4f8bcdc3-18cf-4743-895e-1deb08696fe7.txt

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

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

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

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

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

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

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

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

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

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

    def step(self):

        for group in self.param_groups:

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

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

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

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

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

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

class CastedLinear(nn.Linear):

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

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

class Rotary(torch.nn.Module):

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

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

class CausalSelfAttention(nn.Module):

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

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

class MLP(nn.Module):

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

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

class Block(nn.Module):

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

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

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

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

class GPT(nn.Module):

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

        self.reset()

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

# -------------------------------------------------------------------------
# clean up nice
dist.destroy_process_group()
====================================================================================================
Running pytorch 2.6.0.dev20241203+cu124 compiled for CUDA 12.4
nvidia-smi:
Thu Dec  5 03:35:06 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   31C    P0             113W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   38C    P0             117W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   39C    P0              99W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   30C    P0             110W / 700W |    529MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   39C    P0              94W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   30C    P0             119W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
+---------------------------------------------------------------------------------------+

====================================================================================================
Training DataLoader: total number of tokens: 1100000000 across 11 files
Validation DataLoader: total number of tokens: 100000000 across 1 files
====================================================================================================
step:0/1530 val_loss:10.8258 train_time:0ms step_avg:nanms
step:1/1530 train_loss:10.8258 train_time:31906ms step_avg:nanms
step:2/1530 train_loss:10.0865 train_time:32016ms step_avg:nanms
step:3/1530 train_loss:8.3825 train_time:32176ms step_avg:nanms
step:4/1530 train_loss:7.5261 train_time:32338ms step_avg:nanms
step:5/1530 train_loss:7.4893 train_time:32498ms step_avg:nanms
step:6/1530 train_loss:6.9792 train_time:32658ms step_avg:nanms
step:7/1530 train_loss:7.1948 train_time:32818ms step_avg:nanms
step:8/1530 train_loss:6.7279 train_time:32979ms step_avg:nanms
step:9/1530 train_loss:6.6182 train_time:33140ms step_avg:nanms
step:10/1530 train_loss:6.4870 train_time:33301ms step_avg:nanms
step:11/1530 train_loss:6.4352 train_time:115ms step_avg:nanms
step:12/1530 train_loss:6.3162 train_time:275ms step_avg:nanms
step:13/1530 train_loss:6.2522 train_time:434ms step_avg:144.79ms
step:14/1530 train_loss:6.1979 train_time:595ms step_avg:148.78ms
step:15/1530 train_loss:6.1776 train_time:756ms step_avg:151.18ms
step:16/1530 train_loss:6.1079 train_time:915ms step_avg:152.58ms
step:17/1530 train_loss:6.1595 train_time:1076ms step_avg:153.73ms
step:18/1530 train_loss:5.9458 train_time:1236ms step_avg:154.51ms
step:19/1530 train_loss:5.9823 train_time:1396ms step_avg:155.14ms
step:20/1530 train_loss:5.6711 train_time:1556ms step_avg:155.60ms
step:21/1530 train_loss:5.9676 train_time:1717ms step_avg:156.07ms
step:22/1530 train_loss:6.1869 train_time:1877ms step_avg:156.44ms
step:23/1530 train_loss:5.8462 train_time:2038ms step_avg:156.78ms
step:24/1530 train_loss:6.0146 train_time:2199ms step_avg:157.08ms
step:25/1530 train_loss:5.6915 train_time:2360ms step_avg:157.35ms
step:26/1530 train_loss:5.5899 train_time:2521ms step_avg:157.59ms
step:27/1530 train_loss:5.7757 train_time:2682ms step_avg:157.76ms
step:28/1530 train_loss:5.3907 train_time:2843ms step_avg:157.93ms
step:29/1530 train_loss:5.6686 train_time:3003ms step_avg:158.05ms
step:30/1530 train_loss:5.4577 train_time:3163ms step_avg:158.17ms
step:31/1530 train_loss:5.4375 train_time:3324ms step_avg:158.28ms
step:32/1530 train_loss:5.2745 train_time:3484ms step_avg:158.39ms
step:33/1530 train_loss:5.5856 train_time:3645ms step_avg:158.49ms
step:34/1530 train_loss:5.4954 train_time:3805ms step_avg:158.55ms
step:35/1530 train_loss:5.5982 train_time:3966ms step_avg:158.64ms
step:36/1530 train_loss:5.5232 train_time:4127ms step_avg:158.72ms
step:37/1530 train_loss:5.4462 train_time:4286ms step_avg:158.75ms
step:38/1530 train_loss:5.2996 train_time:4448ms step_avg:158.87ms
step:39/1530 train_loss:5.3272 train_time:4608ms step_avg:158.89ms
step:40/1530 train_loss:5.2311 train_time:4769ms step_avg:158.95ms
step:41/1530 train_loss:5.2184 train_time:4930ms step_avg:159.02ms
step:42/1530 train_loss:5.1525 train_time:5090ms step_avg:159.07ms
step:43/1530 train_loss:5.2567 train_time:5250ms step_avg:159.10ms
step:44/1530 train_loss:5.2122 train_time:5410ms step_avg:159.13ms
step:45/1530 train_loss:5.3627 train_time:5571ms step_avg:159.17ms
step:46/1530 train_loss:5.1576 train_time:5731ms step_avg:159.20ms
step:47/1530 train_loss:5.0684 train_time:5891ms step_avg:159.22ms
step:48/1530 train_loss:5.2051 train_time:6053ms step_avg:159.29ms
step:49/1530 train_loss:5.1622 train_time:6213ms step_avg:159.31ms
step:50/1530 train_loss:5.2614 train_time:6373ms step_avg:159.32ms
step:51/1530 train_loss:5.1350 train_time:6534ms step_avg:159.36ms
step:52/1530 train_loss:5.0166 train_time:6694ms step_avg:159.39ms
step:53/1530 train_loss:5.1759 train_time:6855ms step_avg:159.42ms
step:54/1530 train_loss:5.0088 train_time:7015ms step_avg:159.43ms
step:55/1530 train_loss:5.3933 train_time:7175ms step_avg:159.44ms
step:56/1530 train_loss:5.0066 train_time:7335ms step_avg:159.47ms
step:57/1530 train_loss:4.8856 train_time:7496ms step_avg:159.50ms
step:58/1530 train_loss:5.0569 train_time:7657ms step_avg:159.52ms
step:59/1530 train_loss:5.0250 train_time:7818ms step_avg:159.55ms
step:60/1530 train_loss:5.1568 train_time:7978ms step_avg:159.57ms
step:61/1530 train_loss:4.8578 train_time:8138ms step_avg:159.56ms
step:62/1530 train_loss:4.9683 train_time:8298ms step_avg:159.58ms
step:63/1530 train_loss:4.9644 train_time:8460ms step_avg:159.62ms
step:64/1530 train_loss:4.9599 train_time:8620ms step_avg:159.63ms
step:65/1530 train_loss:4.8047 train_time:8780ms step_avg:159.64ms
step:66/1530 train_loss:4.9090 train_time:8942ms step_avg:159.67ms
step:67/1530 train_loss:4.8116 train_time:9102ms step_avg:159.69ms
step:68/1530 train_loss:5.0848 train_time:9263ms step_avg:159.71ms
step:69/1530 train_loss:4.7019 train_time:9424ms step_avg:159.72ms
step:70/1530 train_loss:4.8195 train_time:9584ms step_avg:159.73ms
step:71/1530 train_loss:4.9764 train_time:9744ms step_avg:159.74ms
step:72/1530 train_loss:4.8938 train_time:9904ms step_avg:159.74ms
step:73/1530 train_loss:4.7614 train_time:10064ms step_avg:159.75ms
step:74/1530 train_loss:4.9025 train_time:10224ms step_avg:159.76ms
step:75/1530 train_loss:4.8652 train_time:10385ms step_avg:159.77ms
step:76/1530 train_loss:4.8036 train_time:10546ms step_avg:159.78ms
step:77/1530 train_loss:4.9172 train_time:10705ms step_avg:159.78ms
step:78/1530 train_loss:5.1262 train_time:10867ms step_avg:159.81ms
step:79/1530 train_loss:4.8182 train_time:11028ms step_avg:159.82ms
step:80/1530 train_loss:4.8575 train_time:11188ms step_avg:159.82ms
step:81/1530 train_loss:4.6597 train_time:11348ms step_avg:159.83ms
step:82/1530 train_loss:4.8303 train_time:11509ms step_avg:159.85ms
step:83/1530 train_loss:4.7794 train_time:11670ms step_avg:159.87ms
step:84/1530 train_loss:4.7616 train_time:11831ms step_avg:159.88ms
step:85/1530 train_loss:4.6170 train_time:11992ms step_avg:159.89ms
step:86/1530 train_loss:4.8245 train_time:12152ms step_avg:159.90ms
step:87/1530 train_loss:4.7423 train_time:12313ms step_avg:159.91ms
step:88/1530 train_loss:4.7531 train_time:12472ms step_avg:159.90ms
step:89/1530 train_loss:4.7034 train_time:12634ms step_avg:159.92ms
step:90/1530 train_loss:4.6414 train_time:12795ms step_avg:159.93ms
step:91/1530 train_loss:4.6354 train_time:12955ms step_avg:159.94ms
step:92/1530 train_loss:4.7818 train_time:13115ms step_avg:159.94ms
step:93/1530 train_loss:4.6018 train_time:13275ms step_avg:159.94ms
step:94/1530 train_loss:4.6346 train_time:13436ms step_avg:159.95ms
step:95/1530 train_loss:4.6842 train_time:13595ms step_avg:159.94ms
step:96/1530 train_loss:4.5942 train_time:13756ms step_avg:159.96ms
step:97/1530 train_loss:4.6583 train_time:13917ms step_avg:159.96ms
step:98/1530 train_loss:4.5926 train_time:14076ms step_avg:159.95ms
step:99/1530 train_loss:4.6812 train_time:14237ms step_avg:159.96ms
step:100/1530 train_loss:4.6859 train_time:14398ms step_avg:159.98ms
step:101/1530 train_loss:4.5330 train_time:14559ms step_avg:159.99ms
step:102/1530 train_loss:4.6914 train_time:14722ms step_avg:160.02ms
step:103/1530 train_loss:4.5806 train_time:14883ms step_avg:160.03ms
step:104/1530 train_loss:4.5594 train_time:15043ms step_avg:160.04ms
step:105/1530 train_loss:4.5526 train_time:15203ms step_avg:160.03ms
step:106/1530 train_loss:4.6185 train_time:15364ms step_avg:160.05ms
step:107/1530 train_loss:4.5150 train_time:15525ms step_avg:160.05ms
step:108/1530 train_loss:4.3654 train_time:15685ms step_avg:160.05ms
step:109/1530 train_loss:4.4820 train_time:15845ms step_avg:160.05ms
step:110/1530 train_loss:4.4834 train_time:16005ms step_avg:160.05ms
step:111/1530 train_loss:4.4255 train_time:16167ms step_avg:160.06ms
step:112/1530 train_loss:4.5830 train_time:16327ms step_avg:160.07ms
step:113/1530 train_loss:4.4907 train_time:16487ms step_avg:160.07ms
step:114/1530 train_loss:4.3600 train_time:16648ms step_avg:160.08ms
step:115/1530 train_loss:4.5142 train_time:16811ms step_avg:160.10ms
step:116/1530 train_loss:4.4741 train_time:16976ms step_avg:160.15ms
step:117/1530 train_loss:4.3801 train_time:17141ms step_avg:160.19ms
step:118/1530 train_loss:4.5980 train_time:17304ms step_avg:160.22ms
step:119/1530 train_loss:4.4459 train_time:17468ms step_avg:160.26ms
step:120/1530 train_loss:4.3275 train_time:17632ms step_avg:160.29ms
step:121/1530 train_loss:4.2912 train_time:17797ms step_avg:160.33ms
step:122/1530 train_loss:4.4461 train_time:17962ms step_avg:160.37ms
step:123/1530 train_loss:4.2736 train_time:18126ms step_avg:160.41ms
step:124/1530 train_loss:4.5898 train_time:18289ms step_avg:160.43ms
step:125/1530 train_loss:4.4568 train_time:18453ms step_avg:160.47ms
step:125/1530 val_loss:4.4012 train_time:18500ms step_avg:160.87ms
step:126/1530 train_loss:4.4064 train_time:18619ms step_avg:160.51ms
step:127/1530 train_loss:4.4313 train_time:18786ms step_avg:160.57ms
step:128/1530 train_loss:4.3712 train_time:18950ms step_avg:160.59ms
step:129/1530 train_loss:4.6804 train_time:19114ms step_avg:160.62ms
step:130/1530 train_loss:4.3686 train_time:19278ms step_avg:160.65ms
step:131/1530 train_loss:4.4002 train_time:19442ms step_avg:160.68ms
step:132/1530 train_loss:4.3353 train_time:19606ms step_avg:160.71ms
step:133/1530 train_loss:4.4388 train_time:19771ms step_avg:160.74ms
step:134/1530 train_loss:4.2572 train_time:19936ms step_avg:160.77ms
step:135/1530 train_loss:4.4508 train_time:20100ms step_avg:160.80ms
step:136/1530 train_loss:4.2192 train_time:20264ms step_avg:160.83ms
step:137/1530 train_loss:4.3627 train_time:20429ms step_avg:160.86ms
step:138/1530 train_loss:4.2801 train_time:20594ms step_avg:160.89ms
step:139/1530 train_loss:4.3795 train_time:20758ms step_avg:160.92ms
step:140/1530 train_loss:4.4725 train_time:20923ms step_avg:160.94ms
step:141/1530 train_loss:4.3051 train_time:21087ms step_avg:160.97ms
step:142/1530 train_loss:4.2983 train_time:21251ms step_avg:160.99ms
step:143/1530 train_loss:4.2456 train_time:21416ms step_avg:161.02ms
step:144/1530 train_loss:4.3522 train_time:21580ms step_avg:161.04ms
step:145/1530 train_loss:4.3108 train_time:21743ms step_avg:161.06ms
step:146/1530 train_loss:4.1725 train_time:21908ms step_avg:161.08ms
step:147/1530 train_loss:4.3220 train_time:22072ms step_avg:161.11ms
step:148/1530 train_loss:4.3606 train_time:22236ms step_avg:161.13ms
step:149/1530 train_loss:4.2971 train_time:22401ms step_avg:161.16ms
step:150/1530 train_loss:4.4363 train_time:22564ms step_avg:161.17ms
step:151/1530 train_loss:4.2557 train_time:22729ms step_avg:161.20ms
step:152/1530 train_loss:4.2623 train_time:22894ms step_avg:161.22ms
step:153/1530 train_loss:4.3766 train_time:23059ms step_avg:161.25ms
step:154/1530 train_loss:4.3671 train_time:23222ms step_avg:161.26ms
step:155/1530 train_loss:4.2723 train_time:23386ms step_avg:161.28ms
