2e527b0b-3540-4bcd-a15d-955f86cb8bd2.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 01:16:35 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.06             Driver Version: 535.183.06   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA H100 80GB HBM3          On  | 00000000:19:00.0 Off |                    0 |
| N/A   38C    P0              75W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          On  | 00000000:3B:00.0 Off |                    0 |
| N/A   30C    P0             107W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          On  | 00000000:4C:00.0 Off |                    0 |
| N/A   30C    P0             117W / 700W |     41MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   37C    P0              92W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   38C    P0              88W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   29C    P0             109W / 700W |     39MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   38C    P0             110W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   29C    P0             104W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
+---------------------------------------------------------------------------------------+

====================================================================================================
Training DataLoader: total number of tokens: 1100000000 across 11 files
Validation DataLoader: total number of tokens: 100000000 across 1 files
====================================================================================================
step:0/1530 val_loss:10.8258 train_time:0ms step_avg:nanms
step:1/1530 train_loss:10.8258 train_time:32524ms step_avg:nanms
step:2/1530 train_loss:10.0751 train_time:32636ms step_avg:nanms
step:3/1530 train_loss:8.3516 train_time:32796ms step_avg:nanms
step:4/1530 train_loss:7.6118 train_time:32959ms step_avg:nanms
step:5/1530 train_loss:7.5069 train_time:33119ms step_avg:nanms
step:6/1530 train_loss:7.0192 train_time:33280ms step_avg:nanms
step:7/1530 train_loss:7.1690 train_time:33441ms step_avg:nanms
step:8/1530 train_loss:6.7471 train_time:33602ms step_avg:nanms
step:9/1530 train_loss:6.6580 train_time:33763ms step_avg:nanms
step:10/1530 train_loss:6.6250 train_time:33922ms step_avg:nanms
step:11/1530 train_loss:6.4494 train_time:116ms step_avg:nanms
step:12/1530 train_loss:6.3327 train_time:276ms step_avg:nanms
step:13/1530 train_loss:6.2269 train_time:437ms step_avg:145.53ms
step:14/1530 train_loss:6.2111 train_time:597ms step_avg:149.35ms
step:15/1530 train_loss:6.1673 train_time:757ms step_avg:151.42ms
step:16/1530 train_loss:6.1258 train_time:918ms step_avg:152.99ms
step:17/1530 train_loss:6.1546 train_time:1078ms step_avg:154.07ms
step:18/1530 train_loss:5.9418 train_time:1239ms step_avg:154.82ms
step:19/1530 train_loss:5.9595 train_time:1399ms step_avg:155.45ms
step:20/1530 train_loss:5.6948 train_time:1558ms step_avg:155.84ms
step:21/1530 train_loss:5.9505 train_time:1719ms step_avg:156.31ms
step:22/1530 train_loss:6.1680 train_time:1879ms step_avg:156.60ms
step:23/1530 train_loss:5.8572 train_time:2039ms step_avg:156.88ms
step:24/1530 train_loss:6.0206 train_time:2201ms step_avg:157.20ms
step:25/1530 train_loss:5.6700 train_time:2360ms step_avg:157.31ms
step:26/1530 train_loss:5.5931 train_time:2520ms step_avg:157.48ms
step:27/1530 train_loss:5.7516 train_time:2681ms step_avg:157.68ms
step:28/1530 train_loss:5.4139 train_time:2840ms step_avg:157.80ms
step:29/1530 train_loss:5.6570 train_time:3000ms step_avg:157.91ms
step:30/1530 train_loss:5.4594 train_time:3161ms step_avg:158.06ms
step:31/1530 train_loss:5.4344 train_time:3321ms step_avg:158.13ms
step:32/1530 train_loss:5.2915 train_time:3484ms step_avg:158.35ms
step:33/1530 train_loss:5.5739 train_time:3643ms step_avg:158.38ms
step:34/1530 train_loss:5.4857 train_time:3804ms step_avg:158.49ms
step:35/1530 train_loss:5.5961 train_time:3964ms step_avg:158.56ms
step:36/1530 train_loss:5.5413 train_time:4123ms step_avg:158.58ms
step:37/1530 train_loss:5.4507 train_time:4284ms step_avg:158.66ms
step:38/1530 train_loss:5.2959 train_time:4445ms step_avg:158.74ms
step:39/1530 train_loss:5.3059 train_time:4604ms step_avg:158.75ms
step:40/1530 train_loss:5.2314 train_time:4764ms step_avg:158.80ms
step:41/1530 train_loss:5.2101 train_time:4923ms step_avg:158.81ms
step:42/1530 train_loss:5.1635 train_time:5085ms step_avg:158.89ms
step:43/1530 train_loss:5.2685 train_time:5243ms step_avg:158.87ms
step:44/1530 train_loss:5.2279 train_time:5404ms step_avg:158.96ms
step:45/1530 train_loss:5.3812 train_time:5565ms step_avg:158.99ms
step:46/1530 train_loss:5.1668 train_time:5724ms step_avg:159.00ms
step:47/1530 train_loss:5.0469 train_time:5885ms step_avg:159.05ms
step:48/1530 train_loss:5.2052 train_time:6046ms step_avg:159.10ms
step:49/1530 train_loss:5.1382 train_time:6206ms step_avg:159.12ms
step:50/1530 train_loss:5.2369 train_time:6365ms step_avg:159.12ms
step:51/1530 train_loss:5.1316 train_time:6527ms step_avg:159.20ms
step:52/1530 train_loss:5.0178 train_time:6688ms step_avg:159.25ms
step:53/1530 train_loss:5.1516 train_time:6848ms step_avg:159.25ms
step:54/1530 train_loss:4.9787 train_time:7009ms step_avg:159.30ms
step:55/1530 train_loss:5.3834 train_time:7170ms step_avg:159.34ms
step:56/1530 train_loss:5.0122 train_time:7332ms step_avg:159.38ms
step:57/1530 train_loss:4.8737 train_time:7493ms step_avg:159.43ms
step:58/1530 train_loss:5.0304 train_time:7653ms step_avg:159.44ms
step:59/1530 train_loss:5.0165 train_time:7814ms step_avg:159.47ms
step:60/1530 train_loss:5.1480 train_time:7974ms step_avg:159.48ms
step:61/1530 train_loss:4.8535 train_time:8136ms step_avg:159.52ms
step:62/1530 train_loss:4.9701 train_time:8297ms step_avg:159.55ms
step:63/1530 train_loss:4.9618 train_time:8456ms step_avg:159.55ms
step:64/1530 train_loss:4.9433 train_time:8616ms step_avg:159.56ms
step:65/1530 train_loss:4.7815 train_time:8777ms step_avg:159.58ms
step:66/1530 train_loss:4.9012 train_time:8938ms step_avg:159.60ms
step:67/1530 train_loss:4.8068 train_time:9098ms step_avg:159.62ms
step:68/1530 train_loss:5.1025 train_time:9258ms step_avg:159.62ms
step:69/1530 train_loss:4.7249 train_time:9419ms step_avg:159.64ms
step:70/1530 train_loss:4.8476 train_time:9578ms step_avg:159.64ms
step:71/1530 train_loss:4.9746 train_time:9738ms step_avg:159.64ms
step:72/1530 train_loss:4.8950 train_time:9899ms step_avg:159.66ms
step:73/1530 train_loss:4.7548 train_time:10059ms step_avg:159.67ms
step:74/1530 train_loss:4.8930 train_time:10219ms step_avg:159.68ms
step:75/1530 train_loss:4.8338 train_time:10379ms step_avg:159.68ms
step:76/1530 train_loss:4.7975 train_time:10539ms step_avg:159.68ms
step:77/1530 train_loss:4.9071 train_time:10699ms step_avg:159.69ms
step:78/1530 train_loss:5.1314 train_time:10860ms step_avg:159.71ms
step:79/1530 train_loss:4.7924 train_time:11020ms step_avg:159.70ms
step:80/1530 train_loss:4.8415 train_time:11180ms step_avg:159.71ms
step:81/1530 train_loss:4.6377 train_time:11340ms step_avg:159.71ms
step:82/1530 train_loss:4.8089 train_time:11500ms step_avg:159.72ms
step:83/1530 train_loss:4.7607 train_time:11659ms step_avg:159.72ms
step:84/1530 train_loss:4.7552 train_time:11819ms step_avg:159.72ms
step:85/1530 train_loss:4.6118 train_time:11980ms step_avg:159.73ms
step:86/1530 train_loss:4.8222 train_time:12139ms step_avg:159.72ms
step:87/1530 train_loss:4.7359 train_time:12300ms step_avg:159.74ms
step:88/1530 train_loss:4.7273 train_time:12460ms step_avg:159.74ms
step:89/1530 train_loss:4.7102 train_time:12620ms step_avg:159.74ms
step:90/1530 train_loss:4.6515 train_time:12780ms step_avg:159.75ms
step:91/1530 train_loss:4.6406 train_time:12939ms step_avg:159.74ms
step:92/1530 train_loss:4.8004 train_time:13101ms step_avg:159.77ms
step:93/1530 train_loss:4.6138 train_time:13260ms step_avg:159.76ms
step:94/1530 train_loss:4.6391 train_time:13420ms step_avg:159.76ms
step:95/1530 train_loss:4.6914 train_time:13580ms step_avg:159.76ms
step:96/1530 train_loss:4.5856 train_time:13740ms step_avg:159.77ms
step:97/1530 train_loss:4.6196 train_time:13900ms step_avg:159.77ms
step:98/1530 train_loss:4.5702 train_time:14059ms step_avg:159.76ms
step:99/1530 train_loss:4.6651 train_time:14219ms step_avg:159.77ms
step:100/1530 train_loss:4.6749 train_time:14379ms step_avg:159.77ms
step:101/1530 train_loss:4.5163 train_time:14540ms step_avg:159.78ms
step:102/1530 train_loss:4.6938 train_time:14700ms step_avg:159.79ms
step:103/1530 train_loss:4.5772 train_time:14860ms step_avg:159.78ms
step:104/1530 train_loss:4.5368 train_time:15021ms step_avg:159.79ms
step:105/1530 train_loss:4.5460 train_time:15180ms step_avg:159.79ms
step:106/1530 train_loss:4.5834 train_time:15340ms step_avg:159.79ms
step:107/1530 train_loss:4.5055 train_time:15500ms step_avg:159.80ms
step:108/1530 train_loss:4.3749 train_time:15660ms step_avg:159.80ms
step:109/1530 train_loss:4.4818 train_time:15819ms step_avg:159.79ms
step:110/1530 train_loss:4.4768 train_time:15980ms step_avg:159.80ms
step:111/1530 train_loss:4.4272 train_time:16141ms step_avg:159.81ms
step:112/1530 train_loss:4.5966 train_time:16300ms step_avg:159.80ms
step:113/1530 train_loss:4.4851 train_time:16459ms step_avg:159.80ms
step:114/1530 train_loss:4.3585 train_time:16619ms step_avg:159.80ms
step:115/1530 train_loss:4.5108 train_time:16782ms step_avg:159.83ms
step:116/1530 train_loss:4.4703 train_time:16945ms step_avg:159.86ms
step:117/1530 train_loss:4.3654 train_time:17110ms step_avg:159.90ms
step:118/1530 train_loss:4.6011 train_time:17273ms step_avg:159.94ms
step:119/1530 train_loss:4.4507 train_time:17437ms step_avg:159.97ms
step:120/1530 train_loss:4.3301 train_time:17601ms step_avg:160.01ms
step:121/1530 train_loss:4.2981 train_time:17763ms step_avg:160.03ms
step:122/1530 train_loss:4.4468 train_time:17927ms step_avg:160.06ms
step:123/1530 train_loss:4.2793 train_time:18091ms step_avg:160.10ms
step:124/1530 train_loss:4.5840 train_time:18255ms step_avg:160.13ms
step:125/1530 train_loss:4.4426 train_time:18419ms step_avg:160.17ms
step:125/1530 val_loss:4.3918 train_time:18467ms step_avg:160.58ms
step:126/1530 train_loss:4.4044 train_time:18585ms step_avg:160.22ms
step:127/1530 train_loss:4.4321 train_time:18751ms step_avg:160.27ms
step:128/1530 train_loss:4.3778 train_time:18915ms step_avg:160.30ms
step:129/1530 train_loss:4.6998 train_time:19079ms step_avg:160.33ms
step:130/1530 train_loss:4.3570 train_time:19243ms step_avg:160.36ms
step:131/1530 train_loss:4.3881 train_time:19406ms step_avg:160.38ms
step:132/1530 train_loss:4.3493 train_time:19572ms step_avg:160.42ms
step:133/1530 train_loss:4.4554 train_time:19736ms step_avg:160.46ms
step:134/1530 train_loss:4.2601 train_time:19899ms step_avg:160.48ms
step:135/1530 train_loss:4.4513 train_time:20064ms step_avg:160.51ms
step:136/1530 train_loss:4.2186 train_time:20228ms step_avg:160.54ms
step:137/1530 train_loss:4.3899 train_time:20392ms step_avg:160.57ms
step:138/1530 train_loss:4.2862 train_time:20556ms step_avg:160.59ms
step:139/1530 train_loss:4.3851 train_time:20719ms step_avg:160.61ms
step:140/1530 train_loss:4.4800 train_time:20882ms step_avg:160.63ms
step:141/1530 train_loss:4.3122 train_time:21045ms step_avg:160.65ms
step:142/1530 train_loss:4.3144 train_time:21210ms step_avg:160.68ms
step:143/1530 train_loss:4.2683 train_time:21375ms step_avg:160.71ms
step:144/1530 train_loss:4.3576 train_time:21539ms step_avg:160.74ms
step:145/1530 train_loss:4.3236 train_time:21702ms step_avg:160.76ms
step:146/1530 train_loss:4.1796 train_time:21867ms step_avg:160.79ms
step:147/1530 train_loss:4.3328 train_time:22031ms step_avg:160.81ms
step:148/1530 train_loss:4.3719 train_time:22195ms step_avg:160.83ms
step:149/1530 train_loss:4.3128 train_time:22360ms step_avg:160.86ms
step:150/1530 train_loss:4.4524 train_time:22523ms step_avg:160.88ms
step:151/1530 train_loss:4.2775 train_time:22686ms step_avg:160.89ms
step:152/1530 train_loss:4.2706 train_time:22851ms step_avg:160.92ms
step:153/1530 train_loss:4.3700 train_time:23015ms step_avg:160.95ms
step:154/1530 train_loss:4.3626 train_time:23180ms step_avg:160.97ms
step:155/1530 train_loss:4.2753 train_time:23344ms step_avg:160.99ms
step:156/1530 train_loss:4.3565 train_time:23508ms step_avg:161.01ms
step:157/1530 train_loss:4.4051 train_time:23672ms step_avg:161.03ms
step:158/1530 train_loss:4.2470 train_time:23837ms step_avg:161.06ms
step:159/1530 train_loss:4.3104 train_time:24001ms step_avg:161.08ms
step:160/1530 train_loss:4.1351 train_time:24164ms step_avg:161.09ms
step:161/1530 train_loss:4.3552 train_time:24327ms step_avg:161.11ms
step:162/1530 train_loss:4.3682 train_time:24490ms step_avg:161.12ms
step:163/1530 train_loss:4.3541 train_time:24655ms step_avg:161.14ms
step:164/1530 train_loss:4.1937 train_time:24818ms step_avg:161.16ms
step:165/1530 train_loss:4.2918 train_time:24982ms step_avg:161.17ms
step:166/1530 train_loss:4.3503 train_time:25145ms step_avg:161.19ms
step:167/1530 train_loss:4.2115 train_time:25309ms step_avg:161.21ms
step:168/1530 train_loss:4.2976 train_time:25474ms step_avg:161.23ms
step:169/1530 train_loss:4.1729 train_time:25638ms step_avg:161.24ms
step:170/1530 train_loss:4.0211 train_time:25802ms step_avg:161.26ms
step:171/1530 train_loss:4.2141 train_time:25965ms step_avg:161.27ms
step:172/1530 train_loss:4.2126 train_time:26127ms step_avg:161.28ms
step:173/1530 train_loss:4.2754 train_time:26290ms step_avg:161.29ms
step:174/1530 train_loss:4.4159 train_time:26454ms step_avg:161.30ms
step:175/1530 train_loss:4.2416 train_time:26617ms step_avg:161.31ms
step:176/1530 train_loss:4.0965 train_time:26779ms step_avg:161.32ms
