ms
step:1462/1530 train_loss:3.2690 train_time:252036ms step_avg:173.58ms
step:1463/1530 train_loss:3.4696 train_time:252220ms step_avg:173.59ms
step:1464/1530 train_loss:3.3675 train_time:252398ms step_avg:173.59ms
step:1465/1530 train_loss:3.3606 train_time:252580ms step_avg:173.59ms
step:1466/1530 train_loss:3.2897 train_time:252759ms step_avg:173.60ms
step:1467/1530 train_loss:3.4016 train_time:252940ms step_avg:173.60ms
step:1468/1530 train_loss:3.2953 train_time:253117ms step_avg:173.61ms
step:1469/1530 train_loss:3.2827 train_time:253298ms step_avg:173.61ms
step:1470/1530 train_loss:3.3375 train_time:253481ms step_avg:173.62ms
step:1471/1530 train_loss:3.2640 train_time:253667ms step_avg:173.63ms
step:1472/1530 train_loss:3.2532 train_time:253853ms step_avg:173.63ms
step:1473/1530 train_loss:3.4496 train_time:254030ms step_avg:173.64ms
step:1474/1530 train_loss:3.3171 train_time:254215ms step_avg:173.64ms
step:1475/1530 train_loss:3.1585 train_time:254400ms step_avg:173.65ms
step:1476/1530 train_loss:3.2718 train_time:254579ms step_avg:173.66ms
step:1477/1530 train_loss:3.2463 train_time:254766ms step_avg:173.66ms
step:1478/1530 train_loss:3.3138 train_time:254953ms step_avg:173.67ms
step:1479/1530 train_loss:3.4034 train_time:255133ms step_avg:173.68ms
step:1480/1530 train_loss:3.2736 train_time:255310ms step_avg:173.68ms
step:1481/1530 train_loss:3.4549 train_time:255492ms step_avg:173.69ms
step:1482/1530 train_loss:3.3705 train_time:255678ms step_avg:173.69ms
step:1483/1530 train_loss:3.2815 train_time:255869ms step_avg:173.71ms
step:1484/1530 train_loss:3.2737 train_time:256058ms step_avg:173.72ms
step:1485/1530 train_loss:3.2866 train_time:256238ms step_avg:173.72ms
step:1486/1530 train_loss:3.2328 train_time:256424ms step_avg:173.73ms
step:1487/1530 train_loss:3.3471 train_time:256606ms step_avg:173.73ms
step:1488/1530 train_loss:3.2515 train_time:256790ms step_avg:173.74ms
step:1489/1530 train_loss:3.3197 train_time:256971ms step_avg:173.75ms
step:1490/1530 train_loss:3.2589 train_time:257152ms step_avg:173.75ms
step:1491/1530 train_loss:3.1659 train_time:257331ms step_avg:173.76ms
step:1492/1530 train_loss:3.2723 train_time:257510ms step_avg:173.76ms
step:1493/1530 train_loss:3.4413 train_time:257689ms step_avg:173.76ms
step:1494/1530 train_loss:3.3027 train_time:257868ms step_avg:173.77ms
step:1495/1530 train_loss:3.0385 train_time:258054ms step_avg:173.77ms
step:1496/1530 train_loss:3.3658 train_time:258237ms step_avg:173.78ms
step:1497/1530 train_loss:3.3184 train_time:258421ms step_avg:173.79ms
step:1498/1530 train_loss:3.3552 train_time:258606ms step_avg:173.79ms
step:1499/1530 train_loss:3.3164 train_time:258795ms step_avg:173.80ms
step:1500/1530 train_loss:3.3011 train_time:258985ms step_avg:173.82ms
step:1500/1530 val_loss:3.2866 train_time:259040ms step_avg:173.85ms
step:1501/1530 train_loss:3.0941 train_time:259177ms step_avg:173.83ms
step:1502/1530 train_loss:3.3672 train_time:259366ms step_avg:173.84ms
step:1503/1530 train_loss:3.2502 train_time:259545ms step_avg:173.84ms
step:1504/1530 train_loss:3.2551 train_time:259726ms step_avg:173.85ms
step:1505/1530 train_loss:3.2202 train_time:259905ms step_avg:173.85ms
step:1506/1530 train_loss:3.2865 train_time:260085ms step_avg:173.85ms
step:1507/1530 train_loss:3.1864 train_time:260281ms step_avg:173.87ms
step:1508/1530 train_loss:3.4881 train_time:260464ms step_avg:173.87ms
step:1509/1530 train_loss:3.2885 train_time:260640ms step_avg:173.88ms
step:1510/1530 train_loss:3.2811 train_time:260821ms step_avg:173.88ms
step:1511/1530 train_loss:3.4218 train_time:261136ms step_avg:173.97ms
step:1512/1530 train_loss:3.4275 train_time:261325ms step_avg:173.98ms
step:1513/1530 train_loss:3.2753 train_time:261510ms step_avg:173.99ms
step:1514/1530 train_loss:3.0900 train_time:261693ms step_avg:174.00ms
step:1515/1530 train_loss:3.2502 train_time:261875ms step_avg:174.00ms
step:1516/1530 train_loss:3.2616 train_time:262063ms step_avg:174.01ms
step:1517/1530 train_loss:3.3087 train_time:262244ms step_avg:174.02ms
step:1518/1530 train_loss:3.2118 train_time:262427ms step_avg:174.02ms
step:1519/1530 train_loss:3.5113 train_time:262766ms step_avg:174.13ms
step:1520/1530 train_loss:3.1367 train_time:262946ms step_avg:174.14ms
step:1521/1530 train_loss:3.2146 train_time:263124ms step_avg:174.14ms
step:1522/1530 train_loss:3.3646 train_time:263307ms step_avg:174.14ms
step:1523/1530 train_loss:3.2364 train_time:263485ms step_avg:174.15ms
step:1524/1530 train_loss:3.3524 train_time:263662ms step_avg:174.15ms
step:1525/1530 train_loss:3.3449 train_time:263851ms step_avg:174.16ms
step:1526/1530 train_loss:3.2855 train_time:264041ms step_avg:174.17ms
step:1527/1530 train_loss:3.2985 train_time:264223ms step_avg:174.17ms
step:1528/1530 train_loss:3.4172 train_time:264403ms step_avg:174.18ms
step:1529/1530 train_loss:3.4153 train_time:264580ms step_avg:174.18ms
step:1530/1530 train_loss:3.2480 train_time:264758ms step_avg:174.18ms
step:1530/1530 val_loss:3.2842 train_time:264813ms step_avg:174.22ms