step:156/1530 train_loss:4.3486 train_time:23550ms step_avg:161.30ms
step:157/1530 train_loss:4.3973 train_time:23714ms step_avg:161.32ms
step:158/1530 train_loss:4.2340 train_time:23877ms step_avg:161.33ms
step:159/1530 train_loss:4.2998 train_time:24041ms step_avg:161.35ms
step:160/1530 train_loss:4.1253 train_time:24205ms step_avg:161.37ms
step:161/1530 train_loss:4.3435 train_time:24370ms step_avg:161.39ms
step:162/1530 train_loss:4.3517 train_time:24534ms step_avg:161.41ms
step:163/1530 train_loss:4.3271 train_time:24697ms step_avg:161.42ms
step:164/1530 train_loss:4.1794 train_time:24860ms step_avg:161.43ms
step:165/1530 train_loss:4.2814 train_time:25025ms step_avg:161.45ms
step:166/1530 train_loss:4.3459 train_time:25189ms step_avg:161.47ms
step:167/1530 train_loss:4.1979 train_time:25353ms step_avg:161.48ms
step:168/1530 train_loss:4.2877 train_time:25516ms step_avg:161.50ms
step:169/1530 train_loss:4.1579 train_time:25679ms step_avg:161.50ms
step:170/1530 train_loss:4.0152 train_time:25843ms step_avg:161.52ms
step:171/1530 train_loss:4.1912 train_time:26007ms step_avg:161.53ms
step:172/1530 train_loss:4.2138 train_time:26170ms step_avg:161.54ms
step:173/1530 train_loss:4.2594 train_time:26334ms step_avg:161.56ms
step:174/1530 train_loss:4.4070 train_time:26496ms step_avg:161.56ms
step:175/1530 train_loss:4.2380 train_time:26659ms step_avg:161.57ms
step:176/1530 train_loss:4.0865 train_time:26821ms step_avg:161.57ms
step:177/1530 train_loss:4.0582 train_time:26984ms step_avg:161.58ms
step:178/1530 train_loss:4.1698 train_time:27148ms step_avg:161.60ms
step:179/1530 train_loss:4.1196 train_time:27312ms step_avg:161.61ms
step:180/1530 train_loss:4.1189 train_time:27475ms step_avg:161.62ms
step:181/1530 train_loss:4.2942 train_time:27638ms step_avg:161.63ms
step:182/1530 train_loss:4.1618 train_time:27801ms step_avg:161.63ms
step:183/1530 train_loss:4.1140 train_time:27965ms step_avg:161.65ms
step:184/1530 train_loss:4.1183 train_time:28128ms step_avg:161.66ms
step:185/1530 train_loss:4.1903 train_time:28291ms step_avg:161.66ms
step:186/1530 train_loss:4.1657 train_time:28453ms step_avg:161.66ms
step:187/1530 train_loss:4.2305 train_time:28616ms step_avg:161.67ms
step:188/1530 train_loss:4.1606 train_time:28913ms step_avg:162.43ms
step:189/1530 train_loss:4.1057 train_time:29236ms step_avg:163.33ms
step:190/1530 train_loss:4.2055 train_time:29398ms step_avg:163.32ms
step:191/1530 train_loss:4.0766 train_time:29561ms step_avg:163.32ms
step:192/1530 train_loss:4.0267 train_time:29723ms step_avg:163.32ms
step:193/1530 train_loss:4.2441 train_time:29886ms step_avg:163.31ms
step:194/1530 train_loss:4.1647 train_time:30049ms step_avg:163.31ms
step:195/1530 train_loss:4.3425 train_time:30213ms step_avg:163.31ms
step:196/1530 train_loss:4.1727 train_time:30376ms step_avg:163.31ms
step:197/1530 train_loss:4.0364 train_time:30538ms step_avg:163.31ms
step:198/1530 train_loss:4.1724 train_time:30700ms step_avg:163.30ms
step:199/1530 train_loss:4.0318 train_time:30863ms step_avg:163.30ms
step:200/1530 train_loss:4.1119 train_time:31027ms step_avg:163.30ms
step:201/1530 train_loss:4.0095 train_time:31188ms step_avg:163.29ms
step:202/1530 train_loss:4.2566 train_time:31351ms step_avg:163.29ms
step:203/1530 train_loss:4.0708 train_time:31515ms step_avg:163.29ms
step:204/1530 train_loss:4.1916 train_time:31677ms step_avg:163.29ms
step:205/1530 train_loss:4.2389 train_time:31840ms step_avg:163.28ms
step:206/1530 train_loss:3.9407 train_time:32003ms step_avg:163.28ms
step:207/1530 train_loss:4.0654 train_time:32166ms step_avg:163.28ms
step:208/1530 train_loss:4.0884 train_time:32331ms step_avg:163.29ms
step:209/1530 train_loss:4.2310 train_time:32494ms step_avg:163.29ms
step:210/1530 train_loss:4.1699 train_time:32658ms step_avg:163.29ms
step:211/1530 train_loss:4.0581 train_time:32820ms step_avg:163.28ms
step:212/1530 train_loss:4.1243 train_time:32983ms step_avg:163.28ms
step:213/1530 train_loss:4.0383 train_time:33147ms step_avg:163.29ms
step:214/1530 train_loss:4.1071 train_time:33310ms step_avg:163.29ms
step:215/1530 train_loss:3.9465 train_time:33473ms step_avg:163.28ms
step:216/1530 train_loss:3.9956 train_time:33636ms step_avg:163.28ms
step:217/1530 train_loss:4.0028 train_time:33799ms step_avg:163.28ms
step:218/1530 train_loss:4.0802 train_time:33961ms step_avg:163.28ms
step:219/1530 train_loss:4.0672 train_time:34124ms step_avg:163.27ms
step:220/1530 train_loss:4.0735 train_time:34287ms step_avg:163.27ms
step:221/1530 train_loss:4.0979 train_time:34450ms step_avg:163.27ms
step:222/1530 train_loss:4.0052 train_time:34613ms step_avg:163.27ms
step:223/1530 train_loss:3.9859 train_time:34776ms step_avg:163.27ms
step:224/1530 train_loss:4.2958 train_time:34939ms step_avg:163.27ms
step:225/1530 train_loss:3.9261 train_time:35101ms step_avg:163.26ms
step:226/1530 train_loss:3.9834 train_time:35263ms step_avg:163.25ms
step:227/1530 train_loss:3.9866 train_time:35428ms step_avg:163.26ms
step:228/1530 train_loss:4.1343 train_time:35594ms step_avg:163.27ms
step:229/1530 train_loss:3.9196 train_time:35760ms step_avg:163.29ms
step:230/1530 train_loss:4.0376 train_time:35925ms step_avg:163.30ms
step:231/1530 train_loss:3.8971 train_time:36091ms step_avg:163.31ms
step:232/1530 train_loss:3.9680 train_time:36257ms step_avg:163.32ms
step:233/1530 train_loss:4.0864 train_time:36423ms step_avg:163.33ms
step:234/1530 train_loss:4.0258 train_time:36588ms step_avg:163.34ms
step:235/1530 train_loss:3.8979 train_time:36757ms step_avg:163.37ms
step:236/1530 train_loss:4.0770 train_time:36923ms step_avg:163.38ms
step:237/1530 train_loss:4.0777 train_time:37090ms step_avg:163.39ms
step:238/1530 train_loss:3.9406 train_time:37257ms step_avg:163.41ms
step:239/1530 train_loss:4.0757 train_time:37423ms step_avg:163.42ms
step:240/1530 train_loss:4.1053 train_time:37590ms step_avg:163.43ms
step:241/1530 train_loss:3.9603 train_time:37755ms step_avg:163.44ms
step:242/1530 train_loss:4.1401 train_time:37921ms step_avg:163.45ms
step:243/1530 train_loss:4.0031 train_time:38087ms step_avg:163.46ms
step:244/1530 train_loss:4.0793 train_time:38253ms step_avg:163.47ms
step:245/1530 train_loss:4.1363 train_time:38419ms step_avg:163.48ms
step:246/1530 train_loss:4.0545 train_time:38583ms step_avg:163.49ms
step:247/1530 train_loss:4.0019 train_time:38750ms step_avg:163.50ms
step:248/1530 train_loss:4.0983 train_time:38918ms step_avg:163.52ms
step:249/1530 train_loss:3.9144 train_time:39082ms step_avg:163.52ms
step:250/1530 train_loss:3.9742 train_time:39249ms step_avg:163.54ms
step:250/1530 val_loss:4.0030 train_time:39298ms step_avg:163.74ms
step:251/1530 train_loss:4.0742 train_time:39419ms step_avg:163.56ms
step:252/1530 train_loss:4.1657 train_time:39586ms step_avg:163.58ms
step:253/1530 train_loss:3.9309 train_time:39753ms step_avg:163.59ms
step:254/1530 train_loss:3.8772 train_time:39919ms step_avg:163.60ms
step:255/1530 train_loss:4.0701 train_time:40085ms step_avg:163.61ms
step:256/1530 train_loss:3.9905 train_time:40253ms step_avg:163.63ms
step:257/1530 train_loss:3.9995 train_time:40418ms step_avg:163.64ms
step:258/1530 train_loss:3.9913 train_time:40584ms step_avg:163.65ms
step:259/1530 train_loss:4.0292 train_time:40751ms step_avg:163.66ms
step:260/1530 train_loss:4.0585 train_time:40917ms step_avg:163.67ms
step:261/1530 train_loss:4.0189 train_time:41083ms step_avg:163.68ms
step:262/1530 train_loss:3.9849 train_time:41249ms step_avg:163.69ms
step:263/1530 train_loss:3.8886 train_time:41416ms step_avg:163.70ms
step:264/1530 train_loss:3.9811 train_time:41582ms step_avg:163.71ms
step:265/1530 train_loss:3.8645 train_time:41749ms step_avg:163.72ms
step:266/1530 train_loss:3.9142 train_time:41915ms step_avg:163.73ms
step:267/1530 train_loss:3.9234 train_time:42081ms step_avg:163.74ms
step:268/1530 train_loss:3.9621 train_time:42246ms step_avg:163.75ms
step:269/1530 train_loss:3.8485 train_time:42412ms step_avg:163.75ms
step:270/1530 train_loss:4.0967 train_time:42579ms step_avg:163.76ms
step:271/1530 train_loss:3.9692 train_time:42744ms step_avg:163.77ms
step:272/1530 train_loss:3.9304 train_time:42911ms step_avg:163.78ms
step:273/1530 train_loss:3.9455 train_time:43077ms step_avg:163.79ms
step:274/1530 train_loss:4.0416 train_time:43243ms step_avg:163.80ms
step:275/1530 train_loss:4.0655 train_time:43409ms step_avg:163.81ms
step:276/1530 train_loss:4.2267 train_time:43577ms step_avg:163.82ms
step:277/1530 train_loss:4.0354 train_time:43742ms step_avg:163.83ms
step:278/1530 train_loss:4.0833 train_time:43909ms step_avg:163.84ms
step:279/1530 train_loss:3.9965 train_time:44076ms step_avg:163.85ms
step:280/1530 train_loss:4.1975 train_time:44242ms step_avg:163.86ms
step:281/1530 train_loss:3.9716 train_time:44408ms step_avg:163.87ms
step:282/1530 train_loss:3.9500 train_time:44576ms step_avg:163.88ms
step:283/1530 train_loss:3.9178 train_time:44741ms step_avg:163.89ms
step:284/1530 train_loss:4.0504 train_time:44908ms step_avg:163.90ms
step:285/1530 train_loss:4.0650 train_time:45074ms step_avg:163.91ms
step:286/1530 train_loss:4.0946 train_time:45238ms step_avg:163.91ms
step:287/1530 train_loss:3.9123 train_time:45403ms step_avg:163.91ms
step:288/1530 train_loss:4.0129 train_time:45568ms step_avg:163.91ms
step:289/1530 train_loss:3.8763 train_time:45734ms step_avg:163.92ms
step:290/1530 train_loss:3.8585 train_time:45900ms step_avg:163.93ms
step:291/1530 train_loss:3.9084 train_time:46064ms step_avg:163.93ms
step:292/1530 train_loss:3.8650 train_time:46230ms step_avg:163.94ms
step:293/1530 train_loss:3.9073 train_time:46396ms step_avg:163.94ms
step:294/1530 train_loss:3.9407 train_time:46561ms step_avg:163.95ms
step:295/1530 train_loss:3.8371 train_time:46727ms step_avg:163.95ms
step:296/1530 train_loss:3.8654 train_time:46894ms step_avg:163.97ms
step:297/1530 train_loss:3.8598 train_time:47059ms step_avg:163.97ms
step:298/1530 train_loss:3.9709 train_time:47224ms step_avg:163.97ms
step:299/1530 train_loss:3.8254 train_time:47389ms step_avg:163.97ms
step:300/1530 train_loss:3.9731 train_time:47554ms step_avg:163.98ms
step:301/1530 train_loss:3.9578 train_time:47720ms step_avg:163.99ms
step:302/1530 train_loss:3.9354 train_time:47885ms step_avg:163.99ms
step:303/1530 train_loss:3.9812 train_time:48050ms step_avg:163.99ms
step:304/1530 train_loss:3.9656 train_time:48216ms step_avg:164.00ms
step:305/1530 train_loss:4.4519 train_time:48381ms step_avg:164.00ms
step:306/1530 train_loss:3.9376 train_time:48546ms step_avg:164.01ms
step:307/1530 train_loss:3.8336 train_time:48712ms step_avg:164.01ms
step:308/1530 train_loss:3.9876 train_time:48877ms step_avg:164.02ms
step:309/1530 train_loss:3.8706 train_time:49041ms step_avg:164.02ms
step:310/1530 train_loss:4.0833 train_time:49206ms step_avg:164.02ms
step:311/1530 train_loss:3.9308 train_time:49372ms step_avg:164.03ms
step:312/1530 train_loss:3.8614 train_time:49538ms step_avg:164.03ms
step:313/1530 train_loss:3.9391 train_time:49704ms step_avg:164.04ms
step:314/1530 train_loss:4.0635 train_time:49870ms step_avg:164.05ms
step:315/1530 train_loss:3.9416 train_time:50035ms step_avg:164.05ms
step:316/1530 train_loss:3.7897 train_time:50201ms step_avg:164.05ms
step:317/1530 train_loss:3.8792 train_time:50366ms step_avg:164.06ms
step:318/1530 train_loss:3.9222 train_time:50531ms step_avg:164.06ms
step:319/1530 train_loss:3.8922 train_time:50697ms step_avg:164.07ms
step:320/1530 train_loss:4.0196 train_time:50862ms step_avg:164.07ms
step:321/1530 train_loss:3.9568 train_time:51026ms step_avg:164.07ms
step:322/1530 train_loss:3.9350 train_time:51194ms step_avg:164.08ms
step:323/1530 train_loss:4.0068 train_time:51360ms step_avg:164.09ms
step:324/1530 train_loss:3.9491 train_time:51526ms step_avg:164.09ms
step:325/1530 train_loss:4.0197 train_time:51692ms step_avg:164.10ms
step:326/1530 train_loss:3.9012 train_time:51857ms step_avg:164.10ms
step:327/1530 train_loss:4.4027 train_time:52023ms step_avg:164.11ms
step:328/1530 train_loss:4.0770 train_time:52188ms step_avg:164.11ms
step:329/1530 train_loss:3.7980 train_time:52353ms step_avg:164.11ms
step:330/1530 train_loss:3.7450 train_time:52518ms step_avg:164.12ms
step:331/1530 train_loss:3.9792 train_time:52683ms step_avg:164.12ms
step:332/1530 train_loss:3.9177 train_time:52848ms step_avg:164.12ms
step:333/1530 train_loss:3.8874 train_time:53014ms step_avg:164.13ms
step:334/1530 train_loss:3.8391 train_time:53180ms step_avg:164.14ms
step:335/1530 train_loss:4.0088 train_time:53345ms step_avg:164.14ms
step:336/1530 train_loss:3.9676 train_time:53511ms step_avg:164.14ms
step:337/1530 train_loss:4.4283 train_time:53677ms step_avg:164.15ms
step:338/1530 train_loss:3.9423 train_time:53842ms step_avg:164.15ms
step:339/1530 train_loss:3.8674 train_time:54008ms step_avg:164.16ms
step:340/1530 train_loss:3.9365 train_time:54174ms step_avg:164.16ms
step:341/1530 train_loss:3.8595 train_time:54340ms step_avg:164.17ms
step:342/1530 train_loss:3.8098 train_time:54508ms step_avg:164.18ms
step:343/1530 train_loss:3.8388 train_time:54677ms step_avg:164.20ms
step:344/1530 train_loss:3.9993 train_time:54844ms step_avg:164.20ms
step:345/1530 train_loss:3.8132 train_time:55014ms step_avg:164.22ms
step:346/1530 train_loss:3.7683 train_time:55182ms step_avg:164.23ms
step:347/1530 train_loss:3.7955 train_time:55350ms step_avg:164.24ms
step:348/1530 train_loss:3.8587 train_time:55518ms step_avg:164.25ms
step:349/1530 train_loss:3.8306 train_time:55686ms step_avg:164.26ms
step:350/1530 train_loss:3.5719 train_time:55855ms step_avg:164.28ms
step:351/1530 train_loss:3.8275 train_time:56023ms step_avg:164.29ms
step:352/1530 train_loss:4.1890 train_time:56192ms step_avg:164.31ms
step:353/1530 train_loss:3.6631 train_time:56360ms step_avg:164.32ms
step:354/1530 train_loss:3.9306 train_time:56527ms step_avg:164.32ms
step:355/1530 train_loss:3.7854 train_time:56697ms step_avg:164.34ms