step:177/1530 train_loss:4.0712 train_time:26942ms step_avg:161.33ms
step:178/1530 train_loss:4.1900 train_time:27103ms step_avg:161.33ms
step:179/1530 train_loss:4.1290 train_time:27266ms step_avg:161.34ms
step:180/1530 train_loss:4.1228 train_time:27429ms step_avg:161.35ms
step:181/1530 train_loss:4.3158 train_time:27592ms step_avg:161.36ms
step:182/1530 train_loss:4.1639 train_time:27755ms step_avg:161.37ms
step:183/1530 train_loss:4.1254 train_time:27918ms step_avg:161.37ms
step:184/1530 train_loss:4.1231 train_time:28081ms step_avg:161.38ms
step:185/1530 train_loss:4.2124 train_time:28243ms step_avg:161.39ms
step:186/1530 train_loss:4.1842 train_time:28404ms step_avg:161.39ms
step:187/1530 train_loss:4.2501 train_time:28568ms step_avg:161.40ms
step:188/1530 train_loss:4.1721 train_time:28875ms step_avg:162.22ms
step:189/1530 train_loss:4.1167 train_time:29212ms step_avg:163.20ms
step:190/1530 train_loss:4.2108 train_time:29376ms step_avg:163.20ms
step:191/1530 train_loss:4.0876 train_time:29538ms step_avg:163.20ms
step:192/1530 train_loss:4.0560 train_time:29700ms step_avg:163.19ms
step:193/1530 train_loss:4.2577 train_time:29863ms step_avg:163.19ms
step:194/1530 train_loss:4.1809 train_time:30026ms step_avg:163.18ms
step:195/1530 train_loss:4.3663 train_time:30188ms step_avg:163.18ms
step:196/1530 train_loss:4.1825 train_time:30353ms step_avg:163.19ms
step:197/1530 train_loss:4.0516 train_time:30517ms step_avg:163.19ms
step:198/1530 train_loss:4.1881 train_time:30680ms step_avg:163.19ms
step:199/1530 train_loss:4.0353 train_time:30843ms step_avg:163.19ms
step:200/1530 train_loss:4.1164 train_time:31006ms step_avg:163.19ms
step:201/1530 train_loss:4.0126 train_time:31168ms step_avg:163.18ms
step:202/1530 train_loss:4.2633 train_time:31332ms step_avg:163.19ms
step:203/1530 train_loss:4.0727 train_time:31496ms step_avg:163.19ms
step:204/1530 train_loss:4.1936 train_time:31659ms step_avg:163.19ms
step:205/1530 train_loss:4.2559 train_time:31822ms step_avg:163.19ms
step:206/1530 train_loss:3.9552 train_time:31984ms step_avg:163.19ms
step:207/1530 train_loss:4.0912 train_time:32146ms step_avg:163.18ms
step:208/1530 train_loss:4.1175 train_time:32309ms step_avg:163.18ms
step:209/1530 train_loss:4.2425 train_time:32472ms step_avg:163.18ms
step:210/1530 train_loss:4.1850 train_time:32636ms step_avg:163.18ms
step:211/1530 train_loss:4.0671 train_time:32799ms step_avg:163.18ms
step:212/1530 train_loss:4.1276 train_time:32961ms step_avg:163.18ms
step:213/1530 train_loss:4.0455 train_time:33123ms step_avg:163.17ms
step:214/1530 train_loss:4.1181 train_time:33285ms step_avg:163.16ms
step:215/1530 train_loss:3.9567 train_time:33447ms step_avg:163.16ms
step:216/1530 train_loss:4.0130 train_time:33611ms step_avg:163.16ms
step:217/1530 train_loss:4.0134 train_time:33775ms step_avg:163.16ms
step:218/1530 train_loss:4.0834 train_time:33938ms step_avg:163.16ms
step:219/1530 train_loss:4.0776 train_time:34100ms step_avg:163.16ms
step:220/1530 train_loss:4.0933 train_time:34263ms step_avg:163.16ms
step:221/1530 train_loss:4.0976 train_time:34425ms step_avg:163.15ms
step:222/1530 train_loss:4.0027 train_time:34587ms step_avg:163.15ms
step:223/1530 train_loss:3.9913 train_time:34751ms step_avg:163.15ms
step:224/1530 train_loss:4.3082 train_time:34914ms step_avg:163.15ms
step:225/1530 train_loss:3.9238 train_time:35078ms step_avg:163.15ms
step:226/1530 train_loss:3.9945 train_time:35240ms step_avg:163.15ms
step:227/1530 train_loss:3.9982 train_time:35403ms step_avg:163.15ms
step:228/1530 train_loss:4.1439 train_time:35567ms step_avg:163.15ms
step:229/1530 train_loss:3.9272 train_time:35734ms step_avg:163.17ms
step:230/1530 train_loss:4.0509 train_time:35900ms step_avg:163.18ms
step:231/1530 train_loss:3.9147 train_time:36065ms step_avg:163.19ms
step:232/1530 train_loss:3.9777 train_time:36233ms step_avg:163.21ms
step:233/1530 train_loss:4.1014 train_time:36398ms step_avg:163.22ms
step:234/1530 train_loss:4.0477 train_time:36564ms step_avg:163.23ms
step:235/1530 train_loss:3.9097 train_time:36730ms step_avg:163.25ms
step:236/1530 train_loss:4.0880 train_time:36896ms step_avg:163.26ms
step:237/1530 train_loss:4.0881 train_time:37061ms step_avg:163.27ms
step:238/1530 train_loss:3.9514 train_time:37227ms step_avg:163.28ms
step:239/1530 train_loss:4.0952 train_time:37394ms step_avg:163.29ms
step:240/1530 train_loss:4.1229 train_time:37561ms step_avg:163.31ms
step:241/1530 train_loss:3.9711 train_time:37726ms step_avg:163.31ms
step:242/1530 train_loss:4.1477 train_time:37891ms step_avg:163.32ms
step:243/1530 train_loss:4.0198 train_time:38057ms step_avg:163.34ms
step:244/1530 train_loss:4.0919 train_time:38222ms step_avg:163.34ms
step:245/1530 train_loss:4.1479 train_time:38388ms step_avg:163.35ms
step:246/1530 train_loss:4.0651 train_time:38555ms step_avg:163.37ms
step:247/1530 train_loss:4.0073 train_time:38721ms step_avg:163.38ms
step:248/1530 train_loss:4.1066 train_time:38886ms step_avg:163.39ms
step:249/1530 train_loss:3.9335 train_time:39053ms step_avg:163.40ms
step:250/1530 train_loss:3.9801 train_time:39219ms step_avg:163.41ms
step:250/1530 val_loss:4.0077 train_time:39268ms step_avg:163.62ms
step:251/1530 train_loss:4.0798 train_time:39388ms step_avg:163.43ms
step:252/1530 train_loss:4.1737 train_time:39556ms step_avg:163.45ms
step:253/1530 train_loss:3.9391 train_time:39722ms step_avg:163.47ms
step:254/1530 train_loss:3.8832 train_time:39888ms step_avg:163.48ms
step:255/1530 train_loss:4.0868 train_time:40053ms step_avg:163.48ms
step:256/1530 train_loss:3.9991 train_time:40219ms step_avg:163.49ms
step:257/1530 train_loss:3.9987 train_time:40385ms step_avg:163.50ms
step:258/1530 train_loss:3.9961 train_time:40550ms step_avg:163.51ms
step:259/1530 train_loss:4.0429 train_time:40717ms step_avg:163.52ms
step:260/1530 train_loss:4.0716 train_time:40882ms step_avg:163.53ms
step:261/1530 train_loss:4.0281 train_time:41048ms step_avg:163.54ms
step:262/1530 train_loss:3.9950 train_time:41216ms step_avg:163.55ms
step:263/1530 train_loss:3.9019 train_time:41381ms step_avg:163.56ms
step:264/1530 train_loss:3.9961 train_time:41546ms step_avg:163.57ms
step:265/1530 train_loss:3.8707 train_time:41714ms step_avg:163.58ms
step:266/1530 train_loss:3.9258 train_time:41879ms step_avg:163.59ms
step:267/1530 train_loss:3.9396 train_time:42045ms step_avg:163.60ms
step:268/1530 train_loss:3.9609 train_time:42211ms step_avg:163.61ms
step:269/1530 train_loss:3.8610 train_time:42376ms step_avg:163.61ms
step:270/1530 train_loss:4.1037 train_time:42541ms step_avg:163.62ms
step:271/1530 train_loss:3.9814 train_time:42707ms step_avg:163.63ms
step:272/1530 train_loss:3.9382 train_time:42873ms step_avg:163.64ms
step:273/1530 train_loss:3.9453 train_time:43039ms step_avg:163.65ms
step:274/1530 train_loss:4.0510 train_time:43205ms step_avg:163.66ms
step:275/1530 train_loss:4.0645 train_time:43369ms step_avg:163.66ms
step:276/1530 train_loss:4.2491 train_time:43537ms step_avg:163.67ms
step:277/1530 train_loss:4.0503 train_time:43703ms step_avg:163.68ms
step:278/1530 train_loss:4.0902 train_time:43869ms step_avg:163.69ms
step:279/1530 train_loss:4.0075 train_time:44035ms step_avg:163.70ms
step:280/1530 train_loss:4.1972 train_time:44202ms step_avg:163.71ms
step:281/1530 train_loss:3.9747 train_time:44368ms step_avg:163.72ms
step:282/1530 train_loss:3.9536 train_time:44537ms step_avg:163.74ms
step:283/1530 train_loss:3.9240 train_time:44702ms step_avg:163.74ms
step:284/1530 train_loss:4.0558 train_time:44869ms step_avg:163.75ms
step:285/1530 train_loss:4.0597 train_time:45035ms step_avg:163.76ms
step:286/1530 train_loss:4.0941 train_time:45200ms step_avg:163.77ms
step:287/1530 train_loss:3.9060 train_time:45365ms step_avg:163.77ms
step:288/1530 train_loss:4.0129 train_time:45529ms step_avg:163.77ms
step:289/1530 train_loss:3.8747 train_time:45696ms step_avg:163.78ms
step:290/1530 train_loss:3.8675 train_time:45861ms step_avg:163.79ms
step:291/1530 train_loss:3.9148 train_time:46025ms step_avg:163.79ms
step:292/1530 train_loss:3.8695 train_time:46190ms step_avg:163.79ms
step:293/1530 train_loss:3.9125 train_time:46356ms step_avg:163.80ms
step:294/1530 train_loss:3.9456 train_time:46521ms step_avg:163.80ms
step:295/1530 train_loss:3.8457 train_time:46686ms step_avg:163.81ms
step:296/1530 train_loss:3.8706 train_time:46850ms step_avg:163.81ms
step:297/1530 train_loss:3.8702 train_time:47016ms step_avg:163.82ms
step:298/1530 train_loss:3.9740 train_time:47181ms step_avg:163.82ms
step:299/1530 train_loss:3.8279 train_time:47345ms step_avg:163.82ms
step:300/1530 train_loss:3.9807 train_time:47510ms step_avg:163.83ms
step:301/1530 train_loss:3.9724 train_time:47676ms step_avg:163.83ms
step:302/1530 train_loss:3.9431 train_time:47841ms step_avg:163.84ms
step:303/1530 train_loss:3.9881 train_time:48005ms step_avg:163.84ms
step:304/1530 train_loss:3.9776 train_time:48170ms step_avg:163.84ms
step:305/1530 train_loss:4.4647 train_time:48335ms step_avg:163.85ms
step:306/1530 train_loss:3.9403 train_time:48500ms step_avg:163.85ms
step:307/1530 train_loss:3.8428 train_time:48665ms step_avg:163.86ms
step:308/1530 train_loss:3.9851 train_time:48829ms step_avg:163.86ms
step:309/1530 train_loss:3.8788 train_time:48996ms step_avg:163.87ms
step:310/1530 train_loss:4.0988 train_time:49163ms step_avg:163.88ms
step:311/1530 train_loss:3.9409 train_time:49329ms step_avg:163.88ms
step:312/1530 train_loss:3.8680 train_time:49494ms step_avg:163.89ms
step:313/1530 train_loss:3.9373 train_time:49660ms step_avg:163.89ms
step:314/1530 train_loss:4.0707 train_time:49825ms step_avg:163.90ms
step:315/1530 train_loss:3.9572 train_time:49989ms step_avg:163.90ms
step:316/1530 train_loss:3.8017 train_time:50154ms step_avg:163.90ms
step:317/1530 train_loss:3.8844 train_time:50320ms step_avg:163.91ms
step:318/1530 train_loss:3.9270 train_time:50484ms step_avg:163.91ms
step:319/1530 train_loss:3.8926 train_time:50648ms step_avg:163.91ms
step:320/1530 train_loss:4.0166 train_time:50815ms step_avg:163.92ms
step:321/1530 train_loss:3.9599 train_time:50980ms step_avg:163.92ms
step:322/1530 train_loss:3.9333 train_time:51146ms step_avg:163.93ms
step:323/1530 train_loss:4.0156 train_time:51311ms step_avg:163.93ms
step:324/1530 train_loss:3.9588 train_time:51477ms step_avg:163.94ms
step:325/1530 train_loss:4.0239 train_time:51642ms step_avg:163.94ms
step:326/1530 train_loss:3.9046 train_time:51807ms step_avg:163.95ms
step:327/1530 train_loss:4.3946 train_time:51974ms step_avg:163.96ms
step:328/1530 train_loss:4.0803 train_time:52139ms step_avg:163.96ms
step:329/1530 train_loss:3.8002 train_time:52306ms step_avg:163.97ms
step:330/1530 train_loss:3.7598 train_time:52470ms step_avg:163.97ms
step:331/1530 train_loss:3.9837 train_time:52636ms step_avg:163.98ms
step:332/1530 train_loss:3.9184 train_time:52801ms step_avg:163.98ms
step:333/1530 train_loss:3.8938 train_time:52965ms step_avg:163.98ms
step:334/1530 train_loss:3.8532 train_time:53130ms step_avg:163.98ms
step:335/1530 train_loss:4.0175 train_time:53295ms step_avg:163.99ms
step:336/1530 train_loss:3.9657 train_time:53460ms step_avg:163.99ms
step:337/1530 train_loss:4.4350 train_time:53627ms step_avg:164.00ms
step:338/1530 train_loss:3.9404 train_time:53792ms step_avg:164.00ms
step:339/1530 train_loss:3.8739 train_time:53958ms step_avg:164.00ms
step:340/1530 train_loss:3.9397 train_time:54123ms step_avg:164.01ms
step:341/1530 train_loss:3.8559 train_time:54288ms step_avg:164.01ms
step:342/1530 train_loss:3.8131 train_time:54457ms step_avg:164.03ms
step:343/1530 train_loss:3.8458 train_time:54626ms step_avg:164.04ms
step:344/1530 train_loss:4.0091 train_time:54793ms step_avg:164.05ms
step:345/1530 train_loss:3.8244 train_time:54964ms step_avg:164.07ms
step:346/1530 train_loss:3.7744 train_time:55130ms step_avg:164.08ms
step:347/1530 train_loss:3.8097 train_time:55300ms step_avg:164.09ms
step:348/1530 train_loss:3.8640 train_time:55468ms step_avg:164.11ms
step:349/1530 train_loss:3.8304 train_time:55638ms step_avg:164.12ms
step:350/1530 train_loss:3.5736 train_time:55806ms step_avg:164.13ms
step:351/1530 train_loss:3.8289 train_time:55972ms step_avg:164.14ms
step:352/1530 train_loss:4.1969 train_time:56141ms step_avg:164.15ms
step:353/1530 train_loss:3.6734 train_time:56309ms step_avg:164.17ms
step:354/1530 train_loss:3.9305 train_time:56476ms step_avg:164.17ms
step:355/1530 train_loss:3.7913 train_time:56644ms step_avg:164.19ms
step:356/1530 train_loss:3.8845 train_time:56810ms step_avg:164.19ms
step:357/1530 train_loss:3.7683 train_time:56979ms step_avg:164.21ms
step:358/1530 train_loss:3.8698 train_time:57146ms step_avg:164.21ms
step:359/1530 train_loss:3.7833 train_time:57317ms step_avg:164.23ms
step:360/1530 train_loss:3.4359 train_time:57486ms step_avg:164.24ms
step:361/1530 train_loss:4.0212 train_time:57653ms step_avg:164.25ms
step:362/1530 train_loss:3.9216 train_time:57822ms step_avg:164.27ms
step:363/1530 train_loss:3.8446 train_time:57990ms step_avg:164.28ms
step:364/1530 train_loss:3.7502 train_time:58158ms step_avg:164.29ms
step:365/1530 train_loss:3.9201 train_time:58326ms step_avg:164.30ms
step:366/1530 train_loss:3.8640 train_time:58493ms step_avg:164.31ms
step:367/1530 train_loss:3.8569 train_time:58663ms step_avg:164.32ms
step:368/1530 train_loss:3.8566 train_time:58830ms step_avg:164.33ms
step:369/1530 train_loss:3.7507 train_time:58998ms step_avg:164.34ms
step:370/1530 train_loss:3.8739 train_time:59165ms step_avg:164.35ms
step:371/1530 train_loss:3.7373 train_time:59332ms step_avg:164.36ms
step:372/1530 train_loss:3.6987 train_time:59502ms step_avg:164.37ms
step:373/1530 train_loss:3.9219 train_time:59670ms step_avg:164.38ms