step:356/1530 train_loss:3.8817 train_time:56865ms step_avg:164.35ms
step:357/1530 train_loss:3.7622 train_time:57033ms step_avg:164.36ms
step:358/1530 train_loss:3.8680 train_time:57202ms step_avg:164.37ms
step:359/1530 train_loss:3.8012 train_time:57371ms step_avg:164.39ms
step:360/1530 train_loss:3.4198 train_time:57540ms step_avg:164.40ms
step:361/1530 train_loss:4.0292 train_time:57709ms step_avg:164.41ms
step:362/1530 train_loss:3.9218 train_time:57878ms step_avg:164.43ms
step:363/1530 train_loss:3.8413 train_time:58044ms step_avg:164.43ms
step:364/1530 train_loss:3.7536 train_time:58213ms step_avg:164.44ms
step:365/1530 train_loss:3.9184 train_time:58381ms step_avg:164.45ms
step:366/1530 train_loss:3.8680 train_time:58548ms step_avg:164.46ms
step:367/1530 train_loss:3.8638 train_time:58717ms step_avg:164.47ms
step:368/1530 train_loss:3.8572 train_time:58884ms step_avg:164.48ms
step:369/1530 train_loss:3.7507 train_time:59052ms step_avg:164.49ms
step:370/1530 train_loss:3.8797 train_time:59220ms step_avg:164.50ms
step:371/1530 train_loss:3.7302 train_time:59388ms step_avg:164.51ms
step:372/1530 train_loss:3.6931 train_time:59558ms step_avg:164.52ms
step:373/1530 train_loss:3.9106 train_time:59726ms step_avg:164.53ms
step:374/1530 train_loss:3.8277 train_time:59895ms step_avg:164.55ms
step:375/1530 train_loss:3.8069 train_time:60062ms step_avg:164.55ms
step:375/1530 val_loss:3.8269 train_time:60110ms step_avg:164.69ms
step:376/1530 train_loss:3.8640 train_time:60231ms step_avg:164.57ms
step:377/1530 train_loss:3.7914 train_time:60532ms step_avg:164.94ms
step:378/1530 train_loss:3.8564 train_time:60710ms step_avg:164.97ms
step:379/1530 train_loss:3.8725 train_time:61025ms step_avg:165.38ms
step:380/1530 train_loss:3.9633 train_time:61192ms step_avg:165.38ms
step:381/1530 train_loss:3.8459 train_time:61360ms step_avg:165.39ms
step:382/1530 train_loss:3.8070 train_time:61528ms step_avg:165.40ms
step:383/1530 train_loss:3.7934 train_time:61697ms step_avg:165.41ms
step:384/1530 train_loss:3.8725 train_time:61864ms step_avg:165.41ms
step:385/1530 train_loss:3.7976 train_time:62032ms step_avg:165.42ms
step:386/1530 train_loss:3.8934 train_time:62200ms step_avg:165.43ms
step:387/1530 train_loss:4.0535 train_time:62367ms step_avg:165.43ms
step:388/1530 train_loss:3.7952 train_time:62535ms step_avg:165.44ms
step:389/1530 train_loss:3.7974 train_time:62704ms step_avg:165.44ms
step:390/1530 train_loss:3.8986 train_time:62872ms step_avg:165.45ms
step:391/1530 train_loss:3.8129 train_time:63040ms step_avg:165.46ms
step:392/1530 train_loss:3.9243 train_time:63207ms step_avg:165.46ms
step:393/1530 train_loss:3.7657 train_time:63375ms step_avg:165.47ms
step:394/1530 train_loss:3.8900 train_time:63543ms step_avg:165.48ms
step:395/1530 train_loss:3.6378 train_time:63710ms step_avg:165.48ms
step:396/1530 train_loss:3.8379 train_time:63879ms step_avg:165.49ms
step:397/1530 train_loss:3.8625 train_time:64047ms step_avg:165.50ms
step:398/1530 train_loss:3.8719 train_time:64216ms step_avg:165.51ms
step:399/1530 train_loss:3.7715 train_time:64382ms step_avg:165.51ms
step:400/1530 train_loss:3.8346 train_time:64549ms step_avg:165.51ms
step:401/1530 train_loss:3.9177 train_time:64717ms step_avg:165.52ms
step:402/1530 train_loss:3.8474 train_time:64883ms step_avg:165.52ms
step:403/1530 train_loss:3.9670 train_time:65051ms step_avg:165.52ms
step:404/1530 train_loss:3.6813 train_time:65220ms step_avg:165.53ms
step:405/1530 train_loss:3.7897 train_time:65387ms step_avg:165.54ms
step:406/1530 train_loss:4.0993 train_time:65554ms step_avg:165.54ms
step:407/1530 train_loss:3.7742 train_time:65723ms step_avg:165.55ms
step:408/1530 train_loss:3.8247 train_time:65890ms step_avg:165.55ms
step:409/1530 train_loss:3.8576 train_time:66057ms step_avg:165.56ms
step:410/1530 train_loss:3.7555 train_time:66224ms step_avg:165.56ms
step:411/1530 train_loss:3.7640 train_time:66391ms step_avg:165.56ms
step:412/1530 train_loss:4.1817 train_time:66559ms step_avg:165.57ms
step:413/1530 train_loss:3.7332 train_time:66725ms step_avg:165.57ms
step:414/1530 train_loss:4.0114 train_time:66892ms step_avg:165.57ms
step:415/1530 train_loss:3.7523 train_time:67060ms step_avg:165.58ms
step:416/1530 train_loss:3.7629 train_time:67227ms step_avg:165.58ms
step:417/1530 train_loss:3.9543 train_time:67395ms step_avg:165.59ms
step:418/1530 train_loss:3.6919 train_time:67562ms step_avg:165.59ms
step:419/1530 train_loss:3.8079 train_time:67729ms step_avg:165.60ms
step:420/1530 train_loss:3.7059 train_time:67896ms step_avg:165.60ms
step:421/1530 train_loss:3.6509 train_time:68062ms step_avg:165.60ms
step:422/1530 train_loss:3.7871 train_time:68229ms step_avg:165.60ms
step:423/1530 train_loss:3.8824 train_time:68395ms step_avg:165.61ms
step:424/1530 train_loss:3.6220 train_time:68562ms step_avg:165.61ms
step:425/1530 train_loss:3.7914 train_time:68730ms step_avg:165.61ms
step:426/1530 train_loss:3.6610 train_time:68898ms step_avg:165.62ms
step:427/1530 train_loss:3.8957 train_time:69065ms step_avg:165.62ms
step:428/1530 train_loss:3.8113 train_time:69232ms step_avg:165.63ms
step:429/1530 train_loss:3.7613 train_time:69400ms step_avg:165.63ms
step:430/1530 train_loss:3.7102 train_time:69568ms step_avg:165.64ms
step:431/1530 train_loss:3.6285 train_time:69738ms step_avg:165.65ms
step:432/1530 train_loss:3.7668 train_time:69904ms step_avg:165.65ms
step:433/1530 train_loss:3.8170 train_time:70070ms step_avg:165.65ms
step:434/1530 train_loss:3.7804 train_time:70240ms step_avg:165.66ms
step:435/1530 train_loss:3.8141 train_time:70406ms step_avg:165.66ms
step:436/1530 train_loss:3.8383 train_time:70572ms step_avg:165.66ms
step:437/1530 train_loss:3.7224 train_time:70741ms step_avg:165.67ms
step:438/1530 train_loss:3.7079 train_time:70908ms step_avg:165.67ms
step:439/1530 train_loss:3.7201 train_time:71076ms step_avg:165.68ms
step:440/1530 train_loss:3.8929 train_time:71243ms step_avg:165.68ms
step:441/1530 train_loss:3.7617 train_time:71410ms step_avg:165.68ms
step:442/1530 train_loss:3.7427 train_time:71579ms step_avg:165.69ms
step:443/1530 train_loss:3.6219 train_time:71746ms step_avg:165.69ms
step:444/1530 train_loss:3.9225 train_time:71912ms step_avg:165.70ms
step:445/1530 train_loss:3.8479 train_time:72079ms step_avg:165.70ms
step:446/1530 train_loss:3.8362 train_time:72246ms step_avg:165.70ms
step:447/1530 train_loss:3.7538 train_time:72414ms step_avg:165.71ms
step:448/1530 train_loss:3.8483 train_time:72580ms step_avg:165.71ms
step:449/1530 train_loss:3.6896 train_time:72747ms step_avg:165.71ms
step:450/1530 train_loss:3.7260 train_time:72915ms step_avg:165.72ms
step:451/1530 train_loss:3.5816 train_time:73081ms step_avg:165.72ms
step:452/1530 train_loss:3.7182 train_time:73249ms step_avg:165.72ms
step:453/1530 train_loss:3.6726 train_time:73417ms step_avg:165.73ms
step:454/1530 train_loss:3.6449 train_time:73583ms step_avg:165.73ms
step:455/1530 train_loss:3.8384 train_time:73752ms step_avg:165.73ms
step:456/1530 train_loss:3.7280 train_time:73922ms step_avg:165.74ms
step:457/1530 train_loss:3.7874 train_time:74092ms step_avg:165.75ms
step:458/1530 train_loss:3.8286 train_time:74261ms step_avg:165.76ms
step:459/1530 train_loss:3.6365 train_time:74431ms step_avg:165.77ms
step:460/1530 train_loss:3.7920 train_time:74600ms step_avg:165.78ms
step:461/1530 train_loss:3.6943 train_time:74770ms step_avg:165.79ms
step:462/1530 train_loss:3.7424 train_time:74942ms step_avg:165.80ms
step:463/1530 train_loss:3.7821 train_time:75112ms step_avg:165.81ms
step:464/1530 train_loss:3.7182 train_time:75282ms step_avg:165.82ms
step:465/1530 train_loss:3.7213 train_time:75450ms step_avg:165.82ms
step:466/1530 train_loss:3.7969 train_time:75621ms step_avg:165.83ms
step:467/1530 train_loss:3.8242 train_time:75790ms step_avg:165.84ms
step:468/1530 train_loss:3.7983 train_time:75959ms step_avg:165.85ms
step:469/1530 train_loss:3.6921 train_time:76129ms step_avg:165.86ms
step:470/1530 train_loss:3.7711 train_time:76300ms step_avg:165.87ms
step:471/1530 train_loss:3.8090 train_time:76471ms step_avg:165.88ms
step:472/1530 train_loss:3.7822 train_time:76643ms step_avg:165.89ms
step:473/1530 train_loss:3.7176 train_time:76813ms step_avg:165.90ms
step:474/1530 train_loss:3.5947 train_time:76982ms step_avg:165.91ms
step:475/1530 train_loss:4.0257 train_time:77151ms step_avg:165.92ms
step:476/1530 train_loss:3.7627 train_time:77323ms step_avg:165.93ms
step:477/1530 train_loss:3.5963 train_time:77492ms step_avg:165.94ms
step:478/1530 train_loss:3.8278 train_time:77662ms step_avg:165.95ms
step:479/1530 train_loss:3.7738 train_time:77833ms step_avg:165.95ms
step:480/1530 train_loss:3.9205 train_time:78003ms step_avg:165.96ms
step:481/1530 train_loss:3.7283 train_time:78172ms step_avg:165.97ms
step:482/1530 train_loss:3.5295 train_time:78343ms step_avg:165.98ms
step:483/1530 train_loss:3.8096 train_time:78512ms step_avg:165.99ms
step:484/1530 train_loss:3.6633 train_time:78682ms step_avg:166.00ms
step:485/1530 train_loss:3.6542 train_time:78852ms step_avg:166.00ms
step:486/1530 train_loss:3.5759 train_time:79023ms step_avg:166.02ms
step:487/1530 train_loss:3.6864 train_time:79192ms step_avg:166.02ms
step:488/1530 train_loss:3.8833 train_time:79363ms step_avg:166.03ms
step:489/1530 train_loss:3.7193 train_time:79533ms step_avg:166.04ms
step:490/1530 train_loss:3.6008 train_time:79702ms step_avg:166.05ms
step:491/1530 train_loss:3.6210 train_time:79869ms step_avg:166.05ms
step:492/1530 train_loss:3.7374 train_time:80042ms step_avg:166.06ms
step:493/1530 train_loss:3.5777 train_time:80212ms step_avg:166.07ms
step:494/1530 train_loss:3.7016 train_time:80382ms step_avg:166.08ms
step:495/1530 train_loss:3.6632 train_time:80551ms step_avg:166.08ms
step:496/1530 train_loss:3.5215 train_time:80725ms step_avg:166.10ms
step:497/1530 train_loss:3.7404 train_time:80894ms step_avg:166.11ms
step:498/1530 train_loss:3.7884 train_time:81064ms step_avg:166.11ms
step:499/1530 train_loss:3.8230 train_time:81235ms step_avg:166.12ms
step:500/1530 train_loss:3.7349 train_time:81405ms step_avg:166.13ms
step:500/1530 val_loss:3.7067 train_time:81453ms step_avg:166.23ms
step:501/1530 train_loss:3.8066 train_time:81576ms step_avg:166.14ms
step:502/1530 train_loss:3.7566 train_time:81750ms step_avg:166.16ms
step:503/1530 train_loss:3.7789 train_time:81919ms step_avg:166.16ms
step:504/1530 train_loss:3.7206 train_time:82087ms step_avg:166.17ms
step:505/1530 train_loss:3.8043 train_time:82256ms step_avg:166.17ms
step:506/1530 train_loss:3.6537 train_time:82426ms step_avg:166.18ms
step:507/1530 train_loss:3.7655 train_time:82594ms step_avg:166.19ms
step:508/1530 train_loss:3.8225 train_time:82766ms step_avg:166.20ms
step:509/1530 train_loss:3.7733 train_time:82936ms step_avg:166.20ms
step:510/1530 train_loss:3.5851 train_time:83106ms step_avg:166.21ms
step:511/1530 train_loss:3.7800 train_time:83275ms step_avg:166.22ms
step:512/1530 train_loss:3.7256 train_time:83449ms step_avg:166.23ms
step:513/1530 train_loss:3.6710 train_time:83617ms step_avg:166.24ms
step:514/1530 train_loss:3.7648 train_time:83789ms step_avg:166.25ms
step:515/1530 train_loss:3.7322 train_time:83959ms step_avg:166.25ms
step:516/1530 train_loss:4.0749 train_time:84131ms step_avg:166.27ms
step:517/1530 train_loss:3.7034 train_time:84300ms step_avg:166.27ms
step:518/1530 train_loss:3.7660 train_time:84469ms step_avg:166.28ms
step:519/1530 train_loss:3.6555 train_time:84638ms step_avg:166.28ms
step:520/1530 train_loss:3.6862 train_time:84808ms step_avg:166.29ms
step:521/1530 train_loss:3.6639 train_time:84977ms step_avg:166.30ms
step:522/1530 train_loss:3.6553 train_time:85149ms step_avg:166.31ms
step:523/1530 train_loss:4.2824 train_time:85318ms step_avg:166.31ms
step:524/1530 train_loss:3.7400 train_time:85487ms step_avg:166.32ms
step:525/1530 train_loss:3.6856 train_time:85656ms step_avg:166.32ms
step:526/1530 train_loss:3.7014 train_time:85826ms step_avg:166.33ms
step:527/1530 train_loss:3.6602 train_time:85994ms step_avg:166.33ms
step:528/1530 train_loss:3.6290 train_time:86163ms step_avg:166.34ms
step:529/1530 train_loss:3.8525 train_time:86333ms step_avg:166.35ms
step:530/1530 train_loss:3.6537 train_time:86503ms step_avg:166.35ms
step:531/1530 train_loss:3.9257 train_time:86672ms step_avg:166.36ms
step:532/1530 train_loss:3.7342 train_time:86840ms step_avg:166.36ms
step:533/1530 train_loss:3.6517 train_time:87011ms step_avg:166.37ms
step:534/1530 train_loss:3.6726 train_time:87180ms step_avg:166.37ms
step:535/1530 train_loss:3.6121 train_time:87351ms step_avg:166.38ms
step:536/1530 train_loss:3.7519 train_time:87520ms step_avg:166.39ms
step:537/1530 train_loss:3.7317 train_time:87690ms step_avg:166.39ms
step:538/1530 train_loss:3.6337 train_time:87859ms step_avg:166.40ms
step:539/1530 train_loss:4.1214 train_time:88032ms step_avg:166.41ms
step:540/1530 train_loss:3.6852 train_time:88201ms step_avg:166.42ms
step:541/1530 train_loss:3.7901 train_time:88370ms step_avg:166.42ms
step:542/1530 train_loss:3.5937 train_time:88538ms step_avg:166.42ms
step:543/1530 train_loss:3.5815 train_time:88707ms step_avg:166.43ms
step:544/1530 train_loss:3.6305 train_time:88874ms step_avg:166.43ms
step:545/1530 train_loss:3.5946 train_time:89046ms step_avg:166.44ms
step:546/1530 train_loss:3.6384 train_time:89214ms step_avg:166.44ms
step:547/1530 train_loss:3.6446 train_time:89384ms step_avg:166.45ms
step:548/1530 train_loss:3.6188 train_time:89554ms step_avg:166.46ms
step:549/1530 train_loss:3.7227 train_time:89722ms step_avg:166.46ms
step:550/1530 train_loss:3.6209 train_time:89890ms step_avg:166.46ms
step:551/1530 train_loss:3.6333 train_time:90058ms step_avg:166.47ms
step:552/1530 train_loss:3.9375 train_time:90229ms step_avg:166.47ms
step:553/1530 train_loss:3.7624 train_time:90398ms step_avg:166.48ms
step:554/1530 train_loss:3.7152 train_time:90568ms step_avg:166.49ms
step:555/1530 train_loss:3.6337 train_time:90736ms step_avg:166.49ms
step:556/1530 train_loss:3.7026 train_time:90906ms step_avg:166.50ms
step:557/1530 train_loss:3.3058 train_time:91075ms step_avg:166.50ms
step:558/1530 train_loss:3.6150 train_time:91245ms step_avg:166.50ms
step:559/1530 train_loss:3.6471 train_time:91412ms step_avg:166.51ms
step:560/1530 train_loss:3.6909 train_time:91581ms step_avg:166.51ms
step:561/1530 train_loss:3.6161 train_time:91751ms step_avg:166.52ms
step:562/1530 train_loss:3.5565 train_time:91919ms step_avg:166.52ms
step:563/1530 train_loss:3.7567 train_time:92088ms step_avg:166.52ms
step:564/1530 train_loss:3.5753 train_time:92257ms step_avg:166.53ms
step:565/1530 train_loss:3.6867 train_time:92426ms step_avg:166.53ms
step:566/1530 train_loss:3.6261 train_time:92727ms step_avg:166.77ms
step:567/1530 train_loss:3.6025 train_time:92905ms step_avg:166.80ms
step:568/1530 train_loss:3.6852 train_time:93075ms step_avg:166.80ms
step:569/1530 train_loss:3.6472 train_time:93395ms step_avg:167.07ms
step:570/1530 train_loss:3.6862 train_time:93563ms step_avg:167.08ms
step:571/1530 train_loss:3.7597 train_time:93735ms step_avg:167.08ms
step:572/1530 train_loss:3.7279 train_time:93906ms step_avg:167.09ms
step:573/1530 train_loss:3.7339 train_time:94078ms step_avg:167.10ms
step:574/1530 train_loss:3.7804 train_time:94254ms step_avg:167.12ms
step:575/1530 train_loss:3.7343 train_time:94425ms step_avg:167.12ms
step:576/1530 train_loss:3.7634 train_time:94595ms step_avg:167.13ms
step:577/1530 train_loss:3.6744 train_time:94768ms step_avg:167.14ms
step:578/1530 train_loss:3.6780 train_time:94940ms step_avg:167.15ms
step:579/1530 train_loss:3.6718 train_time:95111ms step_avg:167.15ms
step:580/1530 train_loss:3.5985 train_time:95281ms step_avg:167.16ms
step:581/1530 train_loss:3.6415 train_time:95453ms step_avg:167.17ms
step:582/1530 train_loss:3.8456 train_time:95624ms step_avg:167.17ms
step:583/1530 train_loss:3.6264 train_time:95794ms step_avg:167.18ms
step:584/1530 train_loss:3.5949 train_time:95966ms step_avg:167.19ms
step:585/1530 train_loss:3.7905 train_time:96136ms step_avg:167.19ms
step:586/1530 train_loss:3.5233 train_time:96309ms step_avg:167.20ms
step:587/1530 train_loss:3.6716 train_time:96480ms step_avg:167.21ms
step:588/1530 train_loss:3.6413 train_time:96651ms step_avg:167.22ms
step:589/1530 train_loss:4.0010 train_time:96823ms step_avg:167.22ms
step:590/1530 train_loss:3.7861 train_time:96995ms step_avg:167.23ms
step:591/1530 train_loss:3.5082 train_time:97166ms step_avg:167.24ms
step:592/1530 train_loss:3.5345 train_time:97339ms step_avg:167.25ms
step:593/1530 train_loss:3.5036 train_time:97513ms step_avg:167.26ms
step:594/1530 train_loss:3.5629 train_time:97686ms step_avg:167.27ms
step:595/1530 train_loss:3.9175 train_time:97858ms step_avg:167.28ms
step:596/1530 train_loss:3.6444 train_time:98032ms step_avg:167.29ms
step:597/1530 train_loss:3.5904 train_time:98202ms step_avg:167.30ms
step:598/1530 train_loss:3.6584 train_time:98374ms step_avg:167.30ms
step:599/1530 train_loss:3.4854 train_time:98545ms step_avg:167.31ms
step:600/1530 train_loss:3.5973 train_time:98715ms step_avg:167.31ms
step:601/1530 train_loss:3.6479 train_time:98890ms step_avg:167.33ms
step:602/1530 train_loss:3.6734 train_time:99062ms step_avg:167.33ms
step:603/1530 train_loss:3.7857 train_time:99234ms step_avg:167.34ms
step:604/1530 train_loss:3.6129 train_time:99406ms step_avg:167.35ms
step:605/1530 train_loss:3.6100 train_time:99577ms step_avg:167.36ms
step:606/1530 train_loss:3.5759 train_time:99752ms step_avg:167.37ms
step:607/1530 train_loss:3.8440 train_time:99922ms step_avg:167.37ms
step:608/1530 train_loss:3.6381 train_time:100093ms step_avg:167.38ms
step:609/1530 train_loss:3.6165 train_time:100263ms step_avg:167.38ms
step:610/1530 train_loss:3.7013 train_time:100433ms step_avg:167.39ms
step:611/1530 train_loss:3.6078 train_time:100605ms step_avg:167.40ms
step:612/1530 train_loss:3.5720 train_time:100776ms step_avg:167.40ms
step:613/1530 train_loss:3.7633 train_time:100947ms step_avg:167.41ms
step:614/1530 train_loss:3.7126 train_time:101118ms step_avg:167.41ms
step:615/1530 train_loss:3.7014 train_time:101289ms step_avg:167.42ms
step:616/1530 train_loss:3.6346 train_time:101460ms step_avg:167.43ms
step:617/1530 train_loss:3.5666 train_time:101634ms step_avg:167.44ms
step:618/1530 train_loss:3.6951 train_time:101804ms step_avg:167.44ms
step:619/1530 train_loss:3.5528 train_time:101975ms step_avg:167.45ms
step:620/1530 train_loss:3.5910 train_time:102147ms step_avg:167.45ms
step:621/1530 train_loss:3.9300 train_time:102319ms step_avg:167.46ms
step:622/1530 train_loss:3.5760 train_time:102492ms step_avg:167.47ms
step:623/1530 train_loss:3.6019 train_time:102664ms step_avg:167.48ms
step:624/1530 train_loss:3.6912 train_time:102836ms step_avg:167.48ms
step:625/1530 train_loss:3.7071 train_time:103006ms step_avg:167.49ms
step:625/1530 val_loss:3.6263 train_time:103054ms step_avg:167.57ms
step:626/1530 train_loss:3.7368 train_time:103176ms step_avg:167.49ms
step:627/1530 train_loss:3.7143 train_time:103351ms step_avg:167.51ms
step:628/1530 train_loss:3.7646 train_time:103521ms step_avg:167.51ms
step:629/1530 train_loss:3.5928 train_time:103691ms step_avg:167.51ms
step:630/1530 train_loss:3.7235 train_time:103861ms step_avg:167.52ms
step:631/1530 train_loss:3.7445 train_time:104031ms step_avg:167.52ms
step:632/1530 train_loss:3.6520 train_time:104203ms step_avg:167.53ms
step:633/1530 train_loss:3.6121 train_time:104373ms step_avg:167.53ms
step:634/1530 train_loss:3.7027 train_time:104546ms step_avg:167.54ms
step:635/1530 train_loss:3.9544 train_time:104716ms step_avg:167.55ms
step:636/1530 train_loss:3.5547 train_time:104888ms step_avg:167.55ms
step:637/1530 train_loss:3.3587 train_time:105059ms step_avg:167.56ms
step:638/1530 train_loss:3.5985 train_time:105230ms step_avg:167.56ms
step:639/1530 train_loss:3.6342 train_time:105400ms step_avg:167.57ms
step:640/1530 train_loss:3.5734 train_time:105570ms step_avg:167.57ms
step:641/1530 train_loss:3.5912 train_time:105742ms step_avg:167.58ms
step:642/1530 train_loss:3.6375 train_time:105911ms step_avg:167.58ms
step:643/1530 train_loss:3.5999 train_time:106084ms step_avg:167.59ms
step:644/1530 train_loss:3.5578 train_time:106254ms step_avg:167.59ms
step:645/1530 train_loss:3.7808 train_time:106425ms step_avg:167.60ms
step:646/1530 train_loss:3.6737 train_time:106595ms step_avg:167.60ms
step:647/1530 train_loss:3.6634 train_time:106766ms step_avg:167.61ms
step:648/1530 train_loss:3.7153 train_time:106939ms step_avg:167.62ms
step:649/1530 train_loss:3.7710 train_time:107109ms step_avg:167.62ms
step:650/1530 train_loss:3.6209 train_time:107281ms step_avg:167.63ms
step:651/1530 train_loss:3.7709 train_time:107452ms step_avg:167.63ms
step:652/1530 train_loss:3.5907 train_time:107623ms step_avg:167.64ms
step:653/1530 train_loss:3.6631 train_time:107793ms step_avg:167.64ms
step:654/1530 train_loss:3.4338 train_time:107965ms step_avg:167.65ms
step:655/1530 train_loss:3.5860 train_time:108135ms step_avg:167.65ms
step:656/1530 train_loss:3.5754 train_time:108306ms step_avg:167.66ms
step:657/1530 train_loss:3.5013 train_time:108476ms step_avg:167.66ms
step:658/1530 train_loss:3.6898 train_time:108648ms step_avg:167.67ms
step:659/1530 train_loss:3.5883 train_time:108817ms step_avg:167.67ms
step:660/1530 train_loss:3.6958 train_time:108988ms step_avg:167.67ms
step:661/1530 train_loss:3.7516 train_time:109158ms step_avg:167.68ms
step:662/1530 train_loss:3.6790 train_time:109329ms step_avg:167.68ms
step:663/1530 train_loss:3.5558 train_time:109499ms step_avg:167.69ms
step:664/1530 train_loss:3.6143 train_time:109671ms step_avg:167.69ms
step:665/1530 train_loss:3.4947 train_time:109843ms step_avg:167.70ms
step:666/1530 train_loss:3.7818 train_time:110012ms step_avg:167.70ms
step:667/1530 train_loss:3.6088 train_time:110183ms step_avg:167.71ms
step:668/1530 train_loss:3.6524 train_time:110353ms step_avg:167.71ms
step:669/1530 train_loss:3.4952 train_time:110525ms step_avg:167.72ms
step:670/1530 train_loss:3.6073 train_time:110694ms step_avg:167.72ms
step:671/1530 train_loss:3.5640 train_time:110865ms step_avg:167.72ms
step:672/1530 train_loss:3.5701 train_time:111036ms step_avg:167.73ms
step:673/1530 train_loss:3.8588 train_time:111208ms step_avg:167.73ms
step:674/1530 train_loss:3.6321 train_time:111378ms step_avg:167.74ms
step:675/1530 train_loss:3.7170 train_time:111550ms step_avg:167.74ms
step:676/1530 train_loss:3.4918 train_time:111721ms step_avg:167.75ms
step:677/1530 train_loss:3.6033 train_time:111892ms step_avg:167.75ms
step:678/1530 train_loss:3.5617 train_time:112066ms step_avg:167.76ms
step:679/1530 train_loss:3.6810 train_time:112235ms step_avg:167.77ms
step:680/1530 train_loss:3.5846 train_time:112406ms step_avg:167.77ms
step:681/1530 train_loss:3.6168 train_time:112576ms step_avg:167.77ms
step:682/1530 train_loss:3.6696 train_time:112753ms step_avg:167.79ms
step:683/1530 train_loss:3.7411 train_time:112926ms step_avg:167.80ms
step:684/1530 train_loss:3.6510 train_time:113098ms step_avg:167.80ms
step:685/1530 train_loss:3.6868 train_time:113273ms step_avg:167.81ms
step:686/1530 train_loss:3.6412 train_time:113447ms step_avg:167.82ms
step:687/1530 train_loss:3.6710 train_time:113619ms step_avg:167.83ms
step:688/1530 train_loss:3.2105 train_time:113794ms step_avg:167.84ms
step:689/1530 train_loss:3.4057 train_time:113968ms step_avg:167.85ms
step:690/1530 train_loss:3.5428 train_time:114142ms step_avg:167.86ms
step:691/1530 train_loss:3.4136 train_time:114314ms step_avg:167.86ms
step:692/1530 train_loss:3.6320 train_time:114486ms step_avg:167.87ms
step:693/1530 train_loss:3.6506 train_time:114658ms step_avg:167.87ms
step:694/1530 train_loss:3.5555 train_time:114832ms step_avg:167.88ms
step:695/1530 train_loss:3.5369 train_time:115003ms step_avg:167.89ms
step:696/1530 train_loss:3.8546 train_time:115175ms step_avg:167.89ms
step:697/1530 train_loss:3.5933 train_time:115349ms step_avg:167.90ms
step:698/1530 train_loss:3.6519 train_time:115520ms step_avg:167.91ms
step:699/1530 train_loss:3.7719 train_time:115694ms step_avg:167.92ms
step:700/1530 train_loss:3.5740 train_time:115866ms step_avg:167.92ms
step:701/1530 train_loss:3.5509 train_time:116038ms step_avg:167.93ms
step:702/1530 train_loss:3.5128 train_time:116212ms step_avg:167.94ms
step:703/1530 train_loss:3.5021 train_time:116384ms step_avg:167.94ms
step:704/1530 train_loss:3.5779 train_time:116557ms step_avg:167.95ms
step:705/1530 train_loss:3.5644 train_time:116734ms step_avg:167.96ms
step:706/1530 train_loss:3.5814 train_time:116910ms step_avg:167.97ms
step:707/1530 train_loss:3.6516 train_time:117084ms step_avg:167.98ms
step:708/1530 train_loss:3.6067 train_time:117254ms step_avg:167.99ms
step:709/1530 train_loss:3.5855 train_time:117429ms step_avg:168.00ms
step:710/1530 train_loss:3.5446 train_time:117599ms step_avg:168.00ms
step:711/1530 train_loss:3.6015 train_time:117773ms step_avg:168.01ms
step:712/1530 train_loss:3.6500 train_time:117949ms step_avg:168.02ms
step:713/1530 train_loss:3.6564 train_time:118126ms step_avg:168.03ms
step:714/1530 train_loss:3.5638 train_time:118298ms step_avg:168.04ms
step:715/1530 train_loss:3.5723 train_time:118472ms step_avg:168.05ms
step:716/1530 train_loss:3.5931 train_time:118644ms step_avg:168.05ms
step:717/1530 train_loss:3.7125 train_time:118816ms step_avg:168.06ms
step:718/1530 train_loss:3.5981 train_time:118988ms step_avg:168.06ms
step:719/1530 train_loss:3.6813 train_time:119160ms step_avg:168.07ms
step:720/1530 train_loss:3.8514 train_time:119335ms step_avg:168.08ms
step:721/1530 train_loss:3.4703 train_time:119509ms step_avg:168.09ms
step:722/1530 train_loss:3.7451 train_time:119681ms step_avg:168.09ms
step:723/1530 train_loss:3.7775 train_time:119853ms step_avg:168.10ms
step:724/1530 train_loss:3.5710 train_time:120028ms step_avg:168.11ms
step:725/1530 train_loss:3.6575 train_time:120200ms step_avg:168.11ms
step:726/1530 train_loss:3.5360 train_time:120374ms step_avg:168.12ms
step:727/1530 train_loss:3.5821 train_time:120550ms step_avg:168.13ms
step:728/1530 train_loss:3.7331 train_time:120721ms step_avg:168.14ms
step:729/1530 train_loss:3.6763 train_time:120893ms step_avg:168.14ms
step:730/1530 train_loss:3.6583 train_time:121067ms step_avg:168.15ms
step:731/1530 train_loss:3.5580 train_time:121239ms step_avg:168.15ms
step:732/1530 train_loss:3.5986 train_time:121411ms step_avg:168.16ms
step:733/1530 train_loss:3.8364 train_time:121585ms step_avg:168.17ms
step:734/1530 train_loss:3.5590 train_time:121758ms step_avg:168.17ms
step:735/1530 train_loss:3.6231 train_time:121931ms step_avg:168.18ms
step:736/1530 train_loss:3.7380 train_time:122104ms step_avg:168.19ms
step:737/1530 train_loss:3.6778 train_time:122275ms step_avg:168.19ms
step:738/1530 train_loss:3.6104 train_time:122449ms step_avg:168.20ms
step:739/1530 train_loss:3.5115 train_time:122619ms step_avg:168.20ms
step:740/1530 train_loss:4.1216 train_time:122795ms step_avg:168.21ms
step:741/1530 train_loss:3.4863 train_time:122966ms step_avg:168.22ms
step:742/1530 train_loss:3.5542 train_time:123139ms step_avg:168.22ms
step:743/1530 train_loss:3.5831 train_time:123311ms step_avg:168.23ms
step:744/1530 train_loss:3.6546 train_time:123482ms step_avg:168.23ms
step:745/1530 train_loss:3.5942 train_time:123656ms step_avg:168.24ms
step:746/1530 train_loss:3.6001 train_time:123829ms step_avg:168.25ms
step:747/1530 train_loss:3.6544 train_time:124002ms step_avg:168.25ms
step:748/1530 train_loss:3.5666 train_time:124178ms step_avg:168.26ms
step:749/1530 train_loss:3.5623 train_time:124353ms step_avg:168.27ms
step:750/1530 train_loss:3.6016 train_time:124523ms step_avg:168.27ms
step:750/1530 val_loss:3.5694 train_time:124572ms step_avg:168.34ms
step:751/1530 train_loss:3.5762 train_time:124697ms step_avg:168.28ms