step:374/1530 train_loss:3.8363 train_time:59838ms step_avg:164.39ms
step:375/1530 train_loss:3.8077 train_time:60008ms step_avg:164.40ms
step:375/1530 val_loss:3.8286 train_time:60055ms step_avg:164.53ms
step:376/1530 train_loss:3.8719 train_time:60176ms step_avg:164.41ms
step:377/1530 train_loss:3.7918 train_time:60487ms step_avg:164.82ms
step:378/1530 train_loss:3.8585 train_time:60668ms step_avg:164.86ms
step:379/1530 train_loss:3.8758 train_time:60989ms step_avg:165.28ms
step:380/1530 train_loss:3.9533 train_time:61154ms step_avg:165.28ms
step:381/1530 train_loss:3.8431 train_time:61321ms step_avg:165.29ms
step:382/1530 train_loss:3.8032 train_time:61490ms step_avg:165.30ms
step:383/1530 train_loss:3.7967 train_time:61658ms step_avg:165.30ms
step:384/1530 train_loss:3.8813 train_time:61825ms step_avg:165.31ms
step:385/1530 train_loss:3.7983 train_time:61993ms step_avg:165.31ms
step:386/1530 train_loss:3.9009 train_time:62161ms step_avg:165.32ms
step:387/1530 train_loss:4.0598 train_time:62327ms step_avg:165.32ms
step:388/1530 train_loss:3.7947 train_time:62495ms step_avg:165.33ms
step:389/1530 train_loss:3.8038 train_time:62662ms step_avg:165.33ms
step:390/1530 train_loss:3.9012 train_time:62830ms step_avg:165.34ms
step:391/1530 train_loss:3.8214 train_time:62998ms step_avg:165.35ms
step:392/1530 train_loss:3.9297 train_time:63166ms step_avg:165.35ms
step:393/1530 train_loss:3.7687 train_time:63332ms step_avg:165.36ms
step:394/1530 train_loss:3.8881 train_time:63500ms step_avg:165.36ms
step:395/1530 train_loss:3.6388 train_time:63668ms step_avg:165.37ms
step:396/1530 train_loss:3.8473 train_time:63835ms step_avg:165.38ms
step:397/1530 train_loss:3.8654 train_time:64003ms step_avg:165.38ms
step:398/1530 train_loss:3.8929 train_time:64171ms step_avg:165.39ms
step:399/1530 train_loss:3.7726 train_time:64337ms step_avg:165.39ms
step:400/1530 train_loss:3.8321 train_time:64506ms step_avg:165.40ms
step:401/1530 train_loss:3.9130 train_time:64673ms step_avg:165.40ms
step:402/1530 train_loss:3.8434 train_time:64841ms step_avg:165.41ms
step:403/1530 train_loss:3.9645 train_time:65009ms step_avg:165.42ms
step:404/1530 train_loss:3.6820 train_time:65176ms step_avg:165.42ms
step:405/1530 train_loss:3.7893 train_time:65343ms step_avg:165.42ms
step:406/1530 train_loss:4.0999 train_time:65510ms step_avg:165.43ms
step:407/1530 train_loss:3.7821 train_time:65678ms step_avg:165.44ms
step:408/1530 train_loss:3.8257 train_time:65845ms step_avg:165.44ms
step:409/1530 train_loss:3.8642 train_time:66011ms step_avg:165.44ms
step:410/1530 train_loss:3.7589 train_time:66179ms step_avg:165.45ms
step:411/1530 train_loss:3.7661 train_time:66346ms step_avg:165.45ms
step:412/1530 train_loss:4.1867 train_time:66513ms step_avg:165.46ms
step:413/1530 train_loss:3.6260 train_time:66681ms step_avg:165.46ms
step:414/1530 train_loss:4.0197 train_time:66848ms step_avg:165.47ms
step:415/1530 train_loss:3.7589 train_time:67015ms step_avg:165.47ms
step:416/1530 train_loss:3.7706 train_time:67183ms step_avg:165.48ms
step:417/1530 train_loss:3.9670 train_time:67350ms step_avg:165.48ms
step:418/1530 train_loss:3.6951 train_time:67516ms step_avg:165.48ms
step:419/1530 train_loss:3.8068 train_time:67685ms step_avg:165.49ms
step:420/1530 train_loss:3.7008 train_time:67851ms step_avg:165.49ms
step:421/1530 train_loss:3.6505 train_time:68017ms step_avg:165.49ms
step:422/1530 train_loss:3.7846 train_time:68186ms step_avg:165.50ms
step:423/1530 train_loss:3.8793 train_time:68352ms step_avg:165.50ms
step:424/1530 train_loss:3.6158 train_time:68518ms step_avg:165.50ms
step:425/1530 train_loss:3.7997 train_time:68686ms step_avg:165.51ms
step:426/1530 train_loss:3.6493 train_time:68853ms step_avg:165.51ms
step:427/1530 train_loss:3.8925 train_time:69019ms step_avg:165.51ms
step:428/1530 train_loss:3.8122 train_time:69187ms step_avg:165.52ms
step:429/1530 train_loss:3.7609 train_time:69354ms step_avg:165.52ms
step:430/1530 train_loss:3.7105 train_time:69521ms step_avg:165.53ms
step:431/1530 train_loss:3.6402 train_time:69688ms step_avg:165.53ms
step:432/1530 train_loss:3.7658 train_time:69855ms step_avg:165.53ms
step:433/1530 train_loss:3.8193 train_time:70022ms step_avg:165.54ms
step:434/1530 train_loss:3.7795 train_time:70189ms step_avg:165.54ms
step:435/1530 train_loss:3.8102 train_time:70356ms step_avg:165.54ms
step:436/1530 train_loss:3.8277 train_time:70524ms step_avg:165.55ms
step:437/1530 train_loss:3.7198 train_time:70690ms step_avg:165.55ms
step:438/1530 train_loss:3.7035 train_time:70857ms step_avg:165.55ms
step:439/1530 train_loss:3.7087 train_time:71025ms step_avg:165.56ms
step:440/1530 train_loss:3.8919 train_time:71191ms step_avg:165.56ms
step:441/1530 train_loss:3.7630 train_time:71358ms step_avg:165.56ms
step:442/1530 train_loss:3.7425 train_time:71526ms step_avg:165.57ms
step:443/1530 train_loss:3.6264 train_time:71692ms step_avg:165.57ms
step:444/1530 train_loss:3.9296 train_time:71858ms step_avg:165.57ms
step:445/1530 train_loss:3.8490 train_time:72025ms step_avg:165.57ms
step:446/1530 train_loss:3.8405 train_time:72192ms step_avg:165.58ms
step:447/1530 train_loss:3.7564 train_time:72358ms step_avg:165.58ms
step:448/1530 train_loss:3.8599 train_time:72525ms step_avg:165.58ms
step:449/1530 train_loss:3.6990 train_time:72692ms step_avg:165.59ms
step:450/1530 train_loss:3.7204 train_time:72859ms step_avg:165.59ms
step:451/1530 train_loss:3.5811 train_time:73026ms step_avg:165.59ms
step:452/1530 train_loss:3.7083 train_time:73193ms step_avg:165.59ms
step:453/1530 train_loss:3.6700 train_time:73361ms step_avg:165.60ms
step:454/1530 train_loss:3.6394 train_time:73528ms step_avg:165.60ms
step:455/1530 train_loss:3.8407 train_time:73696ms step_avg:165.61ms
step:456/1530 train_loss:3.7286 train_time:73866ms step_avg:165.62ms
step:457/1530 train_loss:3.7798 train_time:74034ms step_avg:165.63ms
step:458/1530 train_loss:3.8352 train_time:74204ms step_avg:165.63ms
step:459/1530 train_loss:3.6325 train_time:74374ms step_avg:165.64ms
step:460/1530 train_loss:3.7912 train_time:74543ms step_avg:165.65ms
step:461/1530 train_loss:3.6936 train_time:74714ms step_avg:165.66ms
step:462/1530 train_loss:3.7396 train_time:74884ms step_avg:165.67ms
step:463/1530 train_loss:3.7782 train_time:75053ms step_avg:165.68ms
step:464/1530 train_loss:3.7154 train_time:75222ms step_avg:165.69ms
step:465/1530 train_loss:3.7165 train_time:75390ms step_avg:165.69ms
step:466/1530 train_loss:3.8036 train_time:75559ms step_avg:165.70ms
step:467/1530 train_loss:3.8250 train_time:75730ms step_avg:165.71ms
step:468/1530 train_loss:3.7913 train_time:75898ms step_avg:165.72ms
step:469/1530 train_loss:3.6917 train_time:76067ms step_avg:165.72ms
step:470/1530 train_loss:3.7757 train_time:76237ms step_avg:165.73ms
step:471/1530 train_loss:3.8116 train_time:76408ms step_avg:165.74ms
step:472/1530 train_loss:3.7914 train_time:76579ms step_avg:165.75ms
step:473/1530 train_loss:3.7148 train_time:76748ms step_avg:165.76ms
step:474/1530 train_loss:3.5974 train_time:76916ms step_avg:165.77ms
step:475/1530 train_loss:4.0222 train_time:77088ms step_avg:165.78ms
step:476/1530 train_loss:3.7585 train_time:77256ms step_avg:165.79ms
step:477/1530 train_loss:3.5904 train_time:77427ms step_avg:165.80ms
step:478/1530 train_loss:3.8279 train_time:77596ms step_avg:165.80ms
step:479/1530 train_loss:3.7773 train_time:77768ms step_avg:165.82ms
step:480/1530 train_loss:3.9204 train_time:77938ms step_avg:165.83ms
step:481/1530 train_loss:3.7298 train_time:78107ms step_avg:165.83ms
step:482/1530 train_loss:3.5286 train_time:78277ms step_avg:165.84ms
step:483/1530 train_loss:3.8062 train_time:78446ms step_avg:165.85ms
step:484/1530 train_loss:3.6645 train_time:78616ms step_avg:165.86ms
step:485/1530 train_loss:3.6604 train_time:78788ms step_avg:165.87ms
step:486/1530 train_loss:3.5720 train_time:78957ms step_avg:165.88ms
step:487/1530 train_loss:3.6876 train_time:79127ms step_avg:165.88ms
step:488/1530 train_loss:3.8782 train_time:79296ms step_avg:165.89ms
step:489/1530 train_loss:3.7104 train_time:79467ms step_avg:165.90ms
step:490/1530 train_loss:3.5918 train_time:79635ms step_avg:165.91ms
step:491/1530 train_loss:3.6200 train_time:79806ms step_avg:165.92ms
step:492/1530 train_loss:3.7376 train_time:79976ms step_avg:165.93ms
step:493/1530 train_loss:3.5786 train_time:80147ms step_avg:165.94ms
step:494/1530 train_loss:3.7015 train_time:80316ms step_avg:165.94ms
step:495/1530 train_loss:3.6584 train_time:80488ms step_avg:165.95ms
step:496/1530 train_loss:3.5094 train_time:80660ms step_avg:165.97ms
step:497/1530 train_loss:3.7353 train_time:80829ms step_avg:165.97ms
step:498/1530 train_loss:3.7895 train_time:80998ms step_avg:165.98ms
step:499/1530 train_loss:3.8309 train_time:81169ms step_avg:165.99ms
step:500/1530 train_loss:3.7373 train_time:81339ms step_avg:166.00ms
step:500/1530 val_loss:3.7038 train_time:81388ms step_avg:166.10ms
step:501/1530 train_loss:3.8027 train_time:81510ms step_avg:166.01ms
step:502/1530 train_loss:3.7513 train_time:81682ms step_avg:166.02ms
step:503/1530 train_loss:3.7813 train_time:81852ms step_avg:166.03ms
step:504/1530 train_loss:3.7210 train_time:82020ms step_avg:166.03ms
step:505/1530 train_loss:3.8128 train_time:82190ms step_avg:166.04ms
step:506/1530 train_loss:3.6494 train_time:82360ms step_avg:166.05ms
step:507/1530 train_loss:3.7694 train_time:82529ms step_avg:166.06ms
step:508/1530 train_loss:3.8250 train_time:82702ms step_avg:166.07ms
step:509/1530 train_loss:3.7761 train_time:82870ms step_avg:166.07ms
step:510/1530 train_loss:3.5877 train_time:83040ms step_avg:166.08ms
step:511/1530 train_loss:3.7752 train_time:83211ms step_avg:166.09ms
step:512/1530 train_loss:3.7223 train_time:83382ms step_avg:166.10ms
step:513/1530 train_loss:3.6661 train_time:83551ms step_avg:166.11ms
step:514/1530 train_loss:3.8063 train_time:83723ms step_avg:166.12ms
step:515/1530 train_loss:3.7388 train_time:83891ms step_avg:166.12ms
step:516/1530 train_loss:4.0749 train_time:84063ms step_avg:166.13ms
step:517/1530 train_loss:3.6896 train_time:84232ms step_avg:166.14ms
step:518/1530 train_loss:3.7710 train_time:84400ms step_avg:166.14ms
step:519/1530 train_loss:3.6553 train_time:84569ms step_avg:166.15ms
step:520/1530 train_loss:3.6822 train_time:84737ms step_avg:166.15ms
step:521/1530 train_loss:3.6641 train_time:84907ms step_avg:166.16ms
step:522/1530 train_loss:3.6559 train_time:85077ms step_avg:166.17ms
step:523/1530 train_loss:4.2969 train_time:85247ms step_avg:166.17ms
step:524/1530 train_loss:3.7395 train_time:85415ms step_avg:166.18ms
step:525/1530 train_loss:3.6828 train_time:85583ms step_avg:166.18ms
step:526/1530 train_loss:3.7024 train_time:85753ms step_avg:166.19ms
step:527/1530 train_loss:3.6575 train_time:85922ms step_avg:166.19ms
step:528/1530 train_loss:3.6346 train_time:86090ms step_avg:166.20ms
step:529/1530 train_loss:3.8560 train_time:86260ms step_avg:166.20ms
step:530/1530 train_loss:3.6459 train_time:86430ms step_avg:166.21ms
step:531/1530 train_loss:3.9215 train_time:86600ms step_avg:166.22ms
step:532/1530 train_loss:3.7359 train_time:86769ms step_avg:166.22ms
step:533/1530 train_loss:3.6571 train_time:86938ms step_avg:166.23ms
step:534/1530 train_loss:3.6702 train_time:87108ms step_avg:166.24ms
step:535/1530 train_loss:3.6091 train_time:87277ms step_avg:166.24ms
step:536/1530 train_loss:3.7441 train_time:87448ms step_avg:166.25ms
step:537/1530 train_loss:3.7224 train_time:87619ms step_avg:166.26ms
step:538/1530 train_loss:3.6260 train_time:87789ms step_avg:166.27ms
step:539/1530 train_loss:4.1162 train_time:87959ms step_avg:166.27ms
step:540/1530 train_loss:3.6721 train_time:88129ms step_avg:166.28ms
step:541/1530 train_loss:3.7879 train_time:88297ms step_avg:166.29ms
step:542/1530 train_loss:3.5898 train_time:88466ms step_avg:166.29ms
step:543/1530 train_loss:3.5877 train_time:88634ms step_avg:166.29ms
step:544/1530 train_loss:3.6374 train_time:88805ms step_avg:166.30ms
step:545/1530 train_loss:3.5959 train_time:88973ms step_avg:166.30ms
step:546/1530 train_loss:3.6274 train_time:89142ms step_avg:166.31ms
step:547/1530 train_loss:3.6426 train_time:89311ms step_avg:166.31ms
step:548/1530 train_loss:3.6135 train_time:89479ms step_avg:166.32ms
step:549/1530 train_loss:3.7265 train_time:89648ms step_avg:166.32ms
step:550/1530 train_loss:3.6175 train_time:89818ms step_avg:166.33ms
step:551/1530 train_loss:3.6294 train_time:89986ms step_avg:166.33ms
step:552/1530 train_loss:3.9400 train_time:90155ms step_avg:166.34ms
step:553/1530 train_loss:3.7590 train_time:90325ms step_avg:166.34ms
step:554/1530 train_loss:3.7152 train_time:90493ms step_avg:166.35ms
step:555/1530 train_loss:3.6251 train_time:90662ms step_avg:166.35ms
step:556/1530 train_loss:3.6966 train_time:90830ms step_avg:166.35ms
step:557/1530 train_loss:3.3179 train_time:90999ms step_avg:166.36ms
step:558/1530 train_loss:3.6115 train_time:91168ms step_avg:166.37ms
step:559/1530 train_loss:3.6500 train_time:91336ms step_avg:166.37ms
step:560/1530 train_loss:3.6909 train_time:91507ms step_avg:166.38ms
step:561/1530 train_loss:3.6125 train_time:91675ms step_avg:166.38ms
step:562/1530 train_loss:3.5578 train_time:91843ms step_avg:166.38ms
step:563/1530 train_loss:3.7585 train_time:92013ms step_avg:166.39ms
step:564/1530 train_loss:3.5726 train_time:92185ms step_avg:166.40ms
step:565/1530 train_loss:3.6829 train_time:92354ms step_avg:166.40ms
step:566/1530 train_loss:3.6219 train_time:92664ms step_avg:166.66ms
step:567/1530 train_loss:3.5950 train_time:92843ms step_avg:166.68ms
step:568/1530 train_loss:3.6831 train_time:93013ms step_avg:166.69ms
step:569/1530 train_loss:3.6525 train_time:93342ms step_avg:166.98ms