step:752/1530 train_loss:3.6195 train_time:124869ms step_avg:168.29ms
step:753/1530 train_loss:3.6194 train_time:125042ms step_avg:168.29ms
step:754/1530 train_loss:3.5995 train_time:125214ms step_avg:168.30ms
step:755/1530 train_loss:3.6865 train_time:125518ms step_avg:168.48ms
step:756/1530 train_loss:3.4642 train_time:125701ms step_avg:168.50ms
step:757/1530 train_loss:3.7302 train_time:125875ms step_avg:168.51ms
step:758/1530 train_loss:3.6516 train_time:126046ms step_avg:168.51ms
step:759/1530 train_loss:3.5949 train_time:126365ms step_avg:168.71ms
step:760/1530 train_loss:3.7126 train_time:126536ms step_avg:168.71ms
step:761/1530 train_loss:3.4048 train_time:126708ms step_avg:168.72ms
step:762/1530 train_loss:3.5511 train_time:126880ms step_avg:168.72ms
step:763/1530 train_loss:3.6663 train_time:127054ms step_avg:168.73ms
step:764/1530 train_loss:3.3189 train_time:127226ms step_avg:168.73ms
step:765/1530 train_loss:3.7325 train_time:127397ms step_avg:168.74ms
step:766/1530 train_loss:3.5726 train_time:127572ms step_avg:168.75ms
step:767/1530 train_loss:3.5726 train_time:127744ms step_avg:168.75ms
step:768/1530 train_loss:3.5718 train_time:127917ms step_avg:168.76ms
step:769/1530 train_loss:3.5904 train_time:128091ms step_avg:168.76ms
step:770/1530 train_loss:3.6422 train_time:128263ms step_avg:168.77ms
step:771/1530 train_loss:3.8873 train_time:128436ms step_avg:168.77ms
step:772/1530 train_loss:3.4584 train_time:128609ms step_avg:168.78ms
step:773/1530 train_loss:3.6367 train_time:128780ms step_avg:168.78ms
step:774/1530 train_loss:3.6435 train_time:128953ms step_avg:168.79ms
step:775/1530 train_loss:3.6124 train_time:129126ms step_avg:168.79ms
step:776/1530 train_loss:3.4121 train_time:129299ms step_avg:168.80ms
step:777/1530 train_loss:3.3982 train_time:129474ms step_avg:168.81ms
step:778/1530 train_loss:3.4931 train_time:129646ms step_avg:168.81ms
step:779/1530 train_loss:3.5839 train_time:129817ms step_avg:168.81ms
step:780/1530 train_loss:3.5899 train_time:129989ms step_avg:168.82ms
step:781/1530 train_loss:3.6797 train_time:130161ms step_avg:168.82ms
step:782/1530 train_loss:3.5959 train_time:130333ms step_avg:168.83ms
step:783/1530 train_loss:3.5675 train_time:130504ms step_avg:168.83ms
step:784/1530 train_loss:3.6069 train_time:130677ms step_avg:168.83ms
step:785/1530 train_loss:3.5632 train_time:130849ms step_avg:168.84ms
step:786/1530 train_loss:3.4460 train_time:131020ms step_avg:168.84ms
step:787/1530 train_loss:3.7742 train_time:131195ms step_avg:168.85ms
step:788/1530 train_loss:3.5055 train_time:131368ms step_avg:168.85ms
step:789/1530 train_loss:3.5514 train_time:131540ms step_avg:168.86ms
step:790/1530 train_loss:3.6327 train_time:131713ms step_avg:168.86ms
step:791/1530 train_loss:3.7778 train_time:131887ms step_avg:168.87ms
step:792/1530 train_loss:3.7653 train_time:132060ms step_avg:168.87ms
step:793/1530 train_loss:3.4489 train_time:132231ms step_avg:168.88ms
step:794/1530 train_loss:3.5969 train_time:132403ms step_avg:168.88ms
step:795/1530 train_loss:3.6771 train_time:132577ms step_avg:168.89ms
step:796/1530 train_loss:3.7526 train_time:132755ms step_avg:168.90ms
step:797/1530 train_loss:3.5286 train_time:132929ms step_avg:168.91ms
step:798/1530 train_loss:3.6536 train_time:133104ms step_avg:168.91ms
step:799/1530 train_loss:3.5358 train_time:133281ms step_avg:168.92ms
step:800/1530 train_loss:3.5392 train_time:133455ms step_avg:168.93ms
step:801/1530 train_loss:3.6310 train_time:133628ms step_avg:168.94ms
step:802/1530 train_loss:3.5016 train_time:133805ms step_avg:168.95ms
step:803/1530 train_loss:3.4965 train_time:133978ms step_avg:168.95ms
step:804/1530 train_loss:3.6277 train_time:134152ms step_avg:168.96ms
step:805/1530 train_loss:3.5215 train_time:134327ms step_avg:168.97ms
step:806/1530 train_loss:3.5646 train_time:134499ms step_avg:168.97ms
step:807/1530 train_loss:3.6468 train_time:134674ms step_avg:168.98ms
step:808/1530 train_loss:3.5430 train_time:134851ms step_avg:168.99ms
step:809/1530 train_loss:3.5022 train_time:135023ms step_avg:168.99ms
step:810/1530 train_loss:3.5652 train_time:135197ms step_avg:169.00ms
step:811/1530 train_loss:3.5889 train_time:135371ms step_avg:169.00ms
step:812/1530 train_loss:3.6002 train_time:135543ms step_avg:169.01ms
step:813/1530 train_loss:3.6345 train_time:135716ms step_avg:169.01ms
step:814/1530 train_loss:3.5698 train_time:135891ms step_avg:169.02ms
step:815/1530 train_loss:3.5646 train_time:136064ms step_avg:169.02ms
step:816/1530 train_loss:3.6931 train_time:136239ms step_avg:169.03ms
step:817/1530 train_loss:3.7738 train_time:136413ms step_avg:169.04ms
step:818/1530 train_loss:3.5323 train_time:136585ms step_avg:169.04ms
step:819/1530 train_loss:3.7202 train_time:136759ms step_avg:169.05ms
step:820/1530 train_loss:3.4970 train_time:136935ms step_avg:169.06ms
step:821/1530 train_loss:3.5691 train_time:137109ms step_avg:169.06ms
step:822/1530 train_loss:3.7023 train_time:137283ms step_avg:169.07ms
step:823/1530 train_loss:3.5762 train_time:137457ms step_avg:169.07ms
step:824/1530 train_loss:3.5157 train_time:137630ms step_avg:169.08ms
step:825/1530 train_loss:3.6203 train_time:137804ms step_avg:169.08ms
step:826/1530 train_loss:3.4831 train_time:137978ms step_avg:169.09ms
step:827/1530 train_loss:3.7389 train_time:138153ms step_avg:169.10ms
step:828/1530 train_loss:3.6265 train_time:138325ms step_avg:169.10ms
step:829/1530 train_loss:3.6342 train_time:138500ms step_avg:169.11ms
step:830/1530 train_loss:3.5467 train_time:138675ms step_avg:169.12ms
step:831/1530 train_loss:3.6041 train_time:138848ms step_avg:169.12ms
step:832/1530 train_loss:3.5182 train_time:139022ms step_avg:169.13ms
step:833/1530 train_loss:3.6620 train_time:139198ms step_avg:169.14ms
step:834/1530 train_loss:3.4810 train_time:139373ms step_avg:169.14ms
step:835/1530 train_loss:3.4616 train_time:139546ms step_avg:169.15ms
step:836/1530 train_loss:3.7216 train_time:139720ms step_avg:169.15ms
step:837/1530 train_loss:3.4061 train_time:139895ms step_avg:169.16ms
step:838/1530 train_loss:3.5992 train_time:140070ms step_avg:169.17ms
step:839/1530 train_loss:3.4278 train_time:140243ms step_avg:169.17ms
step:840/1530 train_loss:3.4747 train_time:140415ms step_avg:169.17ms
step:841/1530 train_loss:3.5780 train_time:140587ms step_avg:169.18ms
step:842/1530 train_loss:3.5890 train_time:140763ms step_avg:169.19ms
step:843/1530 train_loss:3.5735 train_time:140936ms step_avg:169.19ms
step:844/1530 train_loss:3.4370 train_time:141109ms step_avg:169.20ms
step:845/1530 train_loss:3.6649 train_time:141282ms step_avg:169.20ms
step:846/1530 train_loss:3.5198 train_time:141458ms step_avg:169.21ms
step:847/1530 train_loss:3.4971 train_time:141633ms step_avg:169.22ms
step:848/1530 train_loss:3.6435 train_time:141805ms step_avg:169.22ms
step:849/1530 train_loss:3.4871 train_time:141981ms step_avg:169.23ms
step:850/1530 train_loss:3.4460 train_time:142156ms step_avg:169.23ms
step:851/1530 train_loss:3.7387 train_time:142332ms step_avg:169.24ms
step:852/1530 train_loss:3.4425 train_time:142504ms step_avg:169.24ms
step:853/1530 train_loss:3.5712 train_time:142678ms step_avg:169.25ms
step:854/1530 train_loss:3.6553 train_time:142854ms step_avg:169.26ms
step:855/1530 train_loss:3.5184 train_time:143028ms step_avg:169.26ms
step:856/1530 train_loss:3.5474 train_time:143200ms step_avg:169.27ms
step:857/1530 train_loss:3.6107 train_time:143376ms step_avg:169.28ms
step:858/1530 train_loss:3.4712 train_time:143553ms step_avg:169.28ms
step:859/1530 train_loss:3.5641 train_time:143727ms step_avg:169.29ms
step:860/1530 train_loss:3.5916 train_time:143898ms step_avg:169.29ms
step:861/1530 train_loss:3.6283 train_time:144077ms step_avg:169.30ms
step:862/1530 train_loss:3.6058 train_time:144254ms step_avg:169.31ms
step:863/1530 train_loss:3.5723 train_time:144430ms step_avg:169.32ms
step:864/1530 train_loss:3.3851 train_time:144603ms step_avg:169.32ms
step:865/1530 train_loss:3.6002 train_time:144776ms step_avg:169.33ms
step:866/1530 train_loss:3.8874 train_time:144954ms step_avg:169.34ms
step:867/1530 train_loss:3.4630 train_time:145128ms step_avg:169.34ms
step:868/1530 train_loss:3.6460 train_time:145299ms step_avg:169.35ms
step:869/1530 train_loss:3.6189 train_time:145475ms step_avg:169.35ms
step:870/1530 train_loss:3.4529 train_time:145649ms step_avg:169.36ms
step:871/1530 train_loss:3.3985 train_time:145823ms step_avg:169.36ms
step:872/1530 train_loss:3.6527 train_time:145998ms step_avg:169.37ms
step:873/1530 train_loss:3.4680 train_time:146172ms step_avg:169.38ms
step:874/1530 train_loss:3.2259 train_time:146350ms step_avg:169.39ms
step:875/1530 train_loss:3.6316 train_time:146524ms step_avg:169.39ms
step:875/1530 val_loss:3.5239 train_time:146573ms step_avg:169.45ms
step:876/1530 train_loss:3.4452 train_time:146698ms step_avg:169.40ms
step:877/1530 train_loss:3.6263 train_time:146875ms step_avg:169.41ms
step:878/1530 train_loss:3.4755 train_time:147051ms step_avg:169.41ms
step:879/1530 train_loss:3.6592 train_time:147224ms step_avg:169.42ms
step:880/1530 train_loss:3.3080 train_time:147396ms step_avg:169.42ms
step:881/1530 train_loss:3.4825 train_time:147570ms step_avg:169.43ms
step:882/1530 train_loss:3.6985 train_time:147742ms step_avg:169.43ms
step:883/1530 train_loss:3.8410 train_time:147916ms step_avg:169.43ms
step:884/1530 train_loss:3.5700 train_time:148093ms step_avg:169.44ms
step:885/1530 train_loss:3.4991 train_time:148266ms step_avg:169.45ms
step:886/1530 train_loss:3.5781 train_time:148439ms step_avg:169.45ms
step:887/1530 train_loss:4.0910 train_time:148615ms step_avg:169.46ms
step:888/1530 train_loss:3.8397 train_time:148795ms step_avg:169.47ms
step:889/1530 train_loss:3.5273 train_time:148969ms step_avg:169.48ms
step:890/1530 train_loss:3.5317 train_time:149139ms step_avg:169.48ms
step:891/1530 train_loss:3.3636 train_time:149315ms step_avg:169.48ms
step:892/1530 train_loss:3.7203 train_time:149488ms step_avg:169.49ms
step:893/1530 train_loss:3.4301 train_time:149658ms step_avg:169.49ms
step:894/1530 train_loss:3.6491 train_time:149833ms step_avg:169.49ms
step:895/1530 train_loss:3.6878 train_time:150008ms step_avg:169.50ms
step:896/1530 train_loss:3.5059 train_time:150180ms step_avg:169.50ms
step:897/1530 train_loss:3.5513 train_time:150356ms step_avg:169.51ms
step:898/1530 train_loss:3.5909 train_time:150532ms step_avg:169.52ms
step:899/1530 train_loss:3.4848 train_time:150705ms step_avg:169.52ms
step:900/1530 train_loss:3.4283 train_time:150877ms step_avg:169.52ms
step:901/1530 train_loss:3.6242 train_time:151051ms step_avg:169.53ms
step:902/1530 train_loss:3.6420 train_time:151226ms step_avg:169.54ms
step:903/1530 train_loss:3.5475 train_time:151401ms step_avg:169.54ms
step:904/1530 train_loss:3.4967 train_time:151575ms step_avg:169.55ms
step:905/1530 train_loss:3.5017 train_time:151746ms step_avg:169.55ms
step:906/1530 train_loss:3.7113 train_time:151921ms step_avg:169.55ms
step:907/1530 train_loss:3.5223 train_time:152095ms step_avg:169.56ms
step:908/1530 train_loss:3.5722 train_time:152269ms step_avg:169.56ms
step:909/1530 train_loss:3.4601 train_time:152444ms step_avg:169.57ms
step:910/1530 train_loss:3.5316 train_time:152624ms step_avg:169.58ms
step:911/1530 train_loss:3.6470 train_time:152800ms step_avg:169.59ms
step:912/1530 train_loss:3.6078 train_time:152979ms step_avg:169.60ms
step:913/1530 train_loss:3.4672 train_time:153158ms step_avg:169.61ms
step:914/1530 train_loss:3.7503 train_time:153336ms step_avg:169.62ms
step:915/1530 train_loss:3.5339 train_time:153517ms step_avg:169.63ms
step:916/1530 train_loss:3.6223 train_time:153694ms step_avg:169.64ms
step:917/1530 train_loss:3.6076 train_time:153869ms step_avg:169.65ms
step:918/1530 train_loss:4.8461 train_time:154049ms step_avg:169.66ms
step:919/1530 train_loss:3.5046 train_time:154229ms step_avg:169.67ms
step:920/1530 train_loss:3.5945 train_time:154403ms step_avg:169.67ms
step:921/1530 train_loss:3.5588 train_time:154579ms step_avg:169.68ms
step:922/1530 train_loss:3.5849 train_time:154758ms step_avg:169.69ms
step:923/1530 train_loss:3.6186 train_time:154934ms step_avg:169.70ms
step:924/1530 train_loss:3.6849 train_time:155111ms step_avg:169.71ms
step:925/1530 train_loss:3.6478 train_time:155287ms step_avg:169.71ms
step:926/1530 train_loss:3.5597 train_time:155460ms step_avg:169.72ms
step:927/1530 train_loss:3.5592 train_time:155636ms step_avg:169.72ms
step:928/1530 train_loss:3.7852 train_time:155814ms step_avg:169.73ms
step:929/1530 train_loss:3.6163 train_time:155989ms step_avg:169.74ms
step:930/1530 train_loss:3.4103 train_time:156166ms step_avg:169.75ms
step:931/1530 train_loss:3.5021 train_time:156339ms step_avg:169.75ms
step:932/1530 train_loss:3.6503 train_time:156518ms step_avg:169.76ms
step:933/1530 train_loss:3.3688 train_time:156694ms step_avg:169.77ms
step:934/1530 train_loss:3.5888 train_time:156872ms step_avg:169.77ms
step:935/1530 train_loss:3.4425 train_time:157049ms step_avg:169.78ms
step:936/1530 train_loss:3.5279 train_time:157226ms step_avg:169.79ms
step:937/1530 train_loss:3.6355 train_time:157403ms step_avg:169.80ms
step:938/1530 train_loss:3.5453 train_time:157577ms step_avg:169.80ms
step:939/1530 train_loss:3.6856 train_time:157758ms step_avg:169.81ms
step:940/1530 train_loss:3.4810 train_time:157933ms step_avg:169.82ms
step:941/1530 train_loss:3.5533 train_time:158108ms step_avg:169.83ms
step:942/1530 train_loss:3.3636 train_time:158285ms step_avg:169.83ms
step:943/1530 train_loss:3.7180 train_time:158465ms step_avg:169.84ms
step:944/1530 train_loss:3.4095 train_time:158784ms step_avg:170.00ms
step:945/1530 train_loss:3.4265 train_time:158969ms step_avg:170.02ms
step:946/1530 train_loss:5.0918 train_time:159148ms step_avg:170.03ms
step:947/1530 train_loss:3.6056 train_time:159324ms step_avg:170.04ms
step:948/1530 train_loss:3.4905 train_time:159498ms step_avg:170.04ms
step:949/1530 train_loss:3.3802 train_time:159821ms step_avg:170.20ms
step:950/1530 train_loss:3.4479 train_time:159997ms step_avg:170.21ms
step:951/1530 train_loss:3.4143 train_time:160176ms step_avg:170.22ms
step:952/1530 train_loss:3.4824 train_time:160352ms step_avg:170.22ms
step:953/1530 train_loss:3.5744 train_time:160528ms step_avg:170.23ms
step:954/1530 train_loss:3.4530 train_time:160704ms step_avg:170.24ms
step:955/1530 train_loss:3.4837 train_time:160879ms step_avg:170.24ms
step:956/1530 train_loss:3.4518 train_time:161054ms step_avg:170.25ms
step:957/1530 train_loss:3.4971 train_time:161233ms step_avg:170.26ms
step:958/1530 train_loss:3.5099 train_time:161412ms step_avg:170.27ms
step:959/1530 train_loss:3.5186 train_time:161589ms step_avg:170.27ms
step:960/1530 train_loss:3.4106 train_time:161766ms step_avg:170.28ms
step:961/1530 train_loss:3.6506 train_time:161941ms step_avg:170.29ms
step:962/1530 train_loss:3.5978 train_time:162116ms step_avg:170.29ms
step:963/1530 train_loss:3.6087 train_time:162294ms step_avg:170.30ms
step:964/1530 train_loss:3.4250 train_time:162473ms step_avg:170.31ms
step:965/1530 train_loss:3.4851 train_time:162645ms step_avg:170.31ms
step:966/1530 train_loss:3.7132 train_time:162820ms step_avg:170.31ms
step:967/1530 train_loss:3.5263 train_time:162996ms step_avg:170.32ms
step:968/1530 train_loss:3.5176 train_time:163174ms step_avg:170.33ms
step:969/1530 train_loss:3.5939 train_time:163348ms step_avg:170.33ms
step:970/1530 train_loss:3.3836 train_time:163520ms step_avg:170.33ms
step:971/1530 train_loss:3.5385 train_time:163694ms step_avg:170.34ms
step:972/1530 train_loss:3.4768 train_time:163869ms step_avg:170.34ms
step:973/1530 train_loss:3.5463 train_time:164042ms step_avg:170.34ms
step:974/1530 train_loss:3.5946 train_time:164219ms step_avg:170.35ms
step:975/1530 train_loss:3.4687 train_time:164395ms step_avg:170.36ms
step:976/1530 train_loss:3.6772 train_time:164571ms step_avg:170.36ms
step:977/1530 train_loss:3.5719 train_time:164745ms step_avg:170.37ms
step:978/1530 train_loss:3.3620 train_time:164920ms step_avg:170.37ms
step:979/1530 train_loss:3.6367 train_time:165096ms step_avg:170.38ms
step:980/1530 train_loss:3.4187 train_time:165274ms step_avg:170.39ms
step:981/1530 train_loss:3.5763 train_time:165450ms step_avg:170.39ms
step:982/1530 train_loss:3.5490 train_time:165623ms step_avg:170.39ms
step:983/1530 train_loss:3.5275 train_time:165798ms step_avg:170.40ms
step:984/1530 train_loss:3.5002 train_time:165975ms step_avg:170.41ms
step:985/1530 train_loss:3.5800 train_time:166153ms step_avg:170.41ms
step:986/1530 train_loss:3.4211 train_time:166329ms step_avg:170.42ms
step:987/1530 train_loss:3.4971 train_time:166502ms step_avg:170.42ms
step:988/1530 train_loss:3.4933 train_time:166678ms step_avg:170.43ms
step:989/1530 train_loss:3.4236 train_time:166852ms step_avg:170.43ms
step:990/1530 train_loss:3.6641 train_time:167029ms step_avg:170.44ms
step:991/1530 train_loss:3.4748 train_time:167203ms step_avg:170.44ms
step:992/1530 train_loss:3.4480 train_time:167384ms step_avg:170.45ms
step:993/1530 train_loss:3.5080 train_time:167564ms step_avg:170.46ms
step:994/1530 train_loss:3.5985 train_time:167738ms step_avg:170.47ms
step:995/1530 train_loss:3.5398 train_time:167911ms step_avg:170.47ms
step:996/1530 train_loss:3.4614 train_time:168083ms step_avg:170.47ms
step:997/1530 train_loss:3.7659 train_time:168256ms step_avg:170.47ms
step:998/1530 train_loss:3.4411 train_time:168429ms step_avg:170.47ms
step:999/1530 train_loss:3.5928 train_time:168603ms step_avg:170.48ms
step:1000/1530 train_loss:3.4395 train_time:168781ms step_avg:170.49ms
step:1000/1530 val_loss:3.4708 train_time:168832ms step_avg:170.54ms
step:1001/1530 train_loss:3.5032 train_time:168957ms step_avg:170.49ms
step:1002/1530 train_loss:3.3788 train_time:169130ms step_avg:170.49ms
step:1003/1530 train_loss:3.5634 train_time:169306ms step_avg:170.50ms
step:1004/1530 train_loss:3.6089 train_time:169484ms step_avg:170.51ms
step:1005/1530 train_loss:3.3904 train_time:169658ms step_avg:170.51ms
step:1006/1530 train_loss:3.4702 train_time:169835ms step_avg:170.52ms
step:1007/1530 train_loss:3.4394 train_time:170009ms step_avg:170.52ms
step:1008/1530 train_loss:3.5686 train_time:170185ms step_avg:170.53ms
step:1009/1530 train_loss:3.6698 train_time:170363ms step_avg:170.53ms
step:1010/1530 train_loss:3.5666 train_time:170538ms step_avg:170.54ms
step:1011/1530 train_loss:3.5416 train_time:170710ms step_avg:170.54ms
step:1012/1530 train_loss:3.3936 train_time:170884ms step_avg:170.54ms
step:1013/1530 train_loss:3.5366 train_time:171059ms step_avg:170.55ms
step:1014/1530 train_loss:3.6261 train_time:171235ms step_avg:170.55ms
step:1015/1530 train_loss:3.3375 train_time:171411ms step_avg:170.56ms
step:1016/1530 train_loss:3.4097 train_time:171587ms step_avg:170.56ms
step:1017/1530 train_loss:3.4004 train_time:171764ms step_avg:170.57ms
step:1018/1530 train_loss:3.3975 train_time:171940ms step_avg:170.58ms
step:1019/1530 train_loss:3.5258 train_time:172115ms step_avg:170.58ms
step:1020/1530 train_loss:3.3866 train_time:172292ms step_avg:170.59ms
step:1021/1530 train_loss:3.3599 train_time:172466ms step_avg:170.59ms
step:1022/1530 train_loss:3.4795 train_time:172642ms step_avg:170.60ms
step:1023/1530 train_loss:3.5124 train_time:172820ms step_avg:170.60ms
step:1024/1530 train_loss:3.4822 train_time:172998ms step_avg:170.61ms
step:1025/1530 train_loss:3.4796 train_time:173174ms step_avg:170.61ms
step:1026/1530 train_loss:3.6193 train_time:173350ms step_avg:170.62ms
step:1027/1530 train_loss:3.3339 train_time:173527ms step_avg:170.63ms
step:1028/1530 train_loss:3.4043 train_time:173708ms step_avg:170.64ms
step:1029/1530 train_loss:3.3201 train_time:173889ms step_avg:170.65ms
step:1030/1530 train_loss:3.5467 train_time:174066ms step_avg:170.65ms
step:1031/1530 train_loss:3.5143 train_time:174242ms step_avg:170.66ms
step:1032/1530 train_loss:3.7003 train_time:174425ms step_avg:170.67ms
step:1033/1530 train_loss:3.4909 train_time:174600ms step_avg:170.67ms
step:1034/1530 train_loss:3.3969 train_time:174776ms step_avg:170.68ms
step:1035/1530 train_loss:3.4478 train_time:174954ms step_avg:170.69ms
step:1036/1530 train_loss:3.4863 train_time:175130ms step_avg:170.69ms
step:1037/1530 train_loss:3.7978 train_time:175309ms step_avg:170.70ms
step:1038/1530 train_loss:3.6193 train_time:175488ms step_avg:170.71ms
step:1039/1530 train_loss:3.5214 train_time:175668ms step_avg:170.72ms
step:1040/1530 train_loss:3.4206 train_time:175844ms step_avg:170.72ms
step:1041/1530 train_loss:3.4894 train_time:176022ms step_avg:170.73ms
step:1042/1530 train_loss:3.5259 train_time:176195ms step_avg:170.73ms
step:1043/1530 train_loss:3.4452 train_time:176370ms step_avg:170.74ms
step:1044/1530 train_loss:3.4641 train_time:176548ms step_avg:170.74ms
step:1045/1530 train_loss:3.5140 train_time:176728ms step_avg:170.75ms
step:1046/1530 train_loss:3.4284 train_time:176905ms step_avg:170.76ms
step:1047/1530 train_loss:3.6317 train_time:177083ms step_avg:170.76ms
step:1048/1530 train_loss:3.5013 train_time:177259ms step_avg:170.77ms
step:1049/1530 train_loss:3.4023 train_time:177435ms step_avg:170.77ms
step:1050/1530 train_loss:3.3978 train_time:177611ms step_avg:170.78ms
step:1051/1530 train_loss:3.4966 train_time:177789ms step_avg:170.79ms
step:1052/1530 train_loss:3.3647 train_time:177967ms step_avg:170.79ms
step:1053/1530 train_loss:3.6940 train_time:178144ms step_avg:170.80ms
step:1054/1530 train_loss:3.5489 train_time:178324ms step_avg:170.81ms
step:1055/1530 train_loss:3.3880 train_time:178501ms step_avg:170.81ms
step:1056/1530 train_loss:3.5068 train_time:178677ms step_avg:170.82ms
step:1057/1530 train_loss:3.5834 train_time:178852ms step_avg:170.82ms
step:1058/1530 train_loss:3.3095 train_time:179031ms step_avg:170.83ms
step:1059/1530 train_loss:3.3738 train_time:179211ms step_avg:170.84ms
step:1060/1530 train_loss:3.4422 train_time:179388ms step_avg:170.85ms
step:1061/1530 train_loss:3.4211 train_time:179563ms step_avg:170.85ms
step:1062/1530 train_loss:3.3849 train_time:179739ms step_avg:170.85ms
step:1063/1530 train_loss:3.4623 train_time:179913ms step_avg:170.86ms
step:1064/1530 train_loss:3.3868 train_time:180086ms step_avg:170.86ms
step:1065/1530 train_loss:3.3656 train_time:180264ms step_avg:170.87ms
step:1066/1530 train_loss:3.4195 train_time:180439ms step_avg:170.87ms
step:1067/1530 train_loss:3.2827 train_time:180617ms step_avg:170.88ms
step:1068/1530 train_loss:3.4426 train_time:180794ms step_avg:170.88ms
step:1069/1530 train_loss:3.3013 train_time:180975ms step_avg:170.89ms
step:1070/1530 train_loss:3.5702 train_time:181150ms step_avg:170.90ms
step:1071/1530 train_loss:3.5184 train_time:181329ms step_avg:170.90ms
step:1072/1530 train_loss:3.4464 train_time:181504ms step_avg:170.91ms
step:1073/1530 train_loss:3.5267 train_time:181678ms step_avg:170.91ms
step:1074/1530 train_loss:3.4339 train_time:181854ms step_avg:170.92ms
step:1075/1530 train_loss:3.4034 train_time:182033ms step_avg:170.92ms
step:1076/1530 train_loss:3.7985 train_time:182208ms step_avg:170.93ms
step:1077/1530 train_loss:3.4372 train_time:182384ms step_avg:170.93ms
step:1078/1530 train_loss:3.0940 train_time:182570ms step_avg:170.95ms
step:1079/1530 train_loss:3.5435 train_time:182746ms step_avg:170.95ms
step:1080/1530 train_loss:3.4326 train_time:182924ms step_avg:170.96ms
step:1081/1530 train_loss:3.5116 train_time:183099ms step_avg:170.96ms
step:1082/1530 train_loss:3.5937 train_time:183273ms step_avg:170.96ms
step:1083/1530 train_loss:3.5009 train_time:183448ms step_avg:170.97ms
step:1084/1530 train_loss:3.4671 train_time:183625ms step_avg:170.97ms
step:1085/1530 train_loss:3.4344 train_time:183801ms step_avg:170.98ms
step:1086/1530 train_loss:3.6361 train_time:183979ms step_avg:170.98ms
step:1087/1530 train_loss:3.5061 train_time:184154ms step_avg:170.99ms
step:1088/1530 train_loss:3.3740 train_time:184331ms step_avg:170.99ms
step:1089/1530 train_loss:3.3782 train_time:184511ms step_avg:171.00ms
step:1090/1530 train_loss:3.4830 train_time:184690ms step_avg:171.01ms
step:1091/1530 train_loss:3.2901 train_time:184866ms step_avg:171.01ms
step:1092/1530 train_loss:3.4873 train_time:185042ms step_avg:171.02ms
step:1093/1530 train_loss:3.6104 train_time:185221ms step_avg:171.03ms
step:1094/1530 train_loss:3.4528 train_time:185396ms step_avg:171.03ms
step:1095/1530 train_loss:3.4255 train_time:185569ms step_avg:171.03ms
step:1096/1530 train_loss:3.4315 train_time:185746ms step_avg:171.04ms
step:1097/1530 train_loss:3.4943 train_time:185925ms step_avg:171.04ms
step:1098/1530 train_loss:3.5665 train_time:186104ms step_avg:171.05ms
step:1099/1530 train_loss:3.5309 train_time:186282ms step_avg:171.06ms
step:1100/1530 train_loss:3.4361 train_time:186461ms step_avg:171.06ms
step:1101/1530 train_loss:3.2944 train_time:186639ms step_avg:171.07ms
step:1102/1530 train_loss:3.3169 train_time:186817ms step_avg:171.08ms
step:1103/1530 train_loss:3.4475 train_time:186999ms step_avg:171.09ms
step:1104/1530 train_loss:3.3234 train_time:187175ms step_avg:171.09ms
step:1105/1530 train_loss:4.0709 train_time:187353ms step_avg:171.10ms
step:1106/1530 train_loss:3.2236 train_time:187528ms step_avg:171.10ms
step:1107/1530 train_loss:3.5705 train_time:187705ms step_avg:171.11ms
step:1108/1530 train_loss:3.3536 train_time:187879ms step_avg:171.11ms
step:1109/1530 train_loss:3.5028 train_time:188053ms step_avg:171.11ms
step:1110/1530 train_loss:3.4296 train_time:188227ms step_avg:171.12ms
step:1111/1530 train_loss:3.4888 train_time:188404ms step_avg:171.12ms
step:1112/1530 train_loss:3.5643 train_time:188584ms step_avg:171.13ms
step:1113/1530 train_loss:3.4357 train_time:188766ms step_avg:171.14ms
step:1114/1530 train_loss:3.3778 train_time:188946ms step_avg:171.15ms
step:1115/1530 train_loss:3.2413 train_time:189126ms step_avg:171.15ms
step:1116/1530 train_loss:3.4330 train_time:189299ms step_avg:171.16ms
step:1117/1530 train_loss:3.5953 train_time:189478ms step_avg:171.16ms
step:1118/1530 train_loss:3.6347 train_time:189654ms step_avg:171.17ms
step:1119/1530 train_loss:3.4829 train_time:189827ms step_avg:171.17ms
step:1120/1530 train_loss:3.4933 train_time:190005ms step_avg:171.18ms
step:1121/1530 train_loss:3.3955 train_time:190183ms step_avg:171.18ms
step:1122/1530 train_loss:3.4677 train_time:190357ms step_avg:171.18ms
step:1123/1530 train_loss:3.5881 train_time:190532ms step_avg:171.19ms
step:1124/1530 train_loss:3.3406 train_time:190708ms step_avg:171.19ms
step:1125/1530 train_loss:3.2404 train_time:190885ms step_avg:171.20ms
step:1125/1530 val_loss:3.4133 train_time:190936ms step_avg:171.24ms
step:1126/1530 train_loss:3.4818 train_time:191062ms step_avg:171.20ms
step:1127/1530 train_loss:3.6756 train_time:191242ms step_avg:171.21ms
step:1128/1530 train_loss:3.2357 train_time:191422ms step_avg:171.22ms
step:1129/1530 train_loss:3.5623 train_time:191602ms step_avg:171.23ms
step:1130/1530 train_loss:3.3794 train_time:191780ms step_avg:171.23ms
step:1131/1530 train_loss:3.4073 train_time:191960ms step_avg:171.24ms
step:1132/1530 train_loss:3.3703 train_time:192133ms step_avg:171.24ms
step:1133/1530 train_loss:3.4955 train_time:192442ms step_avg:171.36ms
step:1134/1530 train_loss:3.4544 train_time:192626ms step_avg:171.38ms
step:1135/1530 train_loss:3.5249 train_time:192806ms step_avg:171.38ms
step:1136/1530 train_loss:3.5700 train_time:192984ms step_avg:171.39ms
step:1137/1530 train_loss:3.4638 train_time:193160ms step_avg:171.39ms
step:1138/1530 train_loss:3.3555 train_time:193340ms step_avg:171.40ms
step:1139/1530 train_loss:3.6596 train_time:193662ms step_avg:171.53ms
step:1140/1530 train_loss:3.4610 train_time:193841ms step_avg:171.54ms
step:1141/1530 train_loss:3.6022 train_time:194023ms step_avg:171.55ms
step:1142/1530 train_loss:3.4502 train_time:194201ms step_avg:171.56ms
step:1143/1530 train_loss:3.3697 train_time:194381ms step_avg:171.56ms