step:570/1530 train_loss:3.6967 train_time:93512ms step_avg:166.99ms
step:571/1530 train_loss:3.7632 train_time:93682ms step_avg:166.99ms
step:572/1530 train_loss:3.7194 train_time:93853ms step_avg:167.00ms
step:573/1530 train_loss:3.7368 train_time:94029ms step_avg:167.01ms
step:574/1530 train_loss:3.7803 train_time:94200ms step_avg:167.02ms
step:575/1530 train_loss:3.7338 train_time:94371ms step_avg:167.03ms
step:576/1530 train_loss:3.7630 train_time:94541ms step_avg:167.03ms
step:577/1530 train_loss:3.6688 train_time:94714ms step_avg:167.04ms
step:578/1530 train_loss:3.6770 train_time:94887ms step_avg:167.05ms
step:579/1530 train_loss:3.6705 train_time:95057ms step_avg:167.06ms
step:580/1530 train_loss:3.5934 train_time:95229ms step_avg:167.07ms
step:581/1530 train_loss:3.6430 train_time:95399ms step_avg:167.07ms
step:582/1530 train_loss:3.8566 train_time:95570ms step_avg:167.08ms
step:583/1530 train_loss:3.6313 train_time:95740ms step_avg:167.09ms
step:584/1530 train_loss:3.5922 train_time:95913ms step_avg:167.10ms
step:585/1530 train_loss:3.7890 train_time:96082ms step_avg:167.10ms
step:586/1530 train_loss:3.5115 train_time:96254ms step_avg:167.11ms
step:587/1530 train_loss:3.6721 train_time:96427ms step_avg:167.12ms
step:588/1530 train_loss:3.6411 train_time:96596ms step_avg:167.12ms
step:589/1530 train_loss:3.9988 train_time:96768ms step_avg:167.13ms
step:590/1530 train_loss:3.7773 train_time:96939ms step_avg:167.14ms
step:591/1530 train_loss:3.5075 train_time:97111ms step_avg:167.14ms
step:592/1530 train_loss:3.5305 train_time:97283ms step_avg:167.15ms
step:593/1530 train_loss:3.5011 train_time:97455ms step_avg:167.16ms
step:594/1530 train_loss:3.5540 train_time:97630ms step_avg:167.17ms
step:595/1530 train_loss:3.9071 train_time:97804ms step_avg:167.19ms
step:596/1530 train_loss:3.6512 train_time:97976ms step_avg:167.19ms
step:597/1530 train_loss:3.5836 train_time:98147ms step_avg:167.20ms
step:598/1530 train_loss:3.6596 train_time:98318ms step_avg:167.21ms
step:599/1530 train_loss:3.4760 train_time:98488ms step_avg:167.21ms
step:600/1530 train_loss:3.5968 train_time:98659ms step_avg:167.22ms
step:601/1530 train_loss:3.6468 train_time:98833ms step_avg:167.23ms
step:602/1530 train_loss:3.6705 train_time:99006ms step_avg:167.24ms
step:603/1530 train_loss:3.7792 train_time:99176ms step_avg:167.24ms
step:604/1530 train_loss:3.6066 train_time:99348ms step_avg:167.25ms
step:605/1530 train_loss:3.6105 train_time:99521ms step_avg:167.26ms
step:606/1530 train_loss:3.5746 train_time:99694ms step_avg:167.27ms
step:607/1530 train_loss:3.8427 train_time:99867ms step_avg:167.28ms
step:608/1530 train_loss:3.6347 train_time:100038ms step_avg:167.29ms
step:609/1530 train_loss:3.6173 train_time:100210ms step_avg:167.30ms
step:610/1530 train_loss:3.7011 train_time:100379ms step_avg:167.30ms
step:611/1530 train_loss:3.5973 train_time:100551ms step_avg:167.31ms
step:612/1530 train_loss:3.5709 train_time:100723ms step_avg:167.31ms
step:613/1530 train_loss:3.7641 train_time:100894ms step_avg:167.32ms
step:614/1530 train_loss:3.7045 train_time:101065ms step_avg:167.33ms
step:615/1530 train_loss:3.7060 train_time:101236ms step_avg:167.33ms
step:616/1530 train_loss:3.6292 train_time:101407ms step_avg:167.34ms
step:617/1530 train_loss:3.5619 train_time:101581ms step_avg:167.35ms
step:618/1530 train_loss:3.6916 train_time:101751ms step_avg:167.35ms
step:619/1530 train_loss:3.5519 train_time:101923ms step_avg:167.36ms
step:620/1530 train_loss:3.5911 train_time:102093ms step_avg:167.37ms
step:621/1530 train_loss:3.9258 train_time:102267ms step_avg:167.38ms
step:622/1530 train_loss:3.5728 train_time:102440ms step_avg:167.39ms
step:623/1530 train_loss:3.6057 train_time:102613ms step_avg:167.39ms
step:624/1530 train_loss:3.6966 train_time:102783ms step_avg:167.40ms
step:625/1530 train_loss:3.7065 train_time:102953ms step_avg:167.40ms
step:625/1530 val_loss:3.6229 train_time:103003ms step_avg:167.48ms
step:626/1530 train_loss:3.7376 train_time:103125ms step_avg:167.41ms
step:627/1530 train_loss:3.7072 train_time:103296ms step_avg:167.42ms
step:628/1530 train_loss:3.7625 train_time:103466ms step_avg:167.42ms
step:629/1530 train_loss:3.5908 train_time:103638ms step_avg:167.43ms
step:630/1530 train_loss:3.7253 train_time:103808ms step_avg:167.43ms
step:631/1530 train_loss:3.7441 train_time:103979ms step_avg:167.44ms
step:632/1530 train_loss:3.6484 train_time:104152ms step_avg:167.45ms
step:633/1530 train_loss:3.6058 train_time:104324ms step_avg:167.45ms
step:634/1530 train_loss:3.6964 train_time:104495ms step_avg:167.46ms
step:635/1530 train_loss:3.9483 train_time:104664ms step_avg:167.46ms
step:636/1530 train_loss:3.5477 train_time:104835ms step_avg:167.47ms
step:637/1530 train_loss:3.3477 train_time:105004ms step_avg:167.47ms
step:638/1530 train_loss:3.5923 train_time:105174ms step_avg:167.47ms
step:639/1530 train_loss:3.6303 train_time:105344ms step_avg:167.48ms
step:640/1530 train_loss:3.5689 train_time:105515ms step_avg:167.48ms
step:641/1530 train_loss:3.5878 train_time:105685ms step_avg:167.49ms
step:642/1530 train_loss:3.6357 train_time:105856ms step_avg:167.49ms
step:643/1530 train_loss:3.5923 train_time:106026ms step_avg:167.50ms
step:644/1530 train_loss:3.5601 train_time:106196ms step_avg:167.50ms
step:645/1530 train_loss:3.7758 train_time:106366ms step_avg:167.51ms
step:646/1530 train_loss:3.6713 train_time:106539ms step_avg:167.51ms
step:647/1530 train_loss:3.6631 train_time:106710ms step_avg:167.52ms
step:648/1530 train_loss:3.7124 train_time:106882ms step_avg:167.53ms
step:649/1530 train_loss:3.7676 train_time:107052ms step_avg:167.53ms
step:650/1530 train_loss:3.6245 train_time:107223ms step_avg:167.54ms
step:651/1530 train_loss:3.7579 train_time:107395ms step_avg:167.54ms
step:652/1530 train_loss:3.5901 train_time:107564ms step_avg:167.55ms
step:653/1530 train_loss:3.6612 train_time:107737ms step_avg:167.55ms
step:654/1530 train_loss:3.4270 train_time:107906ms step_avg:167.56ms
step:655/1530 train_loss:3.5827 train_time:108076ms step_avg:167.56ms
step:656/1530 train_loss:3.5780 train_time:108247ms step_avg:167.56ms
step:657/1530 train_loss:3.4988 train_time:108418ms step_avg:167.57ms
step:658/1530 train_loss:3.6908 train_time:108589ms step_avg:167.58ms
step:659/1530 train_loss:3.5845 train_time:108759ms step_avg:167.58ms
step:660/1530 train_loss:3.6874 train_time:108930ms step_avg:167.58ms
step:661/1530 train_loss:3.7518 train_time:109100ms step_avg:167.59ms
step:662/1530 train_loss:3.6699 train_time:109270ms step_avg:167.59ms
step:663/1530 train_loss:3.5500 train_time:109441ms step_avg:167.60ms
step:664/1530 train_loss:3.6109 train_time:109612ms step_avg:167.60ms
step:665/1530 train_loss:3.4945 train_time:109784ms step_avg:167.61ms
step:666/1530 train_loss:3.7815 train_time:109955ms step_avg:167.61ms
step:667/1530 train_loss:3.6063 train_time:110126ms step_avg:167.62ms
step:668/1530 train_loss:3.6478 train_time:110296ms step_avg:167.62ms
step:669/1530 train_loss:3.4837 train_time:110466ms step_avg:167.63ms
step:670/1530 train_loss:3.5987 train_time:110637ms step_avg:167.63ms
step:671/1530 train_loss:3.5591 train_time:110807ms step_avg:167.64ms
step:672/1530 train_loss:3.5710 train_time:110979ms step_avg:167.64ms
step:673/1530 train_loss:3.8512 train_time:111149ms step_avg:167.65ms
step:674/1530 train_loss:3.6203 train_time:111320ms step_avg:167.65ms
step:675/1530 train_loss:3.7094 train_time:111489ms step_avg:167.65ms
step:676/1530 train_loss:3.4934 train_time:111661ms step_avg:167.66ms
step:677/1530 train_loss:3.6017 train_time:111833ms step_avg:167.67ms
step:678/1530 train_loss:3.5559 train_time:112004ms step_avg:167.67ms
step:679/1530 train_loss:3.6834 train_time:112175ms step_avg:167.68ms
step:680/1530 train_loss:3.5858 train_time:112346ms step_avg:167.68ms
step:681/1530 train_loss:3.6125 train_time:112517ms step_avg:167.69ms
step:682/1530 train_loss:3.6573 train_time:112693ms step_avg:167.70ms
step:683/1530 train_loss:3.7374 train_time:112865ms step_avg:167.70ms
step:684/1530 train_loss:3.6491 train_time:113038ms step_avg:167.71ms
step:685/1530 train_loss:3.6895 train_time:113211ms step_avg:167.72ms
step:686/1530 train_loss:3.6375 train_time:113383ms step_avg:167.73ms
step:687/1530 train_loss:3.6652 train_time:113556ms step_avg:167.73ms
step:688/1530 train_loss:3.2117 train_time:113731ms step_avg:167.75ms
step:689/1530 train_loss:3.4054 train_time:113904ms step_avg:167.75ms
step:690/1530 train_loss:3.5367 train_time:114080ms step_avg:167.76ms
step:691/1530 train_loss:3.4131 train_time:114252ms step_avg:167.77ms
step:692/1530 train_loss:3.6277 train_time:114424ms step_avg:167.78ms
step:693/1530 train_loss:3.6507 train_time:114598ms step_avg:167.79ms
step:694/1530 train_loss:3.5499 train_time:114770ms step_avg:167.79ms
step:695/1530 train_loss:3.5344 train_time:114941ms step_avg:167.80ms
step:696/1530 train_loss:3.8577 train_time:115115ms step_avg:167.81ms
step:697/1530 train_loss:3.5870 train_time:115288ms step_avg:167.81ms
step:698/1530 train_loss:3.6435 train_time:115460ms step_avg:167.82ms
step:699/1530 train_loss:3.7707 train_time:115635ms step_avg:167.83ms
step:700/1530 train_loss:3.5699 train_time:115807ms step_avg:167.84ms
step:701/1530 train_loss:3.5455 train_time:115979ms step_avg:167.84ms
step:702/1530 train_loss:3.5112 train_time:116154ms step_avg:167.85ms
step:703/1530 train_loss:3.4968 train_time:116327ms step_avg:167.86ms
step:704/1530 train_loss:3.5726 train_time:116500ms step_avg:167.87ms
step:705/1530 train_loss:3.5597 train_time:116677ms step_avg:167.88ms
step:706/1530 train_loss:3.5794 train_time:116854ms step_avg:167.89ms
step:707/1530 train_loss:3.6463 train_time:117028ms step_avg:167.90ms
step:708/1530 train_loss:3.6017 train_time:117200ms step_avg:167.91ms
step:709/1530 train_loss:3.5804 train_time:117374ms step_avg:167.92ms
step:710/1530 train_loss:3.5410 train_time:117545ms step_avg:167.92ms
step:711/1530 train_loss:3.5915 train_time:117719ms step_avg:167.93ms
step:712/1530 train_loss:3.6474 train_time:117895ms step_avg:167.94ms
step:713/1530 train_loss:3.6546 train_time:118069ms step_avg:167.95ms
step:714/1530 train_loss:3.5637 train_time:118241ms step_avg:167.96ms
step:715/1530 train_loss:3.5685 train_time:118414ms step_avg:167.96ms
step:716/1530 train_loss:3.5836 train_time:118586ms step_avg:167.97ms
step:717/1530 train_loss:3.7028 train_time:118760ms step_avg:167.98ms
step:718/1530 train_loss:3.5996 train_time:118933ms step_avg:167.98ms
step:719/1530 train_loss:3.6782 train_time:119104ms step_avg:167.99ms
step:720/1530 train_loss:3.8532 train_time:119280ms step_avg:168.00ms
step:721/1530 train_loss:3.4643 train_time:119454ms step_avg:168.01ms
step:722/1530 train_loss:3.7397 train_time:119625ms step_avg:168.01ms
step:723/1530 train_loss:3.7634 train_time:119797ms step_avg:168.02ms
step:724/1530 train_loss:3.5693 train_time:119970ms step_avg:168.02ms
step:725/1530 train_loss:3.6561 train_time:120143ms step_avg:168.03ms
step:726/1530 train_loss:3.5314 train_time:120316ms step_avg:168.04ms
step:727/1530 train_loss:3.5794 train_time:120491ms step_avg:168.05ms
step:728/1530 train_loss:3.7297 train_time:120664ms step_avg:168.06ms
step:729/1530 train_loss:3.6707 train_time:120840ms step_avg:168.07ms
step:730/1530 train_loss:3.6649 train_time:121012ms step_avg:168.07ms
step:731/1530 train_loss:3.5607 train_time:121184ms step_avg:168.08ms
step:732/1530 train_loss:3.5876 train_time:121357ms step_avg:168.08ms
step:733/1530 train_loss:3.8340 train_time:121529ms step_avg:168.09ms
step:734/1530 train_loss:3.5631 train_time:121701ms step_avg:168.10ms
step:735/1530 train_loss:3.6222 train_time:121874ms step_avg:168.10ms
step:736/1530 train_loss:3.7382 train_time:122046ms step_avg:168.11ms
step:737/1530 train_loss:3.6781 train_time:122219ms step_avg:168.11ms
step:738/1530 train_loss:3.6029 train_time:122391ms step_avg:168.12ms
step:739/1530 train_loss:3.5030 train_time:122563ms step_avg:168.12ms
step:740/1530 train_loss:4.1117 train_time:122742ms step_avg:168.14ms
step:741/1530 train_loss:3.4882 train_time:122915ms step_avg:168.15ms
step:742/1530 train_loss:3.5497 train_time:123088ms step_avg:168.15ms
step:743/1530 train_loss:3.5810 train_time:123260ms step_avg:168.16ms
step:744/1530 train_loss:3.6512 train_time:123433ms step_avg:168.17ms
step:745/1530 train_loss:3.5771 train_time:123607ms step_avg:168.17ms
step:746/1530 train_loss:3.5889 train_time:123779ms step_avg:168.18ms
step:747/1530 train_loss:3.6447 train_time:123953ms step_avg:168.19ms
step:748/1530 train_loss:3.5620 train_time:124129ms step_avg:168.20ms
step:749/1530 train_loss:3.5664 train_time:124302ms step_avg:168.20ms
step:750/1530 train_loss:3.5986 train_time:124472ms step_avg:168.21ms
step:750/1530 val_loss:3.5648 train_time:124522ms step_avg:168.27ms
step:751/1530 train_loss:3.5683 train_time:124645ms step_avg:168.21ms
step:752/1530 train_loss:3.6126 train_time:124817ms step_avg:168.22ms
step:753/1530 train_loss:3.6206 train_time:124990ms step_avg:168.22ms
step:754/1530 train_loss:3.5908 train_time:125163ms step_avg:168.23ms
step:755/1530 train_loss:3.6850 train_time:125478ms step_avg:168.43ms
step:756/1530 train_loss:3.4573 train_time:125664ms step_avg:168.45ms
step:757/1530 train_loss:3.7277 train_time:125837ms step_avg:168.46ms
step:758/1530 train_loss:3.6540 train_time:126009ms step_avg:168.46ms
step:759/1530 train_loss:3.5911 train_time:126338ms step_avg:168.68ms
step:760/1530 train_loss:3.7066 train_time:126508ms step_avg:168.68ms
step:761/1530 train_loss:3.3950 train_time:126682ms step_avg:168.68ms
step:762/1530 train_loss:3.5444 train_time:126853ms step_avg:168.69ms
step:763/1530 train_loss:3.6549 train_time:127025ms step_avg:168.69ms
step:764/1530 train_loss:3.3182 train_time:127198ms step_avg:168.70ms
step:765/1530 train_loss:3.7330 train_time:127370ms step_avg:168.70ms
step:766/1530 train_loss:3.5668 train_time:127543ms step_avg:168.71ms
step:767/1530 train_loss:3.5647 train_time:127713ms step_avg:168.71ms
step:768/1530 train_loss:3.5764 train_time:127888ms step_avg:168.72ms
step:769/1530 train_loss:3.5852 train_time:128063ms step_avg:168.73ms
step:770/1530 train_loss:3.6408 train_time:128235ms step_avg:168.73ms
step:771/1530 train_loss:3.8880 train_time:128407ms step_avg:168.73ms
step:772/1530 train_loss:3.4484 train_time:128579ms step_avg:168.74ms
step:773/1530 train_loss:3.6293 train_time:128750ms step_avg:168.74ms
step:774/1530 train_loss:3.6409 train_time:128922ms step_avg:168.75ms
step:775/1530 train_loss:3.6084 train_time:129093ms step_avg:168.75ms
step:776/1530 train_loss:3.3993 train_time:129266ms step_avg:168.76ms
step:777/1530 train_loss:3.3859 train_time:129441ms step_avg:168.76ms
step:778/1530 train_loss:3.4950 train_time:129613ms step_avg:168.77ms
step:779/1530 train_loss:3.5762 train_time:129784ms step_avg:168.77ms
step:780/1530 train_loss:3.5852 train_time:129959ms step_avg:168.78ms
step:781/1530 train_loss:3.6734 train_time:130131ms step_avg:168.78ms
step:782/1530 train_loss:3.5873 train_time:130303ms step_avg:168.79ms
step:783/1530 train_loss:3.5686 train_time:130474ms step_avg:168.79ms
step:784/1530 train_loss:3.6076 train_time:130645ms step_avg:168.79ms
step:785/1530 train_loss:3.5601 train_time:130817ms step_avg:168.80ms
step:786/1530 train_loss:3.4329 train_time:130989ms step_avg:168.80ms
step:787/1530 train_loss:3.7201 train_time:131162ms step_avg:168.81ms
step:788/1530 train_loss:3.5030 train_time:131335ms step_avg:168.81ms
step:789/1530 train_loss:3.5412 train_time:131507ms step_avg:168.81ms
step:790/1530 train_loss:3.6243 train_time:131681ms step_avg:168.82ms
step:791/1530 train_loss:3.7727 train_time:131857ms step_avg:168.83ms
step:792/1530 train_loss:3.7619 train_time:132029ms step_avg:168.83ms
step:793/1530 train_loss:3.4469 train_time:132201ms step_avg:168.84ms
step:794/1530 train_loss:3.5946 train_time:132376ms step_avg:168.85ms
step:795/1530 train_loss:3.6706 train_time:132550ms step_avg:168.85ms
step:796/1530 train_loss:3.7659 train_time:132726ms step_avg:168.86ms
step:797/1530 train_loss:3.5235 train_time:132900ms step_avg:168.87ms
step:798/1530 train_loss:3.6444 train_time:133077ms step_avg:168.88ms
step:799/1530 train_loss:3.5357 train_time:133256ms step_avg:168.89ms
step:800/1530 train_loss:3.5280 train_time:133429ms step_avg:168.90ms
step:801/1530 train_loss:3.6315 train_time:133602ms step_avg:168.90ms
step:802/1530 train_loss:3.4944 train_time:133781ms step_avg:168.92ms
step:803/1530 train_loss:3.4863 train_time:133954ms step_avg:168.92ms
step:804/1530 train_loss:3.6257 train_time:134127ms step_avg:168.93ms
step:805/1530 train_loss:3.5191 train_time:134303ms step_avg:168.93ms
step:806/1530 train_loss:3.5606 train_time:134477ms step_avg:168.94ms
step:807/1530 train_loss:3.6396 train_time:134649ms step_avg:168.95ms
step:808/1530 train_loss:3.5454 train_time:134824ms step_avg:168.95ms
step:809/1530 train_loss:3.4926 train_time:134997ms step_avg:168.96ms
step:810/1530 train_loss:3.5628 train_time:135169ms step_avg:168.96ms
step:811/1530 train_loss:3.5755 train_time:135343ms step_avg:168.97ms
step:812/1530 train_loss:3.5979 train_time:135517ms step_avg:168.97ms
step:813/1530 train_loss:3.6243 train_time:135688ms step_avg:168.98ms
step:814/1530 train_loss:3.5657 train_time:135863ms step_avg:168.98ms
step:815/1530 train_loss:3.5625 train_time:136039ms step_avg:168.99ms
step:816/1530 train_loss:3.6824 train_time:136214ms step_avg:169.00ms
step:817/1530 train_loss:3.7669 train_time:136387ms step_avg:169.00ms
step:818/1530 train_loss:3.5219 train_time:136560ms step_avg:169.01ms
step:819/1530 train_loss:3.7210 train_time:136735ms step_avg:169.02ms
step:820/1530 train_loss:3.4940 train_time:136911ms step_avg:169.03ms
step:821/1530 train_loss:3.5621 train_time:137083ms step_avg:169.03ms
step:822/1530 train_loss:3.6994 train_time:137260ms step_avg:169.04ms
step:823/1530 train_loss:3.5771 train_time:137434ms step_avg:169.05ms
step:824/1530 train_loss:3.5113 train_time:137606ms step_avg:169.05ms
step:825/1530 train_loss:3.6089 train_time:137783ms step_avg:169.06ms
step:826/1530 train_loss:3.4811 train_time:137960ms step_avg:169.07ms
step:827/1530 train_loss:3.7335 train_time:138134ms step_avg:169.07ms
step:828/1530 train_loss:3.6191 train_time:138307ms step_avg:169.08ms
step:829/1530 train_loss:3.6302 train_time:138484ms step_avg:169.09ms
step:830/1530 train_loss:3.5370 train_time:138660ms step_avg:169.10ms
step:831/1530 train_loss:3.5992 train_time:138832ms step_avg:169.10ms
step:832/1530 train_loss:3.5160 train_time:139007ms step_avg:169.11ms
step:833/1530 train_loss:3.6524 train_time:139182ms step_avg:169.12ms
step:834/1530 train_loss:3.4722 train_time:139356ms step_avg:169.12ms
step:835/1530 train_loss:3.4582 train_time:139529ms step_avg:169.13ms
step:836/1530 train_loss:3.7171 train_time:139705ms step_avg:169.13ms
step:837/1530 train_loss:3.4005 train_time:139881ms step_avg:169.14ms
step:838/1530 train_loss:3.5985 train_time:140055ms step_avg:169.15ms
step:839/1530 train_loss:3.4255 train_time:140229ms step_avg:169.15ms
step:840/1530 train_loss:3.4665 train_time:140402ms step_avg:169.16ms
step:841/1530 train_loss:3.5675 train_time:140575ms step_avg:169.16ms
step:842/1530 train_loss:3.5810 train_time:140749ms step_avg:169.17ms
step:843/1530 train_loss:3.5666 train_time:140921ms step_avg:169.17ms
step:844/1530 train_loss:3.4299 train_time:141094ms step_avg:169.18ms
step:845/1530 train_loss:3.6644 train_time:141268ms step_avg:169.18ms
step:846/1530 train_loss:3.5179 train_time:141442ms step_avg:169.19ms
step:847/1530 train_loss:3.4928 train_time:141618ms step_avg:169.20ms
step:848/1530 train_loss:3.6346 train_time:141789ms step_avg:169.20ms
step:849/1530 train_loss:3.4890 train_time:141963ms step_avg:169.20ms
step:850/1530 train_loss:3.4374 train_time:142137ms step_avg:169.21ms
step:851/1530 train_loss:3.7366 train_time:142312ms step_avg:169.22ms
step:852/1530 train_loss:3.4393 train_time:142485ms step_avg:169.22ms
step:853/1530 train_loss:3.5687 train_time:142657ms step_avg:169.23ms
step:854/1530 train_loss:3.6560 train_time:142832ms step_avg:169.23ms
step:855/1530 train_loss:3.5163 train_time:143006ms step_avg:169.24ms
step:856/1530 train_loss:3.5429 train_time:143180ms step_avg:169.24ms
step:857/1530 train_loss:3.6048 train_time:143354ms step_avg:169.25ms
step:858/1530 train_loss:3.4686 train_time:143529ms step_avg:169.26ms
step:859/1530 train_loss:3.5639 train_time:143703ms step_avg:169.26ms
step:860/1530 train_loss:3.5856 train_time:143875ms step_avg:169.27ms
step:861/1530 train_loss:3.6311 train_time:144054ms step_avg:169.28ms
step:862/1530 train_loss:3.6028 train_time:144231ms step_avg:169.29ms
step:863/1530 train_loss:3.5672 train_time:144407ms step_avg:169.29ms
step:864/1530 train_loss:3.3832 train_time:144582ms step_avg:169.30ms
step:865/1530 train_loss:3.5962 train_time:144754ms step_avg:169.30ms
step:866/1530 train_loss:3.8904 train_time:144931ms step_avg:169.31ms
step:867/1530 train_loss:3.4539 train_time:145103ms step_avg:169.32ms
step:868/1530 train_loss:3.6399 train_time:145277ms step_avg:169.32ms
step:869/1530 train_loss:3.6140 train_time:145452ms step_avg:169.33ms
step:870/1530 train_loss:3.4435 train_time:145626ms step_avg:169.33ms
step:871/1530 train_loss:3.3922 train_time:145801ms step_avg:169.34ms
step:872/1530 train_loss:3.6448 train_time:145979ms step_avg:169.35ms
step:873/1530 train_loss:3.4623 train_time:146151ms step_avg:169.35ms
step:874/1530 train_loss:3.2249 train_time:146327ms step_avg:169.36ms
step:875/1530 train_loss:3.6316 train_time:146501ms step_avg:169.37ms
step:875/1530 val_loss:3.5179 train_time:146551ms step_avg:169.42ms
step:876/1530 train_loss:3.4369 train_time:146678ms step_avg:169.37ms
step:877/1530 train_loss:3.6209 train_time:146854ms step_avg:169.38ms
step:878/1530 train_loss:3.4626 train_time:147029ms step_avg:169.39ms
step:879/1530 train_loss:3.6495 train_time:147201ms step_avg:169.39ms
step:880/1530 train_loss:3.3054 train_time:147375ms step_avg:169.40ms
step:881/1530 train_loss:3.4785 train_time:147547ms step_avg:169.40ms
step:882/1530 train_loss:3.6934 train_time:147719ms step_avg:169.40ms
step:883/1530 train_loss:3.8385 train_time:147892ms step_avg:169.41ms
step:884/1530 train_loss:3.5656 train_time:148067ms step_avg:169.41ms
step:885/1530 train_loss:3.4955 train_time:148239ms step_avg:169.42ms
step:886/1530 train_loss:3.5678 train_time:148414ms step_avg:169.42ms
step:887/1530 train_loss:4.0871 train_time:148590ms step_avg:169.43ms
step:888/1530 train_loss:3.8404 train_time:148770ms step_avg:169.44ms
step:889/1530 train_loss:3.5202 train_time:148943ms step_avg:169.45ms
step:890/1530 train_loss:3.5327 train_time:149115ms step_avg:169.45ms
step:891/1530 train_loss:3.3591 train_time:149289ms step_avg:169.45ms
step:892/1530 train_loss:3.7135 train_time:149462ms step_avg:169.46ms
step:893/1530 train_loss:3.4174 train_time:149634ms step_avg:169.46ms
step:894/1530 train_loss:3.6369 train_time:149810ms step_avg:169.47ms
step:895/1530 train_loss:3.6750 train_time:149984ms step_avg:169.47ms
step:896/1530 train_loss:3.4970 train_time:150159ms step_avg:169.48ms
step:897/1530 train_loss:3.5426 train_time:150336ms step_avg:169.49ms
step:898/1530 train_loss:3.5905 train_time:150512ms step_avg:169.50ms
step:899/1530 train_loss:3.4773 train_time:150684ms step_avg:169.50ms
step:900/1530 train_loss:3.4257 train_time:150856ms step_avg:169.50ms
step:901/1530 train_loss:3.6188 train_time:151030ms step_avg:169.51ms
step:902/1530 train_loss:3.6278 train_time:151201ms step_avg:169.51ms
step:903/1530 train_loss:3.5395 train_time:151378ms step_avg:169.52ms
step:904/1530 train_loss:3.4883 train_time:151552ms step_avg:169.52ms
step:905/1530 train_loss:3.5001 train_time:151723ms step_avg:169.52ms
step:906/1530 train_loss:3.7051 train_time:151899ms step_avg:169.53ms
step:907/1530 train_loss:3.5107 train_time:152073ms step_avg:169.54ms
step:908/1530 train_loss:3.5652 train_time:152244ms step_avg:169.54ms
step:909/1530 train_loss:3.4528 train_time:152421ms step_avg:169.54ms
step:910/1530 train_loss:3.5321 train_time:152600ms step_avg:169.56ms
step:911/1530 train_loss:3.6459 train_time:152778ms step_avg:169.57ms
step:912/1530 train_loss:3.5944 train_time:152956ms step_avg:169.57ms
step:913/1530 train_loss:3.4626 train_time:153136ms step_avg:169.59ms
step:914/1530 train_loss:3.7416 train_time:153314ms step_avg:169.60ms
step:915/1530 train_loss:3.5313 train_time:153497ms step_avg:169.61ms
step:916/1530 train_loss:3.6177 train_time:153674ms step_avg:169.62ms
step:917/1530 train_loss:3.5971 train_time:153848ms step_avg:169.62ms
step:918/1530 train_loss:4.8343 train_time:154027ms step_avg:169.63ms
step:919/1530 train_loss:3.5016 train_time:154205ms step_avg:169.64ms
step:920/1530 train_loss:3.5897 train_time:154380ms step_avg:169.65ms
step:921/1530 train_loss:3.5471 train_time:154557ms step_avg:169.66ms
step:922/1530 train_loss:3.5827 train_time:154735ms step_avg:169.67ms
step:923/1530 train_loss:3.6056 train_time:154910ms step_avg:169.67ms
step:924/1530 train_loss:3.6779 train_time:155088ms step_avg:169.68ms
step:925/1530 train_loss:3.6479 train_time:155263ms step_avg:169.69ms
step:926/1530 train_loss:3.5578 train_time:155436ms step_avg:169.69ms
step:927/1530 train_loss:3.5529 train_time:155611ms step_avg:169.70ms
step:928/1530 train_loss:3.7808 train_time:155789ms step_avg:169.70ms
step:929/1530 train_loss:3.6096 train_time:155962ms step_avg:169.71ms
step:930/1530 train_loss:3.4030 train_time:156138ms step_avg:169.72ms
step:931/1530 train_loss:3.4942 train_time:156313ms step_avg:169.72ms
step:932/1530 train_loss:3.6450 train_time:156493ms step_avg:169.73ms
step:933/1530 train_loss:3.3766 train_time:156668ms step_avg:169.74ms
step:934/1530 train_loss:3.5896 train_time:156845ms step_avg:169.75ms
step:935/1530 train_loss:3.4397 train_time:157024ms step_avg:169.76ms
step:936/1530 train_loss:3.5170 train_time:157202ms step_avg:169.76ms
step:937/1530 train_loss:3.6213 train_time:157380ms step_avg:169.77ms
step:938/1530 train_loss:3.5440 train_time:157554ms step_avg:169.78ms
step:939/1530 train_loss:3.6730 train_time:157733ms step_avg:169.79ms
step:940/1530 train_loss:3.4811 train_time:157907ms step_avg:169.79ms
step:941/1530 train_loss:3.5441 train_time:158082ms step_avg:169.80ms
step:942/1530 train_loss:3.3552 train_time:158260ms step_avg:169.81ms
step:943/1530 train_loss:3.7099 train_time:158441ms step_avg:169.82ms
step:944/1530 train_loss:3.4034 train_time:158764ms step_avg:169.98ms
step:945/1530 train_loss:3.4279 train_time:158947ms step_avg:170.00ms
step:946/1530 train_loss:5.0783 train_time:159127ms step_avg:170.01ms
step:947/1530 train_loss:3.5986 train_time:159303ms step_avg:170.01ms
step:948/1530 train_loss:3.4810 train_time:159478ms step_avg:170.02ms
step:949/1530 train_loss:3.3748 train_time:159844ms step_avg:170.23ms
step:950/1530 train_loss:3.4415 train_time:160019ms step_avg:170.23ms
step:951/1530 train_loss:3.4091 train_time:160198ms step_avg:170.24ms
step:952/1530 train_loss:3.4775 train_time:160374ms step_avg:170.25ms
step:953/1530 train_loss:3.5660 train_time:160551ms step_avg:170.26ms
step:954/1530 train_loss:3.4464 train_time:160727ms step_avg:170.26ms
step:955/1530 train_loss:3.4771 train_time:160902ms step_avg:170.27ms
step:956/1530 train_loss:3.4411 train_time:161078ms step_avg:170.27ms
step:957/1530 train_loss:3.4930 train_time:161258ms step_avg:170.28ms
step:958/1530 train_loss:3.5023 train_time:161437ms step_avg:170.29ms