step:1144/1530 train_loss:3.4417 train_time:194558ms step_avg:171.57ms
step:1145/1530 train_loss:3.5919 train_time:194733ms step_avg:171.57ms
step:1146/1530 train_loss:3.5595 train_time:194913ms step_avg:171.58ms
step:1147/1530 train_loss:3.4873 train_time:195090ms step_avg:171.58ms
step:1148/1530 train_loss:3.5022 train_time:195268ms step_avg:171.59ms
step:1149/1530 train_loss:3.3338 train_time:195448ms step_avg:171.60ms
step:1150/1530 train_loss:3.3855 train_time:195624ms step_avg:171.60ms
step:1151/1530 train_loss:3.3235 train_time:195803ms step_avg:171.61ms
step:1152/1530 train_loss:3.3961 train_time:195985ms step_avg:171.62ms
step:1153/1530 train_loss:3.4307 train_time:196165ms step_avg:171.62ms
step:1154/1530 train_loss:3.5253 train_time:196341ms step_avg:171.63ms
step:1155/1530 train_loss:3.3227 train_time:196524ms step_avg:171.64ms
step:1156/1530 train_loss:3.5438 train_time:196706ms step_avg:171.65ms
step:1157/1530 train_loss:3.5020 train_time:196884ms step_avg:171.65ms
step:1158/1530 train_loss:3.2529 train_time:197060ms step_avg:171.65ms
step:1159/1530 train_loss:3.3564 train_time:197237ms step_avg:171.66ms
step:1160/1530 train_loss:3.3407 train_time:197413ms step_avg:171.66ms
step:1161/1530 train_loss:3.0840 train_time:197591ms step_avg:171.67ms
step:1162/1530 train_loss:3.4270 train_time:197768ms step_avg:171.67ms
step:1163/1530 train_loss:3.3900 train_time:197946ms step_avg:171.68ms
step:1164/1530 train_loss:3.2975 train_time:198124ms step_avg:171.68ms
step:1165/1530 train_loss:3.2540 train_time:198300ms step_avg:171.69ms
step:1166/1530 train_loss:3.3947 train_time:198480ms step_avg:171.70ms
step:1167/1530 train_loss:3.4191 train_time:198656ms step_avg:171.70ms
step:1168/1530 train_loss:3.7285 train_time:198831ms step_avg:171.70ms
step:1169/1530 train_loss:3.3820 train_time:199007ms step_avg:171.71ms
step:1170/1530 train_loss:3.3917 train_time:199184ms step_avg:171.71ms
step:1171/1530 train_loss:3.3078 train_time:199361ms step_avg:171.72ms
step:1172/1530 train_loss:3.4301 train_time:199536ms step_avg:171.72ms
step:1173/1530 train_loss:3.5429 train_time:199716ms step_avg:171.73ms
step:1174/1530 train_loss:3.3910 train_time:199901ms step_avg:171.74ms
step:1175/1530 train_loss:3.3631 train_time:200081ms step_avg:171.74ms
step:1176/1530 train_loss:3.4287 train_time:200263ms step_avg:171.75ms
step:1177/1530 train_loss:3.4510 train_time:200446ms step_avg:171.76ms
step:1178/1530 train_loss:3.4985 train_time:200623ms step_avg:171.77ms
step:1179/1530 train_loss:3.4078 train_time:200800ms step_avg:171.77ms
step:1180/1530 train_loss:3.3546 train_time:200987ms step_avg:171.78ms
step:1181/1530 train_loss:3.3415 train_time:201165ms step_avg:171.79ms
step:1182/1530 train_loss:3.3808 train_time:201344ms step_avg:171.79ms
step:1183/1530 train_loss:3.3393 train_time:201520ms step_avg:171.80ms
step:1184/1530 train_loss:3.5160 train_time:201698ms step_avg:171.80ms
step:1185/1530 train_loss:3.5498 train_time:201880ms step_avg:171.81ms
step:1186/1530 train_loss:3.3677 train_time:202060ms step_avg:171.82ms
step:1187/1530 train_loss:3.4174 train_time:202246ms step_avg:171.83ms
step:1188/1530 train_loss:3.4470 train_time:202422ms step_avg:171.84ms
step:1189/1530 train_loss:3.2795 train_time:202604ms step_avg:171.84ms
step:1190/1530 train_loss:3.4491 train_time:202783ms step_avg:171.85ms
step:1191/1530 train_loss:3.5842 train_time:202963ms step_avg:171.86ms
step:1192/1530 train_loss:3.3968 train_time:203138ms step_avg:171.86ms
step:1193/1530 train_loss:3.2764 train_time:203314ms step_avg:171.86ms
step:1194/1530 train_loss:3.5631 train_time:203491ms step_avg:171.87ms
step:1195/1530 train_loss:3.3765 train_time:203670ms step_avg:171.87ms
step:1196/1530 train_loss:3.3901 train_time:203855ms step_avg:171.88ms
step:1197/1530 train_loss:3.2978 train_time:204034ms step_avg:171.89ms
step:1198/1530 train_loss:3.3078 train_time:204220ms step_avg:171.90ms
step:1199/1530 train_loss:3.3514 train_time:204399ms step_avg:171.91ms
step:1200/1530 train_loss:3.4570 train_time:204575ms step_avg:171.91ms
step:1201/1530 train_loss:3.4858 train_time:204752ms step_avg:171.92ms
step:1202/1530 train_loss:3.6358 train_time:204941ms step_avg:171.93ms
step:1203/1530 train_loss:3.4096 train_time:205120ms step_avg:171.94ms
step:1204/1530 train_loss:3.3145 train_time:205301ms step_avg:171.94ms
step:1205/1530 train_loss:3.4421 train_time:205477ms step_avg:171.95ms
step:1206/1530 train_loss:3.4858 train_time:205654ms step_avg:171.95ms
step:1207/1530 train_loss:3.5259 train_time:205830ms step_avg:171.96ms
step:1208/1530 train_loss:3.4021 train_time:206006ms step_avg:171.96ms
step:1209/1530 train_loss:3.2510 train_time:206185ms step_avg:171.96ms
step:1210/1530 train_loss:3.3101 train_time:206364ms step_avg:171.97ms
step:1211/1530 train_loss:3.4026 train_time:206542ms step_avg:171.98ms
step:1212/1530 train_loss:3.3975 train_time:206721ms step_avg:171.98ms
step:1213/1530 train_loss:3.4127 train_time:206901ms step_avg:171.99ms
step:1214/1530 train_loss:3.2562 train_time:207082ms step_avg:171.99ms
step:1215/1530 train_loss:3.4013 train_time:207256ms step_avg:172.00ms
step:1216/1530 train_loss:3.3368 train_time:207434ms step_avg:172.00ms
step:1217/1530 train_loss:3.3318 train_time:207611ms step_avg:172.01ms
step:1218/1530 train_loss:3.4140 train_time:207787ms step_avg:172.01ms
step:1219/1530 train_loss:3.2649 train_time:207971ms step_avg:172.02ms
step:1220/1530 train_loss:3.4815 train_time:208147ms step_avg:172.02ms
step:1221/1530 train_loss:3.5082 train_time:208322ms step_avg:172.02ms
step:1222/1530 train_loss:3.4371 train_time:208498ms step_avg:172.03ms
step:1223/1530 train_loss:3.3035 train_time:208676ms step_avg:172.03ms
step:1224/1530 train_loss:3.2579 train_time:208858ms step_avg:172.04ms
step:1225/1530 train_loss:3.3755 train_time:209035ms step_avg:172.05ms
step:1226/1530 train_loss:3.3398 train_time:209214ms step_avg:172.05ms
step:1227/1530 train_loss:3.2805 train_time:209394ms step_avg:172.06ms
step:1228/1530 train_loss:3.4481 train_time:209571ms step_avg:172.06ms
step:1229/1530 train_loss:3.3711 train_time:209749ms step_avg:172.07ms
step:1230/1530 train_loss:3.4094 train_time:209930ms step_avg:172.07ms
step:1231/1530 train_loss:3.5848 train_time:210111ms step_avg:172.08ms
step:1232/1530 train_loss:3.5055 train_time:210291ms step_avg:172.09ms
step:1233/1530 train_loss:3.4317 train_time:210467ms step_avg:172.09ms
step:1234/1530 train_loss:3.5867 train_time:210646ms step_avg:172.10ms
step:1235/1530 train_loss:3.3290 train_time:210827ms step_avg:172.10ms
step:1236/1530 train_loss:3.2902 train_time:211003ms step_avg:172.11ms
step:1237/1530 train_loss:3.2785 train_time:211180ms step_avg:172.11ms
step:1238/1530 train_loss:3.2786 train_time:211364ms step_avg:172.12ms
step:1239/1530 train_loss:3.3379 train_time:211545ms step_avg:172.13ms
step:1240/1530 train_loss:3.3890 train_time:211724ms step_avg:172.13ms
step:1241/1530 train_loss:3.4277 train_time:211904ms step_avg:172.14ms
step:1242/1530 train_loss:3.3072 train_time:212082ms step_avg:172.14ms
step:1243/1530 train_loss:3.4135 train_time:212262ms step_avg:172.15ms
step:1244/1530 train_loss:3.4131 train_time:212436ms step_avg:172.15ms
step:1245/1530 train_loss:3.4148 train_time:212613ms step_avg:172.16ms
step:1246/1530 train_loss:3.2505 train_time:212790ms step_avg:172.16ms
step:1247/1530 train_loss:3.3796 train_time:212965ms step_avg:172.16ms
step:1248/1530 train_loss:3.4319 train_time:213142ms step_avg:172.17ms
step:1249/1530 train_loss:3.4284 train_time:213321ms step_avg:172.17ms
step:1250/1530 train_loss:3.3141 train_time:213500ms step_avg:172.18ms
step:1250/1530 val_loss:3.3608 train_time:213554ms step_avg:172.22ms
step:1251/1530 train_loss:3.4989 train_time:213686ms step_avg:172.19ms
step:1252/1530 train_loss:3.3647 train_time:213862ms step_avg:172.19ms
step:1253/1530 train_loss:3.3155 train_time:214038ms step_avg:172.19ms
step:1254/1530 train_loss:3.4221 train_time:214219ms step_avg:172.20ms
step:1255/1530 train_loss:3.5202 train_time:214408ms step_avg:172.21ms
step:1256/1530 train_loss:3.3137 train_time:214590ms step_avg:172.22ms
step:1257/1530 train_loss:3.3813 train_time:214768ms step_avg:172.23ms
step:1258/1530 train_loss:3.3738 train_time:214951ms step_avg:172.24ms
step:1259/1530 train_loss:3.3318 train_time:215130ms step_avg:172.24ms
step:1260/1530 train_loss:3.2151 train_time:215309ms step_avg:172.25ms
step:1261/1530 train_loss:3.3138 train_time:215490ms step_avg:172.25ms
step:1262/1530 train_loss:3.3260 train_time:215674ms step_avg:172.26ms
step:1263/1530 train_loss:3.2456 train_time:215855ms step_avg:172.27ms
step:1264/1530 train_loss:3.4482 train_time:216031ms step_avg:172.27ms
step:1265/1530 train_loss:3.4310 train_time:216206ms step_avg:172.28ms
step:1266/1530 train_loss:3.4459 train_time:216385ms step_avg:172.28ms
step:1267/1530 train_loss:3.3767 train_time:216565ms step_avg:172.29ms
step:1268/1530 train_loss:3.4163 train_time:216744ms step_avg:172.29ms
step:1269/1530 train_loss:3.2607 train_time:216926ms step_avg:172.30ms
step:1270/1530 train_loss:3.1076 train_time:217103ms step_avg:172.30ms
step:1271/1530 train_loss:3.4126 train_time:217283ms step_avg:172.31ms
step:1272/1530 train_loss:3.3600 train_time:217458ms step_avg:172.31ms
step:1273/1530 train_loss:3.3851 train_time:217638ms step_avg:172.32ms
step:1274/1530 train_loss:3.3649 train_time:217818ms step_avg:172.32ms
step:1275/1530 train_loss:3.4387 train_time:217994ms step_avg:172.33ms
step:1276/1530 train_loss:3.4734 train_time:218168ms step_avg:172.33ms
step:1277/1530 train_loss:3.4148 train_time:218348ms step_avg:172.33ms
step:1278/1530 train_loss:3.4122 train_time:218523ms step_avg:172.34ms
step:1279/1530 train_loss:3.2704 train_time:218706ms step_avg:172.35ms
step:1280/1530 train_loss:3.3741 train_time:218892ms step_avg:172.36ms
step:1281/1530 train_loss:3.4330 train_time:219070ms step_avg:172.36ms
step:1282/1530 train_loss:3.4799 train_time:219246ms step_avg:172.36ms
step:1283/1530 train_loss:3.3397 train_time:219426ms step_avg:172.37ms
step:1284/1530 train_loss:3.3743 train_time:219605ms step_avg:172.37ms
step:1285/1530 train_loss:3.3688 train_time:219784ms step_avg:172.38ms
step:1286/1530 train_loss:3.3376 train_time:219961ms step_avg:172.38ms
step:1287/1530 train_loss:3.4944 train_time:220139ms step_avg:172.39ms
step:1288/1530 train_loss:3.2989 train_time:220318ms step_avg:172.39ms
step:1289/1530 train_loss:3.3860 train_time:220505ms step_avg:172.40ms
step:1290/1530 train_loss:3.4639 train_time:220690ms step_avg:172.41ms
step:1291/1530 train_loss:3.3901 train_time:220870ms step_avg:172.42ms
step:1292/1530 train_loss:3.4862 train_time:221052ms step_avg:172.43ms
step:1293/1530 train_loss:3.5209 train_time:221231ms step_avg:172.43ms
step:1294/1530 train_loss:3.4637 train_time:221411ms step_avg:172.44ms
step:1295/1530 train_loss:3.2871 train_time:221590ms step_avg:172.44ms
step:1296/1530 train_loss:3.3754 train_time:221773ms step_avg:172.45ms
step:1297/1530 train_loss:3.2836 train_time:221952ms step_avg:172.46ms
step:1298/1530 train_loss:3.2767 train_time:222133ms step_avg:172.46ms
step:1299/1530 train_loss:3.4068 train_time:222311ms step_avg:172.47ms
step:1300/1530 train_loss:3.4121 train_time:222487ms step_avg:172.47ms
step:1301/1530 train_loss:3.4086 train_time:222665ms step_avg:172.47ms
step:1302/1530 train_loss:3.5832 train_time:222848ms step_avg:172.48ms
step:1303/1530 train_loss:3.3159 train_time:223030ms step_avg:172.49ms
step:1304/1530 train_loss:3.5215 train_time:223209ms step_avg:172.50ms
step:1305/1530 train_loss:3.2672 train_time:223385ms step_avg:172.50ms
step:1306/1530 train_loss:3.4610 train_time:223568ms step_avg:172.51ms
step:1307/1530 train_loss:3.4641 train_time:223742ms step_avg:172.51ms
step:1308/1530 train_loss:3.2937 train_time:223921ms step_avg:172.51ms
step:1309/1530 train_loss:3.3144 train_time:224100ms step_avg:172.52ms
step:1310/1530 train_loss:3.2948 train_time:224278ms step_avg:172.52ms
step:1311/1530 train_loss:3.3053 train_time:224454ms step_avg:172.52ms
step:1312/1530 train_loss:3.3835 train_time:224634ms step_avg:172.53ms
step:1313/1530 train_loss:3.3484 train_time:224810ms step_avg:172.53ms
step:1314/1530 train_loss:3.0522 train_time:224993ms step_avg:172.54ms
step:1315/1530 train_loss:3.2843 train_time:225169ms step_avg:172.54ms
step:1316/1530 train_loss:3.4036 train_time:225344ms step_avg:172.55ms
step:1317/1530 train_loss:3.4263 train_time:225521ms step_avg:172.55ms
step:1318/1530 train_loss:3.3070 train_time:225708ms step_avg:172.56ms
step:1319/1530 train_loss:3.4355 train_time:225888ms step_avg:172.57ms
step:1320/1530 train_loss:3.4650 train_time:226070ms step_avg:172.57ms
step:1321/1530 train_loss:3.3674 train_time:226250ms step_avg:172.58ms
step:1322/1530 train_loss:3.3296 train_time:226560ms step_avg:172.68ms
step:1323/1530 train_loss:3.3240 train_time:226750ms step_avg:172.70ms
step:1324/1530 train_loss:3.4451 train_time:226932ms step_avg:172.70ms
step:1325/1530 train_loss:3.4984 train_time:227118ms step_avg:172.71ms
step:1326/1530 train_loss:3.2206 train_time:227297ms step_avg:172.72ms
step:1327/1530 train_loss:3.1744 train_time:227473ms step_avg:172.72ms
step:1328/1530 train_loss:3.5006 train_time:227652ms step_avg:172.73ms
step:1329/1530 train_loss:3.3110 train_time:227988ms step_avg:172.85ms
step:1330/1530 train_loss:3.4367 train_time:228169ms step_avg:172.86ms
step:1331/1530 train_loss:3.3359 train_time:228345ms step_avg:172.86ms
step:1332/1530 train_loss:3.7482 train_time:228528ms step_avg:172.87ms
step:1333/1530 train_loss:3.4854 train_time:228711ms step_avg:172.87ms
step:1334/1530 train_loss:3.3781 train_time:228889ms step_avg:172.88ms
step:1335/1530 train_loss:3.2963 train_time:229069ms step_avg:172.88ms
step:1336/1530 train_loss:3.3062 train_time:229254ms step_avg:172.89ms
step:1337/1530 train_loss:3.5588 train_time:229433ms step_avg:172.90ms