step:959/1530 train_loss:3.5102 train_time:161616ms step_avg:170.30ms
step:960/1530 train_loss:3.4051 train_time:161794ms step_avg:170.31ms
step:961/1530 train_loss:3.6424 train_time:161969ms step_avg:170.31ms
step:962/1530 train_loss:3.5872 train_time:162145ms step_avg:170.32ms
step:963/1530 train_loss:3.4910 train_time:162322ms step_avg:170.33ms
step:964/1530 train_loss:3.4288 train_time:162500ms step_avg:170.34ms
step:965/1530 train_loss:3.4733 train_time:162673ms step_avg:170.34ms
step:966/1530 train_loss:3.7056 train_time:162848ms step_avg:170.34ms
step:967/1530 train_loss:3.5150 train_time:163022ms step_avg:170.35ms
step:968/1530 train_loss:3.5175 train_time:163199ms step_avg:170.35ms
step:969/1530 train_loss:3.5804 train_time:163375ms step_avg:170.36ms
step:970/1530 train_loss:3.3699 train_time:163546ms step_avg:170.36ms
step:971/1530 train_loss:3.5279 train_time:163720ms step_avg:170.36ms
step:972/1530 train_loss:3.4706 train_time:163896ms step_avg:170.37ms
step:973/1530 train_loss:3.5399 train_time:164069ms step_avg:170.37ms
step:974/1530 train_loss:3.5867 train_time:164244ms step_avg:170.38ms
step:975/1530 train_loss:3.4618 train_time:164419ms step_avg:170.38ms
step:976/1530 train_loss:3.6691 train_time:164594ms step_avg:170.39ms
step:977/1530 train_loss:3.5671 train_time:164768ms step_avg:170.39ms
step:978/1530 train_loss:3.3511 train_time:164941ms step_avg:170.39ms
step:979/1530 train_loss:3.6233 train_time:165117ms step_avg:170.40ms
step:980/1530 train_loss:3.4173 train_time:165293ms step_avg:170.41ms
step:981/1530 train_loss:3.5706 train_time:165471ms step_avg:170.41ms
step:982/1530 train_loss:3.5380 train_time:165646ms step_avg:170.42ms
step:983/1530 train_loss:3.5058 train_time:165822ms step_avg:170.42ms
step:984/1530 train_loss:3.4944 train_time:165998ms step_avg:170.43ms
step:985/1530 train_loss:3.5749 train_time:166176ms step_avg:170.44ms
step:986/1530 train_loss:3.4116 train_time:166351ms step_avg:170.44ms
step:987/1530 train_loss:3.4864 train_time:166523ms step_avg:170.44ms
step:988/1530 train_loss:3.4783 train_time:166698ms step_avg:170.45ms
step:989/1530 train_loss:3.4141 train_time:166873ms step_avg:170.45ms
step:990/1530 train_loss:3.6616 train_time:167047ms step_avg:170.46ms
step:991/1530 train_loss:3.4669 train_time:167221ms step_avg:170.46ms
step:992/1530 train_loss:3.4423 train_time:167400ms step_avg:170.47ms
step:993/1530 train_loss:3.4998 train_time:167579ms step_avg:170.48ms
step:994/1530 train_loss:3.5934 train_time:167754ms step_avg:170.48ms
step:995/1530 train_loss:3.5261 train_time:167925ms step_avg:170.48ms
step:996/1530 train_loss:3.4558 train_time:168099ms step_avg:170.49ms
step:997/1530 train_loss:3.7488 train_time:168274ms step_avg:170.49ms
step:998/1530 train_loss:3.4409 train_time:168446ms step_avg:170.49ms
step:999/1530 train_loss:3.5846 train_time:168621ms step_avg:170.50ms
step:1000/1530 train_loss:3.4390 train_time:168798ms step_avg:170.50ms
step:1000/1530 val_loss:3.4652 train_time:168849ms step_avg:170.55ms
step:1001/1530 train_loss:3.4988 train_time:168972ms step_avg:170.51ms
step:1002/1530 train_loss:3.3759 train_time:169148ms step_avg:170.51ms
step:1003/1530 train_loss:3.5546 train_time:169324ms step_avg:170.52ms
step:1004/1530 train_loss:3.5970 train_time:169500ms step_avg:170.52ms
step:1005/1530 train_loss:3.3933 train_time:169675ms step_avg:170.53ms
step:1006/1530 train_loss:3.4605 train_time:169850ms step_avg:170.53ms
step:1007/1530 train_loss:3.4352 train_time:170026ms step_avg:170.54ms
step:1008/1530 train_loss:3.5567 train_time:170204ms step_avg:170.55ms
step:1009/1530 train_loss:3.6636 train_time:170383ms step_avg:170.55ms
step:1010/1530 train_loss:3.5645 train_time:170556ms step_avg:170.56ms
step:1011/1530 train_loss:3.5302 train_time:170730ms step_avg:170.56ms
step:1012/1530 train_loss:3.3868 train_time:170905ms step_avg:170.56ms
step:1013/1530 train_loss:3.5329 train_time:171081ms step_avg:170.57ms
step:1014/1530 train_loss:3.6155 train_time:171256ms step_avg:170.57ms
step:1015/1530 train_loss:3.3296 train_time:171433ms step_avg:170.58ms
step:1016/1530 train_loss:3.4077 train_time:171608ms step_avg:170.58ms
step:1017/1530 train_loss:3.3920 train_time:171785ms step_avg:170.59ms
step:1018/1530 train_loss:3.3938 train_time:171961ms step_avg:170.60ms
step:1019/1530 train_loss:3.5180 train_time:172136ms step_avg:170.60ms
step:1020/1530 train_loss:3.3790 train_time:172312ms step_avg:170.61ms
step:1021/1530 train_loss:3.3486 train_time:172487ms step_avg:170.61ms
step:1022/1530 train_loss:3.4762 train_time:172664ms step_avg:170.62ms
step:1023/1530 train_loss:3.5045 train_time:172839ms step_avg:170.62ms
step:1024/1530 train_loss:3.4702 train_time:173015ms step_avg:170.63ms
step:1025/1530 train_loss:3.4788 train_time:173194ms step_avg:170.63ms
step:1026/1530 train_loss:3.6211 train_time:173370ms step_avg:170.64ms
step:1027/1530 train_loss:3.3202 train_time:173548ms step_avg:170.65ms
step:1028/1530 train_loss:3.3953 train_time:173728ms step_avg:170.66ms
step:1029/1530 train_loss:3.3075 train_time:173909ms step_avg:170.67ms
step:1030/1530 train_loss:3.5353 train_time:174087ms step_avg:170.67ms
step:1031/1530 train_loss:3.5081 train_time:174264ms step_avg:170.68ms
step:1032/1530 train_loss:3.6975 train_time:174445ms step_avg:170.69ms
step:1033/1530 train_loss:3.4906 train_time:174620ms step_avg:170.69ms
step:1034/1530 train_loss:3.3932 train_time:174798ms step_avg:170.70ms
step:1035/1530 train_loss:3.4459 train_time:174974ms step_avg:170.71ms
step:1036/1530 train_loss:3.4762 train_time:175150ms step_avg:170.71ms
step:1037/1530 train_loss:3.7850 train_time:175329ms step_avg:170.72ms
step:1038/1530 train_loss:3.6145 train_time:175509ms step_avg:170.73ms
step:1039/1530 train_loss:3.5062 train_time:175691ms step_avg:170.74ms
step:1040/1530 train_loss:3.4111 train_time:175868ms step_avg:170.75ms
step:1041/1530 train_loss:3.4858 train_time:176046ms step_avg:170.75ms
step:1042/1530 train_loss:3.5242 train_time:176220ms step_avg:170.76ms
step:1043/1530 train_loss:3.4421 train_time:176397ms step_avg:170.76ms
step:1044/1530 train_loss:3.4551 train_time:176573ms step_avg:170.77ms
step:1045/1530 train_loss:3.5151 train_time:176750ms step_avg:170.77ms
step:1046/1530 train_loss:3.4256 train_time:176927ms step_avg:170.78ms
step:1047/1530 train_loss:3.6299 train_time:177105ms step_avg:170.79ms
step:1048/1530 train_loss:3.4923 train_time:177280ms step_avg:170.79ms
step:1049/1530 train_loss:3.3987 train_time:177456ms step_avg:170.79ms
step:1050/1530 train_loss:3.3929 train_time:177633ms step_avg:170.80ms
step:1051/1530 train_loss:3.4955 train_time:177810ms step_avg:170.81ms
step:1052/1530 train_loss:3.3612 train_time:177990ms step_avg:170.82ms
step:1053/1530 train_loss:3.6882 train_time:178168ms step_avg:170.82ms
step:1054/1530 train_loss:3.5395 train_time:178347ms step_avg:170.83ms
step:1055/1530 train_loss:3.3849 train_time:178523ms step_avg:170.84ms
step:1056/1530 train_loss:3.4955 train_time:178698ms step_avg:170.84ms
step:1057/1530 train_loss:3.5799 train_time:178876ms step_avg:170.85ms
step:1058/1530 train_loss:3.3006 train_time:179055ms step_avg:170.85ms
step:1059/1530 train_loss:3.3677 train_time:179235ms step_avg:170.86ms
step:1060/1530 train_loss:3.4340 train_time:179411ms step_avg:170.87ms
step:1061/1530 train_loss:3.4180 train_time:179587ms step_avg:170.87ms
step:1062/1530 train_loss:3.3806 train_time:179763ms step_avg:170.88ms
step:1063/1530 train_loss:3.4554 train_time:179937ms step_avg:170.88ms
step:1064/1530 train_loss:3.3776 train_time:180112ms step_avg:170.88ms
step:1065/1530 train_loss:3.3590 train_time:180290ms step_avg:170.89ms
step:1066/1530 train_loss:3.4063 train_time:180467ms step_avg:170.90ms
step:1067/1530 train_loss:3.2774 train_time:180646ms step_avg:170.90ms
step:1068/1530 train_loss:3.4373 train_time:180822ms step_avg:170.91ms
step:1069/1530 train_loss:3.2993 train_time:181004ms step_avg:170.92ms
step:1070/1530 train_loss:3.5629 train_time:181180ms step_avg:170.92ms
step:1071/1530 train_loss:3.5136 train_time:181361ms step_avg:170.93ms
step:1072/1530 train_loss:3.4342 train_time:181534ms step_avg:170.94ms
step:1073/1530 train_loss:3.5196 train_time:181708ms step_avg:170.94ms
step:1074/1530 train_loss:3.4322 train_time:181886ms step_avg:170.95ms
step:1075/1530 train_loss:3.4014 train_time:182064ms step_avg:170.95ms
step:1076/1530 train_loss:3.7986 train_time:182241ms step_avg:170.96ms
step:1077/1530 train_loss:3.4396 train_time:182414ms step_avg:170.96ms
step:1078/1530 train_loss:3.0923 train_time:182599ms step_avg:170.97ms
step:1079/1530 train_loss:3.5322 train_time:182776ms step_avg:170.98ms
step:1080/1530 train_loss:3.4276 train_time:182952ms step_avg:170.98ms
step:1081/1530 train_loss:3.4967 train_time:183127ms step_avg:170.99ms
step:1082/1530 train_loss:3.5861 train_time:183304ms step_avg:170.99ms
step:1083/1530 train_loss:3.4913 train_time:183478ms step_avg:171.00ms
step:1084/1530 train_loss:3.4599 train_time:183654ms step_avg:171.00ms
step:1085/1530 train_loss:3.4321 train_time:183829ms step_avg:171.00ms
step:1086/1530 train_loss:3.6260 train_time:184006ms step_avg:171.01ms
step:1087/1530 train_loss:3.5023 train_time:184181ms step_avg:171.01ms
step:1088/1530 train_loss:3.3698 train_time:184357ms step_avg:171.02ms
step:1089/1530 train_loss:3.3747 train_time:184537ms step_avg:171.03ms
step:1090/1530 train_loss:3.4795 train_time:184715ms step_avg:171.03ms
step:1091/1530 train_loss:3.2812 train_time:184892ms step_avg:171.04ms
step:1092/1530 train_loss:3.4808 train_time:185070ms step_avg:171.04ms
step:1093/1530 train_loss:3.5998 train_time:185247ms step_avg:171.05ms
step:1094/1530 train_loss:3.4456 train_time:185422ms step_avg:171.05ms
step:1095/1530 train_loss:3.4183 train_time:185597ms step_avg:171.06ms
step:1096/1530 train_loss:3.4225 train_time:185774ms step_avg:171.06ms
step:1097/1530 train_loss:3.4856 train_time:185951ms step_avg:171.07ms
step:1098/1530 train_loss:3.5620 train_time:186130ms step_avg:171.08ms
step:1099/1530 train_loss:3.5251 train_time:186309ms step_avg:171.08ms
step:1100/1530 train_loss:3.4224 train_time:186488ms step_avg:171.09ms
step:1101/1530 train_loss:3.2876 train_time:186667ms step_avg:171.10ms
step:1102/1530 train_loss:3.3042 train_time:186846ms step_avg:171.10ms
step:1103/1530 train_loss:3.4413 train_time:187028ms step_avg:171.11ms
step:1104/1530 train_loss:3.3240 train_time:187205ms step_avg:171.12ms
step:1105/1530 train_loss:4.0618 train_time:187383ms step_avg:171.13ms
step:1106/1530 train_loss:3.2227 train_time:187559ms step_avg:171.13ms
step:1107/1530 train_loss:3.5622 train_time:187733ms step_avg:171.13ms
step:1108/1530 train_loss:3.3408 train_time:187908ms step_avg:171.14ms
step:1109/1530 train_loss:3.5021 train_time:188084ms step_avg:171.14ms
step:1110/1530 train_loss:3.4286 train_time:188257ms step_avg:171.14ms
step:1111/1530 train_loss:3.4872 train_time:188433ms step_avg:171.15ms
step:1112/1530 train_loss:3.5558 train_time:188613ms step_avg:171.16ms
step:1113/1530 train_loss:3.4324 train_time:188797ms step_avg:171.17ms
step:1114/1530 train_loss:3.3671 train_time:188976ms step_avg:171.17ms
step:1115/1530 train_loss:3.2430 train_time:189153ms step_avg:171.18ms
step:1116/1530 train_loss:3.4214 train_time:189327ms step_avg:171.18ms
step:1117/1530 train_loss:3.5873 train_time:189507ms step_avg:171.19ms
step:1118/1530 train_loss:3.6221 train_time:189685ms step_avg:171.20ms
step:1119/1530 train_loss:3.4790 train_time:189858ms step_avg:171.20ms
step:1120/1530 train_loss:3.4860 train_time:190036ms step_avg:171.20ms
step:1121/1530 train_loss:3.3899 train_time:190213ms step_avg:171.21ms
step:1122/1530 train_loss:3.4604 train_time:190390ms step_avg:171.21ms
step:1123/1530 train_loss:3.5761 train_time:190566ms step_avg:171.22ms
step:1124/1530 train_loss:3.3406 train_time:190740ms step_avg:171.22ms
step:1125/1530 train_loss:3.2235 train_time:190916ms step_avg:171.23ms
step:1125/1530 val_loss:3.4058 train_time:190966ms step_avg:171.27ms
step:1126/1530 train_loss:3.4738 train_time:191095ms step_avg:171.23ms
step:1127/1530 train_loss:3.6660 train_time:191274ms step_avg:171.24ms
step:1128/1530 train_loss:3.2251 train_time:191453ms step_avg:171.25ms
step:1129/1530 train_loss:3.5500 train_time:191633ms step_avg:171.25ms
step:1130/1530 train_loss:3.3803 train_time:191812ms step_avg:171.26ms
step:1131/1530 train_loss:3.3990 train_time:191997ms step_avg:171.27ms
step:1132/1530 train_loss:3.3649 train_time:192171ms step_avg:171.28ms
step:1133/1530 train_loss:3.4855 train_time:192490ms step_avg:171.41ms
step:1134/1530 train_loss:3.4412 train_time:192676ms step_avg:171.42ms
step:1135/1530 train_loss:3.5158 train_time:192854ms step_avg:171.43ms
step:1136/1530 train_loss:3.5577 train_time:193032ms step_avg:171.43ms
step:1137/1530 train_loss:3.4526 train_time:193208ms step_avg:171.44ms
step:1138/1530 train_loss:3.3531 train_time:193388ms step_avg:171.44ms
step:1139/1530 train_loss:3.6527 train_time:193718ms step_avg:171.58ms
step:1140/1530 train_loss:3.4527 train_time:193893ms step_avg:171.59ms
step:1141/1530 train_loss:3.5907 train_time:194073ms step_avg:171.59ms
step:1142/1530 train_loss:3.4393 train_time:194252ms step_avg:171.60ms
step:1143/1530 train_loss:3.3591 train_time:194431ms step_avg:171.61ms
step:1144/1530 train_loss:3.4418 train_time:194608ms step_avg:171.61ms
step:1145/1530 train_loss:3.5868 train_time:194784ms step_avg:171.62ms
step:1146/1530 train_loss:3.5537 train_time:194964ms step_avg:171.62ms
step:1147/1530 train_loss:3.4878 train_time:195143ms step_avg:171.63ms
step:1148/1530 train_loss:3.4995 train_time:195320ms step_avg:171.63ms
step:1149/1530 train_loss:3.3194 train_time:195501ms step_avg:171.64ms
step:1150/1530 train_loss:3.3754 train_time:195676ms step_avg:171.65ms
step:1151/1530 train_loss:3.3201 train_time:195855ms step_avg:171.65ms
step:1152/1530 train_loss:3.3919 train_time:196036ms step_avg:171.66ms
step:1153/1530 train_loss:3.4290 train_time:196217ms step_avg:171.67ms
step:1154/1530 train_loss:3.5128 train_time:196394ms step_avg:171.67ms
step:1155/1530 train_loss:3.3188 train_time:196579ms step_avg:171.68ms
step:1156/1530 train_loss:3.5329 train_time:196760ms step_avg:171.69ms
step:1157/1530 train_loss:3.4909 train_time:196937ms step_avg:171.70ms
step:1158/1530 train_loss:3.2483 train_time:197115ms step_avg:171.70ms
step:1159/1530 train_loss:3.3453 train_time:197292ms step_avg:171.71ms
step:1160/1530 train_loss:3.3310 train_time:197465ms step_avg:171.71ms
step:1161/1530 train_loss:3.0672 train_time:197644ms step_avg:171.71ms
step:1162/1530 train_loss:3.4211 train_time:197821ms step_avg:171.72ms
step:1163/1530 train_loss:3.3837 train_time:197999ms step_avg:171.73ms
step:1164/1530 train_loss:3.2932 train_time:198176ms step_avg:171.73ms
step:1165/1530 train_loss:3.2480 train_time:198352ms step_avg:171.73ms
step:1166/1530 train_loss:3.3872 train_time:198532ms step_avg:171.74ms
step:1167/1530 train_loss:3.4129 train_time:198710ms step_avg:171.75ms
step:1168/1530 train_loss:3.7154 train_time:198885ms step_avg:171.75ms
step:1169/1530 train_loss:3.3719 train_time:199062ms step_avg:171.75ms
step:1170/1530 train_loss:3.3901 train_time:199238ms step_avg:171.76ms
step:1171/1530 train_loss:3.2932 train_time:199415ms step_avg:171.76ms
step:1172/1530 train_loss:3.4226 train_time:199592ms step_avg:171.77ms
step:1173/1530 train_loss:3.5390 train_time:199773ms step_avg:171.77ms
step:1174/1530 train_loss:3.3826 train_time:199958ms step_avg:171.79ms
step:1175/1530 train_loss:3.3583 train_time:200137ms step_avg:171.79ms
step:1176/1530 train_loss:3.4220 train_time:200319ms step_avg:171.80ms
step:1177/1530 train_loss:3.4472 train_time:200500ms step_avg:171.81ms
step:1178/1530 train_loss:3.4919 train_time:200677ms step_avg:171.81ms
step:1179/1530 train_loss:3.3988 train_time:200852ms step_avg:171.82ms
step:1180/1530 train_loss:3.3521 train_time:201040ms step_avg:171.83ms
step:1181/1530 train_loss:3.3327 train_time:201219ms step_avg:171.83ms
step:1182/1530 train_loss:3.3671 train_time:201397ms step_avg:171.84ms
step:1183/1530 train_loss:3.3314 train_time:201574ms step_avg:171.84ms
step:1184/1530 train_loss:3.5072 train_time:201751ms step_avg:171.85ms
step:1185/1530 train_loss:3.5379 train_time:201932ms step_avg:171.86ms
step:1186/1530 train_loss:3.3624 train_time:202113ms step_avg:171.86ms
step:1187/1530 train_loss:3.4144 train_time:202299ms step_avg:171.88ms
step:1188/1530 train_loss:3.4392 train_time:202476ms step_avg:171.88ms
step:1189/1530 train_loss:3.2727 train_time:202656ms step_avg:171.89ms
step:1190/1530 train_loss:3.4430 train_time:202835ms step_avg:171.89ms
step:1191/1530 train_loss:3.5804 train_time:203016ms step_avg:171.90ms
step:1192/1530 train_loss:3.3881 train_time:203191ms step_avg:171.90ms
step:1193/1530 train_loss:3.2764 train_time:203366ms step_avg:171.91ms
step:1194/1530 train_loss:3.5503 train_time:203542ms step_avg:171.91ms
step:1195/1530 train_loss:3.3709 train_time:203724ms step_avg:171.92ms
step:1196/1530 train_loss:3.3780 train_time:203914ms step_avg:171.93ms
step:1197/1530 train_loss:3.2873 train_time:204093ms step_avg:171.94ms
step:1198/1530 train_loss:3.3006 train_time:204278ms step_avg:171.95ms
step:1199/1530 train_loss:3.3412 train_time:204457ms step_avg:171.96ms
step:1200/1530 train_loss:3.4447 train_time:204636ms step_avg:171.96ms
step:1201/1530 train_loss:3.4769 train_time:204815ms step_avg:171.97ms
step:1202/1530 train_loss:3.6122 train_time:205003ms step_avg:171.98ms
step:1203/1530 train_loss:3.4012 train_time:205181ms step_avg:171.99ms
step:1204/1530 train_loss:3.3065 train_time:205361ms step_avg:171.99ms
step:1205/1530 train_loss:3.4325 train_time:205538ms step_avg:172.00ms
step:1206/1530 train_loss:3.4761 train_time:205716ms step_avg:172.00ms
step:1207/1530 train_loss:3.5153 train_time:205894ms step_avg:172.01ms
step:1208/1530 train_loss:3.3948 train_time:206070ms step_avg:172.01ms
step:1209/1530 train_loss:3.2443 train_time:206251ms step_avg:172.02ms
step:1210/1530 train_loss:3.3039 train_time:206429ms step_avg:172.02ms
step:1211/1530 train_loss:3.3979 train_time:206607ms step_avg:172.03ms
step:1212/1530 train_loss:3.3902 train_time:206782ms step_avg:172.03ms
step:1213/1530 train_loss:3.4085 train_time:206960ms step_avg:172.04ms
step:1214/1530 train_loss:3.2514 train_time:207140ms step_avg:172.04ms
step:1215/1530 train_loss:3.3902 train_time:207317ms step_avg:172.05ms
step:1216/1530 train_loss:3.3300 train_time:207496ms step_avg:172.05ms
step:1217/1530 train_loss:3.3159 train_time:207673ms step_avg:172.06ms
step:1218/1530 train_loss:3.4093 train_time:207850ms step_avg:172.06ms
step:1219/1530 train_loss:3.2549 train_time:208035ms step_avg:172.07ms
step:1220/1530 train_loss:3.4730 train_time:208212ms step_avg:172.08ms
step:1221/1530 train_loss:3.5012 train_time:208388ms step_avg:172.08ms
step:1222/1530 train_loss:3.4252 train_time:208561ms step_avg:172.08ms
step:1223/1530 train_loss:3.2897 train_time:208739ms step_avg:172.09ms
step:1224/1530 train_loss:3.2567 train_time:208923ms step_avg:172.09ms
step:1225/1530 train_loss:3.3611 train_time:209100ms step_avg:172.10ms
step:1226/1530 train_loss:3.3292 train_time:209281ms step_avg:172.11ms
step:1227/1530 train_loss:3.2713 train_time:209459ms step_avg:172.11ms
step:1228/1530 train_loss:3.4379 train_time:209636ms step_avg:172.11ms
step:1229/1530 train_loss:3.3643 train_time:209816ms step_avg:172.12ms
step:1230/1530 train_loss:3.3979 train_time:210000ms step_avg:172.13ms
step:1231/1530 train_loss:3.5738 train_time:210180ms step_avg:172.14ms
step:1232/1530 train_loss:3.4907 train_time:210360ms step_avg:172.14ms
step:1233/1530 train_loss:3.4269 train_time:210538ms step_avg:172.15ms
step:1234/1530 train_loss:3.5817 train_time:210717ms step_avg:172.15ms
step:1235/1530 train_loss:3.3214 train_time:210898ms step_avg:172.16ms
step:1236/1530 train_loss:3.2872 train_time:211075ms step_avg:172.17ms
step:1237/1530 train_loss:3.2662 train_time:211253ms step_avg:172.17ms
step:1238/1530 train_loss:3.2758 train_time:211438ms step_avg:172.18ms
step:1239/1530 train_loss:3.3282 train_time:211616ms step_avg:172.19ms
step:1240/1530 train_loss:3.3807 train_time:211794ms step_avg:172.19ms
step:1241/1530 train_loss:3.4267 train_time:211972ms step_avg:172.19ms
step:1242/1530 train_loss:3.2950 train_time:212150ms step_avg:172.20ms
step:1243/1530 train_loss:3.4036 train_time:212329ms step_avg:172.21ms
step:1244/1530 train_loss:3.3994 train_time:212503ms step_avg:172.21ms
step:1245/1530 train_loss:3.4076 train_time:212679ms step_avg:172.21ms
step:1246/1530 train_loss:3.2395 train_time:212857ms step_avg:172.21ms
step:1247/1530 train_loss:3.3718 train_time:213033ms step_avg:172.22ms
step:1248/1530 train_loss:3.4260 train_time:213210ms step_avg:172.22ms
step:1249/1530 train_loss:3.4205 train_time:213387ms step_avg:172.23ms
step:1250/1530 train_loss:3.3021 train_time:213566ms step_avg:172.23ms
step:1250/1530 val_loss:3.3528 train_time:213619ms step_avg:172.27ms
step:1251/1530 train_loss:3.4878 train_time:213750ms step_avg:172.24ms
step:1252/1530 train_loss:3.3566 train_time:213926ms step_avg:172.24ms
step:1253/1530 train_loss:3.3043 train_time:214103ms step_avg:172.25ms
step:1254/1530 train_loss:3.4177 train_time:214287ms step_avg:172.26ms
step:1255/1530 train_loss:3.5172 train_time:214475ms step_avg:172.27ms
step:1256/1530 train_loss:3.3065 train_time:214658ms step_avg:172.28ms
step:1257/1530 train_loss:3.3714 train_time:214836ms step_avg:172.28ms
step:1258/1530 train_loss:3.3633 train_time:215018ms step_avg:172.29ms
step:1259/1530 train_loss:3.3234 train_time:215196ms step_avg:172.29ms
step:1260/1530 train_loss:3.2051 train_time:215372ms step_avg:172.30ms
step:1261/1530 train_loss:3.3011 train_time:215551ms step_avg:172.30ms
step:1262/1530 train_loss:3.3210 train_time:215734ms step_avg:172.31ms
step:1263/1530 train_loss:3.2343 train_time:215915ms step_avg:172.32ms
step:1264/1530 train_loss:3.4409 train_time:216091ms step_avg:172.32ms
step:1265/1530 train_loss:3.4254 train_time:216265ms step_avg:172.32ms
step:1266/1530 train_loss:3.4390 train_time:216445ms step_avg:172.33ms
step:1267/1530 train_loss:3.3683 train_time:216623ms step_avg:172.33ms
step:1268/1530 train_loss:3.4114 train_time:216806ms step_avg:172.34ms
step:1269/1530 train_loss:3.2533 train_time:216990ms step_avg:172.35ms
step:1270/1530 train_loss:3.1047 train_time:217168ms step_avg:172.36ms
step:1271/1530 train_loss:3.4019 train_time:217346ms step_avg:172.36ms
step:1272/1530 train_loss:3.3515 train_time:217523ms step_avg:172.36ms
step:1273/1530 train_loss:3.3718 train_time:217706ms step_avg:172.37ms
step:1274/1530 train_loss:3.3515 train_time:217888ms step_avg:172.38ms
step:1275/1530 train_loss:3.4285 train_time:218065ms step_avg:172.38ms
step:1276/1530 train_loss:3.4677 train_time:218240ms step_avg:172.39ms
step:1277/1530 train_loss:3.4081 train_time:218420ms step_avg:172.39ms
step:1278/1530 train_loss:3.4078 train_time:218597ms step_avg:172.40ms
step:1279/1530 train_loss:3.2624 train_time:218777ms step_avg:172.40ms
step:1280/1530 train_loss:3.3627 train_time:218964ms step_avg:172.41ms
step:1281/1530 train_loss:3.4239 train_time:219142ms step_avg:172.42ms
step:1282/1530 train_loss:3.4686 train_time:219317ms step_avg:172.42ms
step:1283/1530 train_loss:3.3344 train_time:219496ms step_avg:172.42ms
step:1284/1530 train_loss:3.3659 train_time:219675ms step_avg:172.43ms
step:1285/1530 train_loss:3.3636 train_time:219853ms step_avg:172.43ms
step:1286/1530 train_loss:3.3366 train_time:220030ms step_avg:172.44ms
step:1287/1530 train_loss:3.4878 train_time:220209ms step_avg:172.44ms
step:1288/1530 train_loss:3.2962 train_time:220389ms step_avg:172.45ms
step:1289/1530 train_loss:3.3832 train_time:220575ms step_avg:172.46ms
step:1290/1530 train_loss:3.4572 train_time:220760ms step_avg:172.47ms
step:1291/1530 train_loss:3.3862 train_time:220939ms step_avg:172.47ms
step:1292/1530 train_loss:3.4794 train_time:221123ms step_avg:172.48ms
step:1293/1530 train_loss:3.5143 train_time:221304ms step_avg:172.49ms
step:1294/1530 train_loss:3.4508 train_time:221485ms step_avg:172.50ms
step:1295/1530 train_loss:3.2812 train_time:221666ms step_avg:172.50ms
step:1296/1530 train_loss:3.3733 train_time:221847ms step_avg:172.51ms
step:1297/1530 train_loss:3.2777 train_time:222027ms step_avg:172.52ms
step:1298/1530 train_loss:3.2734 train_time:222208ms step_avg:172.52ms
step:1299/1530 train_loss:3.3985 train_time:222386ms step_avg:172.53ms
step:1300/1530 train_loss:3.4027 train_time:222563ms step_avg:172.53ms
step:1301/1530 train_loss:3.4000 train_time:222742ms step_avg:172.53ms
step:1302/1530 train_loss:3.5781 train_time:222924ms step_avg:172.54ms
step:1303/1530 train_loss:3.3067 train_time:223107ms step_avg:172.55ms
step:1304/1530 train_loss:3.5151 train_time:223288ms step_avg:172.56ms
step:1305/1530 train_loss:3.2593 train_time:223465ms step_avg:172.56ms
step:1306/1530 train_loss:3.4515 train_time:223646ms step_avg:172.57ms
step:1307/1530 train_loss:3.4520 train_time:223822ms step_avg:172.57ms
step:1308/1530 train_loss:3.2817 train_time:224002ms step_avg:172.57ms
step:1309/1530 train_loss:3.3094 train_time:224182ms step_avg:172.58ms
step:1310/1530 train_loss:3.2836 train_time:224359ms step_avg:172.58ms
step:1311/1530 train_loss:3.2950 train_time:224537ms step_avg:172.59ms
step:1312/1530 train_loss:3.3723 train_time:224717ms step_avg:172.59ms
step:1313/1530 train_loss:3.3395 train_time:224893ms step_avg:172.60ms
step:1314/1530 train_loss:3.0392 train_time:225076ms step_avg:172.60ms
step:1315/1530 train_loss:3.2742 train_time:225254ms step_avg:172.61ms
step:1316/1530 train_loss:3.4014 train_time:225428ms step_avg:172.61ms
step:1317/1530 train_loss:3.4156 train_time:225606ms step_avg:172.61ms
step:1318/1530 train_loss:3.3021 train_time:225791ms step_avg:172.62ms
step:1319/1530 train_loss:3.4214 train_time:225970ms step_avg:172.63ms
step:1320/1530 train_loss:3.4598 train_time:226152ms step_avg:172.63ms
step:1321/1530 train_loss:3.3634 train_time:226330ms step_avg:172.64ms
step:1322/1530 train_loss:3.3218 train_time:226650ms step_avg:172.75ms
step:1323/1530 train_loss:3.3206 train_time:226838ms step_avg:172.76ms
step:1324/1530 train_loss:3.4345 train_time:227022ms step_avg:172.77ms
step:1325/1530 train_loss:3.4935 train_time:227208ms step_avg:172.78ms
step:1326/1530 train_loss:3.2105 train_time:227389ms step_avg:172.79ms
step:1327/1530 train_loss:3.1632 train_time:227565ms step_avg:172.79ms
step:1328/1530 train_loss:3.4907 train_time:227745ms step_avg:172.80ms
step:1329/1530 train_loss:3.2959 train_time:228084ms step_avg:172.92ms
step:1330/1530 train_loss:3.4278 train_time:228266ms step_avg:172.93ms
step:1331/1530 train_loss:3.3279 train_time:228442ms step_avg:172.93ms
step:1332/1530 train_loss:3.7412 train_time:228626ms step_avg:172.94ms
step:1333/1530 train_loss:3.4789 train_time:228807ms step_avg:172.95ms
step:1334/1530 train_loss:3.3700 train_time:228987ms step_avg:172.95ms
step:1335/1530 train_loss:3.2881 train_time:229165ms step_avg:172.95ms
step:1336/1530 train_loss:3.2957 train_time:229348ms step_avg:172.96ms
step:1337/1530 train_loss:3.5497 train_time:229528ms step_avg:172.97ms
step:1338/1530 train_loss:3.5193 train_time:229709ms step_avg:172.97ms
step:1339/1530 train_loss:3.3317 train_time:229888ms step_avg:172.98ms
step:1340/1530 train_loss:3.2805 train_time:230065ms step_avg:172.98ms