step:1338/1530 train_loss:3.5269 train_time:229612ms step_avg:172.90ms
step:1339/1530 train_loss:3.3452 train_time:229790ms step_avg:172.90ms
step:1340/1530 train_loss:3.2902 train_time:229969ms step_avg:172.91ms
step:1341/1530 train_loss:3.6033 train_time:230147ms step_avg:172.91ms
step:1342/1530 train_loss:3.3602 train_time:230327ms step_avg:172.92ms
step:1343/1530 train_loss:3.3673 train_time:230504ms step_avg:172.92ms
step:1344/1530 train_loss:3.4201 train_time:230687ms step_avg:172.93ms
step:1345/1530 train_loss:3.3894 train_time:230869ms step_avg:172.94ms
step:1346/1530 train_loss:3.3043 train_time:231046ms step_avg:172.94ms
step:1347/1530 train_loss:3.2832 train_time:231223ms step_avg:172.94ms
step:1348/1530 train_loss:3.3587 train_time:231402ms step_avg:172.95ms
step:1349/1530 train_loss:3.2817 train_time:231578ms step_avg:172.95ms
step:1350/1530 train_loss:3.3994 train_time:231759ms step_avg:172.95ms
step:1351/1530 train_loss:3.2542 train_time:231935ms step_avg:172.96ms
step:1352/1530 train_loss:3.3129 train_time:232112ms step_avg:172.96ms
step:1353/1530 train_loss:3.4114 train_time:232292ms step_avg:172.96ms
step:1354/1530 train_loss:3.2652 train_time:232470ms step_avg:172.97ms
step:1355/1530 train_loss:3.1952 train_time:232647ms step_avg:172.97ms
step:1356/1530 train_loss:3.5149 train_time:232828ms step_avg:172.98ms
step:1357/1530 train_loss:3.4279 train_time:233009ms step_avg:172.98ms
step:1358/1530 train_loss:3.1931 train_time:233187ms step_avg:172.99ms
step:1359/1530 train_loss:3.4467 train_time:233367ms step_avg:172.99ms
step:1360/1530 train_loss:3.3558 train_time:233547ms step_avg:173.00ms
step:1361/1530 train_loss:3.1353 train_time:233733ms step_avg:173.01ms
step:1362/1530 train_loss:3.4002 train_time:233915ms step_avg:173.01ms
step:1363/1530 train_loss:3.2959 train_time:234103ms step_avg:173.03ms
step:1364/1530 train_loss:3.3091 train_time:234282ms step_avg:173.03ms
step:1365/1530 train_loss:3.3200 train_time:234461ms step_avg:173.03ms
step:1366/1530 train_loss:3.4329 train_time:234640ms step_avg:173.04ms
step:1367/1530 train_loss:3.4073 train_time:234817ms step_avg:173.04ms
step:1368/1530 train_loss:3.3524 train_time:234997ms step_avg:173.05ms
step:1369/1530 train_loss:3.2815 train_time:235185ms step_avg:173.06ms
step:1370/1530 train_loss:3.6115 train_time:235365ms step_avg:173.06ms
step:1371/1530 train_loss:3.3236 train_time:235546ms step_avg:173.07ms
step:1372/1530 train_loss:3.3789 train_time:235730ms step_avg:173.08ms
step:1373/1530 train_loss:3.3777 train_time:235914ms step_avg:173.08ms
step:1374/1530 train_loss:3.1594 train_time:236094ms step_avg:173.09ms
step:1375/1530 train_loss:3.5461 train_time:236274ms step_avg:173.09ms
step:1375/1530 val_loss:3.3194 train_time:236325ms step_avg:173.13ms
step:1376/1530 train_loss:3.3536 train_time:236452ms step_avg:173.10ms
step:1377/1530 train_loss:3.4875 train_time:236632ms step_avg:173.10ms
step:1378/1530 train_loss:3.4776 train_time:236810ms step_avg:173.11ms
step:1379/1530 train_loss:3.1232 train_time:236993ms step_avg:173.11ms
step:1380/1530 train_loss:3.3211 train_time:237173ms step_avg:173.12ms
step:1381/1530 train_loss:3.7034 train_time:237357ms step_avg:173.13ms
step:1382/1530 train_loss:3.2153 train_time:237536ms step_avg:173.13ms
step:1383/1530 train_loss:3.3994 train_time:237718ms step_avg:173.14ms
step:1384/1530 train_loss:3.4863 train_time:237903ms step_avg:173.15ms
step:1385/1530 train_loss:3.4164 train_time:238078ms step_avg:173.15ms
step:1386/1530 train_loss:3.3471 train_time:238257ms step_avg:173.15ms
step:1387/1530 train_loss:3.2083 train_time:238437ms step_avg:173.16ms
step:1388/1530 train_loss:3.3550 train_time:238614ms step_avg:173.16ms
step:1389/1530 train_loss:3.3254 train_time:238798ms step_avg:173.17ms
step:1390/1530 train_loss:3.5758 train_time:238975ms step_avg:173.17ms
step:1391/1530 train_loss:3.3035 train_time:239155ms step_avg:173.17ms
step:1392/1530 train_loss:3.2936 train_time:239334ms step_avg:173.18ms
step:1393/1530 train_loss:3.2433 train_time:239514ms step_avg:173.18ms
step:1394/1530 train_loss:3.5056 train_time:239692ms step_avg:173.19ms
step:1395/1530 train_loss:3.3997 train_time:239871ms step_avg:173.19ms
step:1396/1530 train_loss:3.4158 train_time:240049ms step_avg:173.20ms
step:1397/1530 train_loss:3.3167 train_time:240225ms step_avg:173.20ms
step:1398/1530 train_loss:3.2651 train_time:240401ms step_avg:173.20ms
step:1399/1530 train_loss:3.3254 train_time:240578ms step_avg:173.20ms
step:1400/1530 train_loss:3.3260 train_time:240762ms step_avg:173.21ms
step:1401/1530 train_loss:3.3566 train_time:240938ms step_avg:173.21ms
step:1402/1530 train_loss:3.3088 train_time:241118ms step_avg:173.22ms
step:1403/1530 train_loss:3.4987 train_time:241304ms step_avg:173.23ms
step:1404/1530 train_loss:3.2884 train_time:241481ms step_avg:173.23ms
step:1405/1530 train_loss:3.3224 train_time:241663ms step_avg:173.23ms
step:1406/1530 train_loss:3.3245 train_time:241843ms step_avg:173.24ms
step:1407/1530 train_loss:3.1841 train_time:242020ms step_avg:173.24ms
step:1408/1530 train_loss:3.3198 train_time:242200ms step_avg:173.25ms
step:1409/1530 train_loss:3.3109 train_time:242387ms step_avg:173.26ms
step:1410/1530 train_loss:3.2971 train_time:242565ms step_avg:173.26ms
step:1411/1530 train_loss:3.3744 train_time:242742ms step_avg:173.26ms
step:1412/1530 train_loss:3.3424 train_time:242919ms step_avg:173.27ms
step:1413/1530 train_loss:3.3726 train_time:243098ms step_avg:173.27ms
step:1414/1530 train_loss:3.3358 train_time:243278ms step_avg:173.27ms
step:1415/1530 train_loss:3.4131 train_time:243462ms step_avg:173.28ms
step:1416/1530 train_loss:3.2372 train_time:243651ms step_avg:173.29ms
step:1417/1530 train_loss:3.2937 train_time:243834ms step_avg:173.30ms
step:1418/1530 train_loss:3.3980 train_time:244014ms step_avg:173.31ms
step:1419/1530 train_loss:3.3436 train_time:244196ms step_avg:173.31ms
step:1420/1530 train_loss:3.3705 train_time:244378ms step_avg:173.32ms
step:1421/1530 train_loss:3.3764 train_time:244559ms step_avg:173.32ms
step:1422/1530 train_loss:3.3410 train_time:244735ms step_avg:173.33ms
step:1423/1530 train_loss:3.3240 train_time:244916ms step_avg:173.33ms
step:1424/1530 train_loss:3.3414 train_time:245101ms step_avg:173.34ms
step:1425/1530 train_loss:3.2014 train_time:245289ms step_avg:173.35ms
step:1426/1530 train_loss:3.3323 train_time:245467ms step_avg:173.35ms
step:1427/1530 train_loss:3.2933 train_time:245650ms step_avg:173.36ms
step:1428/1530 train_loss:3.3879 train_time:245827ms step_avg:173.36ms
step:1429/1530 train_loss:3.3595 train_time:246003ms step_avg:173.36ms
step:1430/1530 train_loss:3.2655 train_time:246185ms step_avg:173.37ms
step:1431/1530 train_loss:3.3299 train_time:246368ms step_avg:173.38ms
step:1432/1530 train_loss:3.3440 train_time:246549ms step_avg:173.38ms
step:1433/1530 train_loss:3.1365 train_time:246733ms step_avg:173.39ms
step:1434/1530 train_loss:3.2991 train_time:246917ms step_avg:173.40ms
step:1435/1530 train_loss:3.1224 train_time:247098ms step_avg:173.40ms
step:1436/1530 train_loss:3.2391 train_time:247277ms step_avg:173.41ms
step:1437/1530 train_loss:3.4162 train_time:247455ms step_avg:173.41ms
step:1438/1530 train_loss:3.3887 train_time:247632ms step_avg:173.41ms
step:1439/1530 train_loss:3.3264 train_time:247812ms step_avg:173.42ms
step:1440/1530 train_loss:3.1980 train_time:247988ms step_avg:173.42ms
step:1441/1530 train_loss:3.3461 train_time:248168ms step_avg:173.42ms
step:1442/1530 train_loss:3.3980 train_time:248352ms step_avg:173.43ms
step:1443/1530 train_loss:3.4987 train_time:248540ms step_avg:173.44ms
step:1444/1530 train_loss:3.4562 train_time:248718ms step_avg:173.44ms
step:1445/1530 train_loss:3.3469 train_time:248895ms step_avg:173.45ms
step:1446/1530 train_loss:3.2055 train_time:249076ms step_avg:173.45ms
step:1447/1530 train_loss:3.3052 train_time:249258ms step_avg:173.46ms
step:1448/1530 train_loss:3.3068 train_time:249436ms step_avg:173.46ms
step:1449/1530 train_loss:3.4014 train_time:249616ms step_avg:173.46ms
step:1450/1530 train_loss:3.3975 train_time:249798ms step_avg:173.47ms
step:1451/1530 train_loss:3.2111 train_time:249977ms step_avg:173.47ms
step:1452/1530 train_loss:3.3317 train_time:250156ms step_avg:173.48ms
step:1453/1530 train_loss:3.2688 train_time:250331ms step_avg:173.48ms
step:1454/1530 train_loss:3.2980 train_time:250509ms step_avg:173.48ms
step:1455/1530 train_loss:3.3388 train_time:250691ms step_avg:173.49ms
step:1456/1530 train_loss:3.2917 train_time:250869ms step_avg:173.49ms
step:1457/1530 train_loss:3.1617 train_time:251046ms step_avg:173.49ms
step:1458/1530 train_loss:3.4314 train_time:251225ms step_avg:173.50ms
step:1459/1530 train_loss:3.2748 train_time:251406ms step_avg:173.50ms
step:1460/1530 train_loss:3.3250 train_time:251584ms step_avg:173.51ms
step:1461/1530 train_loss:3.4381 train_time:251765ms step_avg:173.51ms
step:1462/1530 train_loss:3.2679 train_time:251941ms step_avg:173.51ms
step:1463/1530 train_loss:3.4709 train_time:252123ms step_avg:173.52ms
step:1464/1530 train_loss:3.3674 train_time:252302ms step_avg:173.52ms
step:1465/1530 train_loss:3.3657 train_time:252482ms step_avg:173.53ms
step:1466/1530 train_loss:3.2967 train_time:252660ms step_avg:173.53ms
step:1467/1530 train_loss:3.4026 train_time:252838ms step_avg:173.53ms
step:1468/1530 train_loss:3.2960 train_time:253014ms step_avg:173.54ms
step:1469/1530 train_loss:3.2828 train_time:253194ms step_avg:173.54ms
step:1470/1530 train_loss:3.3367 train_time:253377ms step_avg:173.55ms
step:1471/1530 train_loss:3.2664 train_time:253563ms step_avg:173.55ms
step:1472/1530 train_loss:3.2520 train_time:253747ms step_avg:173.56ms
step:1473/1530 train_loss:3.4455 train_time:253924ms step_avg:173.56ms
step:1474/1530 train_loss:3.3199 train_time:254107ms step_avg:173.57ms
step:1475/1530 train_loss:3.1575 train_time:254293ms step_avg:173.58ms
step:1476/1530 train_loss:3.2717 train_time:254472ms step_avg:173.58ms
step:1477/1530 train_loss:3.2473 train_time:254659ms step_avg:173.59ms
step:1478/1530 train_loss:3.3153 train_time:254844ms step_avg:173.60ms
step:1479/1530 train_loss:3.4039 train_time:255026ms step_avg:173.61ms
step:1480/1530 train_loss:3.2753 train_time:255204ms step_avg:173.61ms
step:1481/1530 train_loss:3.4598 train_time:255386ms step_avg:173.61ms
step:1482/1530 train_loss:3.3741 train_time:255573ms step_avg:173.62ms
step:1483/1530 train_loss:3.2874 train_time:255763ms step_avg:173.63ms
step:1484/1530 train_loss:3.2651 train_time:255950ms step_avg:173.64ms
step:1485/1530 train_loss:3.2917 train_time:256130ms step_avg:173.65ms
step:1486/1530 train_loss:3.2330 train_time:256314ms step_avg:173.65ms
step:1487/1530 train_loss:3.3425 train_time:256497ms step_avg:173.66ms
step:1488/1530 train_loss:3.2477 train_time:256679ms step_avg:173.67ms
step:1489/1530 train_loss:3.3223 train_time:256860ms step_avg:173.67ms
step:1490/1530 train_loss:3.2591 train_time:257040ms step_avg:173.68ms
step:1491/1530 train_loss:3.1705 train_time:257220ms step_avg:173.68ms
step:1492/1530 train_loss:3.2817 train_time:257399ms step_avg:173.68ms
step:1493/1530 train_loss:3.4399 train_time:257578ms step_avg:173.69ms
step:1494/1530 train_loss:3.3091 train_time:257757ms step_avg:173.69ms
step:1495/1530 train_loss:3.0373 train_time:257943ms step_avg:173.70ms
step:1496/1530 train_loss:3.3682 train_time:258126ms step_avg:173.71ms
step:1497/1530 train_loss:3.3227 train_time:258309ms step_avg:173.71ms
step:1498/1530 train_loss:3.3530 train_time:258496ms step_avg:173.72ms
step:1499/1530 train_loss:3.3188 train_time:258684ms step_avg:173.73ms
step:1500/1530 train_loss:3.3079 train_time:258877ms step_avg:173.74ms
step:1500/1530 val_loss:3.2875 train_time:258932ms step_avg:173.78ms
step:1501/1530 train_loss:3.0941 train_time:259069ms step_avg:173.75ms
step:1502/1530 train_loss:3.3690 train_time:259258ms step_avg:173.77ms
step:1503/1530 train_loss:3.2536 train_time:259438ms step_avg:173.77ms
step:1504/1530 train_loss:3.2545 train_time:259618ms step_avg:173.77ms
step:1505/1530 train_loss:3.2230 train_time:259798ms step_avg:173.78ms
step:1506/1530 train_loss:3.2879 train_time:259980ms step_avg:173.78ms
step:1507/1530 train_loss:3.1828 train_time:260177ms step_avg:173.80ms
step:1508/1530 train_loss:3.4859 train_time:260360ms step_avg:173.80ms
step:1509/1530 train_loss:3.2891 train_time:260537ms step_avg:173.81ms
step:1510/1530 train_loss:3.2811 train_time:260717ms step_avg:173.81ms
step:1511/1530 train_loss:3.4222 train_time:261027ms step_avg:173.90ms
step:1512/1530 train_loss:3.4229 train_time:261215ms step_avg:173.91ms
step:1513/1530 train_loss:3.2781 train_time:261400ms step_avg:173.92ms
step:1514/1530 train_loss:3.0959 train_time:261583ms step_avg:173.92ms
step:1515/1530 train_loss:3.2503 train_time:261763ms step_avg:173.93ms
step:1516/1530 train_loss:3.2639 train_time:261948ms step_avg:173.94ms
step:1517/1530 train_loss:3.3086 train_time:262129ms step_avg:173.94ms
step:1518/1530 train_loss:3.2103 train_time:262312ms step_avg:173.95ms
step:1519/1530 train_loss:3.5098 train_time:262636ms step_avg:174.05ms
step:1520/1530 train_loss:3.1354 train_time:262820ms step_avg:174.05ms
step:1521/1530 train_loss:3.2117 train_time:262996ms step_avg:174.05ms
step:1522/1530 train_loss:3.3635 train_time:263180ms step_avg:174.06ms
step:1523/1530 train_loss:3.2390 train_time:263358ms step_avg:174.06ms
step:1524/1530 train_loss:3.3592 train_time:263538ms step_avg:174.07ms
step:1525/1530 train_loss:3.3496 train_time:263725ms step_avg:174.08ms
step:1526/1530 train_loss:3.2859 train_time:263915ms step_avg:174.09ms
step:1527/1530 train_loss:3.3009 train_time:264096ms step_avg:174.09ms
step:1528/1530 train_loss:3.4168 train_time:264277ms step_avg:174.10ms
step:1529/1530 train_loss:3.4148 train_time:264455ms step_avg:174.10ms
step:1530/1530 train_loss:3.2449 train_time:264636ms step_avg:174.10ms
step:1530/1530 val_loss:3.2851 train_time:264690ms step_avg:174.14ms