step:1341/1530 train_loss:3.5917 train_time:230243ms step_avg:172.98ms
step:1342/1530 train_loss:3.3523 train_time:230422ms step_avg:172.99ms
step:1343/1530 train_loss:3.3613 train_time:230601ms step_avg:172.99ms
step:1344/1530 train_loss:3.4193 train_time:230781ms step_avg:173.00ms
step:1345/1530 train_loss:3.3819 train_time:230963ms step_avg:173.01ms
step:1346/1530 train_loss:3.2963 train_time:231140ms step_avg:173.01ms
step:1347/1530 train_loss:3.2776 train_time:231316ms step_avg:173.01ms
step:1348/1530 train_loss:3.3502 train_time:231494ms step_avg:173.01ms
step:1349/1530 train_loss:3.2748 train_time:231671ms step_avg:173.02ms
step:1350/1530 train_loss:3.3885 train_time:231851ms step_avg:173.02ms
step:1351/1530 train_loss:3.2399 train_time:232027ms step_avg:173.03ms
step:1352/1530 train_loss:3.3066 train_time:232205ms step_avg:173.03ms
step:1353/1530 train_loss:3.4015 train_time:232386ms step_avg:173.03ms
step:1354/1530 train_loss:3.2605 train_time:232564ms step_avg:173.04ms
step:1355/1530 train_loss:3.1934 train_time:232741ms step_avg:173.04ms
step:1356/1530 train_loss:3.5110 train_time:232921ms step_avg:173.05ms
step:1357/1530 train_loss:3.4236 train_time:233104ms step_avg:173.05ms
step:1358/1530 train_loss:3.1846 train_time:233284ms step_avg:173.06ms
step:1359/1530 train_loss:3.4399 train_time:233463ms step_avg:173.06ms
step:1360/1530 train_loss:3.3464 train_time:233644ms step_avg:173.07ms
step:1361/1530 train_loss:3.1235 train_time:233831ms step_avg:173.08ms
step:1362/1530 train_loss:3.3939 train_time:234011ms step_avg:173.09ms
step:1363/1530 train_loss:3.2801 train_time:234200ms step_avg:173.10ms
step:1364/1530 train_loss:3.2989 train_time:234379ms step_avg:173.10ms
step:1365/1530 train_loss:3.3128 train_time:234555ms step_avg:173.10ms
step:1366/1530 train_loss:3.4211 train_time:234734ms step_avg:173.11ms
step:1367/1530 train_loss:3.3987 train_time:234912ms step_avg:173.11ms
step:1368/1530 train_loss:3.3469 train_time:235092ms step_avg:173.12ms
step:1369/1530 train_loss:3.2775 train_time:235281ms step_avg:173.13ms
step:1370/1530 train_loss:3.6041 train_time:235461ms step_avg:173.13ms
step:1371/1530 train_loss:3.3143 train_time:235643ms step_avg:173.14ms
step:1372/1530 train_loss:3.3721 train_time:235826ms step_avg:173.15ms
step:1373/1530 train_loss:3.3697 train_time:236006ms step_avg:173.15ms
step:1374/1530 train_loss:3.1488 train_time:236188ms step_avg:173.16ms
step:1375/1530 train_loss:3.5377 train_time:236367ms step_avg:173.16ms
step:1375/1530 val_loss:3.3096 train_time:236418ms step_avg:173.20ms
step:1376/1530 train_loss:3.3412 train_time:236547ms step_avg:173.17ms
step:1377/1530 train_loss:3.4757 train_time:236724ms step_avg:173.17ms
step:1378/1530 train_loss:3.4692 train_time:236902ms step_avg:173.17ms
step:1379/1530 train_loss:3.1116 train_time:237085ms step_avg:173.18ms
step:1380/1530 train_loss:3.3159 train_time:237264ms step_avg:173.19ms
step:1381/1530 train_loss:3.7029 train_time:237450ms step_avg:173.19ms
step:1382/1530 train_loss:3.2086 train_time:237628ms step_avg:173.20ms
step:1383/1530 train_loss:3.3905 train_time:237811ms step_avg:173.21ms
step:1384/1530 train_loss:3.4759 train_time:237996ms step_avg:173.21ms
step:1385/1530 train_loss:3.4072 train_time:238173ms step_avg:173.22ms
step:1386/1530 train_loss:3.3383 train_time:238352ms step_avg:173.22ms
step:1387/1530 train_loss:3.1976 train_time:238531ms step_avg:173.23ms
step:1388/1530 train_loss:3.3457 train_time:238711ms step_avg:173.23ms
step:1389/1530 train_loss:3.3141 train_time:238894ms step_avg:173.24ms
step:1390/1530 train_loss:3.5666 train_time:239072ms step_avg:173.24ms
step:1391/1530 train_loss:3.2914 train_time:239251ms step_avg:173.25ms
step:1392/1530 train_loss:3.2893 train_time:239431ms step_avg:173.25ms
step:1393/1530 train_loss:3.2330 train_time:239613ms step_avg:173.26ms
step:1394/1530 train_loss:3.4978 train_time:239792ms step_avg:173.26ms
step:1395/1530 train_loss:3.3901 train_time:239972ms step_avg:173.26ms
step:1396/1530 train_loss:3.4071 train_time:240149ms step_avg:173.27ms
step:1397/1530 train_loss:3.3069 train_time:240325ms step_avg:173.27ms
step:1398/1530 train_loss:3.2541 train_time:240499ms step_avg:173.27ms
step:1399/1530 train_loss:3.3134 train_time:240677ms step_avg:173.27ms
step:1400/1530 train_loss:3.3187 train_time:240861ms step_avg:173.28ms
step:1401/1530 train_loss:3.3514 train_time:241036ms step_avg:173.28ms
step:1402/1530 train_loss:3.2983 train_time:241216ms step_avg:173.29ms
step:1403/1530 train_loss:3.4915 train_time:241401ms step_avg:173.30ms
step:1404/1530 train_loss:3.2790 train_time:241578ms step_avg:173.30ms
step:1405/1530 train_loss:3.3123 train_time:241758ms step_avg:173.30ms
step:1406/1530 train_loss:3.3145 train_time:241937ms step_avg:173.31ms
step:1407/1530 train_loss:3.1720 train_time:242114ms step_avg:173.31ms
step:1408/1530 train_loss:3.3118 train_time:242294ms step_avg:173.31ms
step:1409/1530 train_loss:3.3006 train_time:242480ms step_avg:173.32ms
step:1410/1530 train_loss:3.2864 train_time:242657ms step_avg:173.33ms
step:1411/1530 train_loss:3.3620 train_time:242832ms step_avg:173.33ms
step:1412/1530 train_loss:3.3332 train_time:243011ms step_avg:173.33ms
step:1413/1530 train_loss:3.3617 train_time:243191ms step_avg:173.34ms
step:1414/1530 train_loss:3.3256 train_time:243372ms step_avg:173.34ms
step:1415/1530 train_loss:3.4048 train_time:243557ms step_avg:173.35ms
step:1416/1530 train_loss:3.2314 train_time:243745ms step_avg:173.36ms
step:1417/1530 train_loss:3.2797 train_time:243927ms step_avg:173.37ms
step:1418/1530 train_loss:3.3892 train_time:244109ms step_avg:173.37ms
step:1419/1530 train_loss:3.3440 train_time:244293ms step_avg:173.38ms
step:1420/1530 train_loss:3.3657 train_time:244474ms step_avg:173.39ms
step:1421/1530 train_loss:3.3669 train_time:244654ms step_avg:173.39ms
step:1422/1530 train_loss:3.3321 train_time:244831ms step_avg:173.39ms
step:1423/1530 train_loss:3.3152 train_time:245012ms step_avg:173.40ms
step:1424/1530 train_loss:3.3362 train_time:245196ms step_avg:173.41ms
step:1425/1530 train_loss:3.1899 train_time:245382ms step_avg:173.41ms
step:1426/1530 train_loss:3.3203 train_time:245559ms step_avg:173.42ms
step:1427/1530 train_loss:3.2845 train_time:245741ms step_avg:173.42ms
step:1428/1530 train_loss:3.3788 train_time:245919ms step_avg:173.43ms
step:1429/1530 train_loss:3.3529 train_time:246096ms step_avg:173.43ms
step:1430/1530 train_loss:3.2603 train_time:246278ms step_avg:173.43ms
step:1431/1530 train_loss:3.3206 train_time:246461ms step_avg:173.44ms
step:1432/1530 train_loss:3.3386 train_time:246641ms step_avg:173.45ms
step:1433/1530 train_loss:3.1282 train_time:246824ms step_avg:173.45ms
step:1434/1530 train_loss:3.2821 train_time:247007ms step_avg:173.46ms
step:1435/1530 train_loss:3.1157 train_time:247188ms step_avg:173.47ms
step:1436/1530 train_loss:3.2260 train_time:247369ms step_avg:173.47ms
step:1437/1530 train_loss:3.4051 train_time:247546ms step_avg:173.47ms
step:1438/1530 train_loss:3.3815 train_time:247722ms step_avg:173.47ms
step:1439/1530 train_loss:3.3162 train_time:247901ms step_avg:173.48ms
step:1440/1530 train_loss:3.1944 train_time:248076ms step_avg:173.48ms
step:1441/1530 train_loss:3.3318 train_time:248254ms step_avg:173.48ms
step:1442/1530 train_loss:3.3885 train_time:248437ms step_avg:173.49ms
step:1443/1530 train_loss:3.4901 train_time:248622ms step_avg:173.50ms
step:1444/1530 train_loss:3.4446 train_time:248799ms step_avg:173.50ms
step:1445/1530 train_loss:3.3369 train_time:248977ms step_avg:173.50ms
step:1446/1530 train_loss:3.1977 train_time:249156ms step_avg:173.51ms
step:1447/1530 train_loss:3.2971 train_time:249337ms step_avg:173.51ms
step:1448/1530 train_loss:3.2951 train_time:249516ms step_avg:173.52ms
step:1449/1530 train_loss:3.3888 train_time:249695ms step_avg:173.52ms
step:1450/1530 train_loss:3.3875 train_time:249876ms step_avg:173.53ms
step:1451/1530 train_loss:3.2073 train_time:250055ms step_avg:173.53ms
step:1452/1530 train_loss:3.3229 train_time:250234ms step_avg:173.53ms
step:1453/1530 train_loss:3.2583 train_time:250410ms step_avg:173.53ms
step:1454/1530 train_loss:3.2868 train_time:250588ms step_avg:173.54ms
step:1455/1530 train_loss:3.3314 train_time:250772ms step_avg:173.54ms
step:1456/1530 train_loss:3.2858 train_time:250949ms step_avg:173.55ms
step:1457/1530 train_loss:3.1550 train_time:251128ms step_avg:173.55ms
step:1458/1530 train_loss:3.4234 train_time:251305ms step_avg:173.55ms
step:1459/1530 train_loss:3.2708 train_time:251487ms step_avg:173.56ms
step:1460/1530 train_loss:3.3161 train_time:251668ms step_avg:173.56ms
step:1461/1530 train_loss:3.4293 train_time:251847ms step_avg:173.57ms
step:1462/1530 train_loss:3.2606 train_time:252024ms step_avg:173.57ms
step:1463/1530 train_loss:3.4670 train_time:252206ms step_avg:173.58ms
step:1464/1530 train_loss:3.3580 train_time:252383ms step_avg:173.58ms
step:1465/1530 train_loss:3.3571 train_time:252563ms step_avg:173.58ms
step:1466/1530 train_loss:3.2840 train_time:252739ms step_avg:173.58ms
step:1467/1530 train_loss:3.3980 train_time:252919ms step_avg:173.59ms
step:1468/1530 train_loss:3.2864 train_time:253096ms step_avg:173.59ms
step:1469/1530 train_loss:3.2714 train_time:253276ms step_avg:173.60ms
step:1470/1530 train_loss:3.3271 train_time:253458ms step_avg:173.60ms
step:1471/1530 train_loss:3.2553 train_time:253644ms step_avg:173.61ms
step:1472/1530 train_loss:3.2465 train_time:253828ms step_avg:173.62ms
step:1473/1530 train_loss:3.4408 train_time:254006ms step_avg:173.62ms
step:1474/1530 train_loss:3.3096 train_time:254193ms step_avg:173.63ms
step:1475/1530 train_loss:3.1485 train_time:254376ms step_avg:173.64ms
step:1476/1530 train_loss:3.2666 train_time:254555ms step_avg:173.64ms
step:1477/1530 train_loss:3.2385 train_time:254740ms step_avg:173.65ms
step:1478/1530 train_loss:3.3081 train_time:254924ms step_avg:173.65ms
step:1479/1530 train_loss:3.3930 train_time:255105ms step_avg:173.66ms
step:1480/1530 train_loss:3.2696 train_time:255283ms step_avg:173.66ms
step:1481/1530 train_loss:3.4497 train_time:255464ms step_avg:173.67ms
step:1482/1530 train_loss:3.3641 train_time:255653ms step_avg:173.68ms
step:1483/1530 train_loss:3.2746 train_time:255844ms step_avg:173.69ms
step:1484/1530 train_loss:3.2636 train_time:256032ms step_avg:173.70ms
step:1485/1530 train_loss:3.2825 train_time:256213ms step_avg:173.70ms
step:1486/1530 train_loss:3.2278 train_time:256398ms step_avg:173.71ms
step:1487/1530 train_loss:3.3399 train_time:256579ms step_avg:173.72ms
step:1488/1530 train_loss:3.2430 train_time:256762ms step_avg:173.72ms
step:1489/1530 train_loss:3.3118 train_time:256942ms step_avg:173.73ms
step:1490/1530 train_loss:3.2501 train_time:257123ms step_avg:173.73ms
step:1491/1530 train_loss:3.1580 train_time:257302ms step_avg:173.74ms
step:1492/1530 train_loss:3.2667 train_time:257484ms step_avg:173.74ms
step:1493/1530 train_loss:3.4294 train_time:257662ms step_avg:173.74ms
step:1494/1530 train_loss:3.2947 train_time:257838ms step_avg:173.75ms
step:1495/1530 train_loss:3.0286 train_time:258022ms step_avg:173.75ms
step:1496/1530 train_loss:3.3583 train_time:258204ms step_avg:173.76ms
step:1497/1530 train_loss:3.3127 train_time:258393ms step_avg:173.77ms
step:1498/1530 train_loss:3.3457 train_time:258578ms step_avg:173.78ms
step:1499/1530 train_loss:3.3146 train_time:258765ms step_avg:173.78ms
step:1500/1530 train_loss:3.2945 train_time:258959ms step_avg:173.80ms
step:1500/1530 val_loss:3.2783 train_time:259015ms step_avg:173.84ms
step:1501/1530 train_loss:3.0892 train_time:259151ms step_avg:173.81ms
step:1502/1530 train_loss:3.3588 train_time:259342ms step_avg:173.82ms
step:1503/1530 train_loss:3.2390 train_time:259520ms step_avg:173.82ms
step:1504/1530 train_loss:3.2475 train_time:259702ms step_avg:173.83ms
step:1505/1530 train_loss:3.2108 train_time:259880ms step_avg:173.83ms
step:1506/1530 train_loss:3.2809 train_time:260064ms step_avg:173.84ms
step:1507/1530 train_loss:3.1729 train_time:260258ms step_avg:173.85ms
step:1508/1530 train_loss:3.4836 train_time:260440ms step_avg:173.86ms
step:1509/1530 train_loss:3.2810 train_time:260617ms step_avg:173.86ms
step:1510/1530 train_loss:3.2729 train_time:260797ms step_avg:173.86ms
step:1511/1530 train_loss:3.4178 train_time:261117ms step_avg:173.96ms
step:1512/1530 train_loss:3.4179 train_time:261303ms step_avg:173.97ms
step:1513/1530 train_loss:3.2667 train_time:261488ms step_avg:173.98ms
step:1514/1530 train_loss:3.0852 train_time:261672ms step_avg:173.98ms
step:1515/1530 train_loss:3.2372 train_time:261854ms step_avg:173.99ms
step:1516/1530 train_loss:3.2554 train_time:262039ms step_avg:174.00ms
step:1517/1530 train_loss:3.3028 train_time:262221ms step_avg:174.00ms
step:1518/1530 train_loss:3.2069 train_time:262403ms step_avg:174.01ms
step:1519/1530 train_loss:3.5082 train_time:262740ms step_avg:174.12ms
step:1520/1530 train_loss:3.1249 train_time:262920ms step_avg:174.12ms
step:1521/1530 train_loss:3.2058 train_time:263097ms step_avg:174.12ms
step:1522/1530 train_loss:3.3587 train_time:263282ms step_avg:174.13ms
step:1523/1530 train_loss:3.2291 train_time:263460ms step_avg:174.13ms
step:1524/1530 train_loss:3.3420 train_time:263641ms step_avg:174.14ms
step:1525/1530 train_loss:3.3376 train_time:263828ms step_avg:174.14ms
step:1526/1530 train_loss:3.2767 train_time:264018ms step_avg:174.15ms
step:1527/1530 train_loss:3.2918 train_time:264199ms step_avg:174.16ms
step:1528/1530 train_loss:3.4082 train_time:264378ms step_avg:174.16ms
step:1529/1530 train_loss:3.4079 train_time:264557ms step_avg:174.17ms
step:1530/1530 train_loss:3.2303 train_time:264736ms step_avg:174.17ms
step:1530/1530 val_loss:3.2758 train_time:264790ms step_avg:174.20ms