8bd08106-6eb1-4cd1-9779-ccf1192bda1b.txt

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

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

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

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

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

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

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

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

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

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

    def step(self):

        for group in self.param_groups:

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

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

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

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

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

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

class CastedLinear(nn.Linear):

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

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

class Rotary(torch.nn.Module):

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

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

class CausalSelfAttention(nn.Module):

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

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

class MLP(nn.Module):

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

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

class Block(nn.Module):

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

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

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

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

class GPT(nn.Module):

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

        self.reset()

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

# -------------------------------------------------------------------------
# clean up nice
dist.destroy_process_group()
====================================================================================================
Running pytorch 2.6.0.dev20241203+cu124 compiled for CUDA 12.4
nvidia-smi:
Thu Dec  5 03:22:32 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.06             Driver Version: 535.183.06   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA H100 80GB HBM3          On  | 00000000:19:00.0 Off |                    0 |
| N/A   39C    P0              75W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          On  | 00000000:3B:00.0 Off |                    0 |
| N/A   30C    P0              91W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          On  | 00000000:4C:00.0 Off |                    0 |
| N/A   31C    P0             117W / 700W |    529MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   38C    P0             118W / 700W |     39MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   39C    P0             123W / 700W |     43MiB / 81559MiB |      1%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   30C    P0             110W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   39C    P0             127W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   30C    P0             111W / 700W |     23MiB / 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:31706ms step_avg:nanms
step:2/1530 train_loss:10.0695 train_time:31819ms step_avg:nanms
step:3/1530 train_loss:8.3650 train_time:31978ms step_avg:nanms
step:4/1530 train_loss:7.5826 train_time:32140ms step_avg:nanms
step:5/1530 train_loss:7.4873 train_time:32300ms step_avg:nanms
step:6/1530 train_loss:6.9794 train_time:32461ms step_avg:nanms
step:7/1530 train_loss:7.2095 train_time:32621ms step_avg:nanms
step:8/1530 train_loss:6.7352 train_time:32781ms step_avg:nanms
step:9/1530 train_loss:6.6229 train_time:32942ms step_avg:nanms
step:10/1530 train_loss:6.4962 train_time:33103ms step_avg:nanms
step:11/1530 train_loss:6.4808 train_time:115ms step_avg:nanms
step:12/1530 train_loss:6.3722 train_time:276ms step_avg:nanms
step:13/1530 train_loss:6.2568 train_time:435ms step_avg:145.06ms
step:14/1530 train_loss:6.1783 train_time:597ms step_avg:149.20ms
step:15/1530 train_loss:6.1548 train_time:757ms step_avg:151.40ms
step:16/1530 train_loss:6.0964 train_time:918ms step_avg:152.98ms
step:17/1530 train_loss:6.1749 train_time:1078ms step_avg:154.01ms
step:18/1530 train_loss:5.9492 train_time:1238ms step_avg:154.80ms
step:19/1530 train_loss:6.0128 train_time:1399ms step_avg:155.42ms
step:20/1530 train_loss:5.6703 train_time:1559ms step_avg:155.93ms
step:21/1530 train_loss:5.9501 train_time:1720ms step_avg:156.38ms
step:22/1530 train_loss:6.1783 train_time:1881ms step_avg:156.74ms
step:23/1530 train_loss:5.8536 train_time:2040ms step_avg:156.95ms
step:24/1530 train_loss:6.0183 train_time:2200ms step_avg:157.18ms
step:25/1530 train_loss:5.6858 train_time:2361ms step_avg:157.38ms
step:26/1530 train_loss:5.5864 train_time:2521ms step_avg:157.54ms
step:27/1530 train_loss:5.8094 train_time:2682ms step_avg:157.77ms
step:28/1530 train_loss:5.3988 train_time:2842ms step_avg:157.90ms
step:29/1530 train_loss:5.6748 train_time:3002ms step_avg:157.98ms
step:30/1530 train_loss:5.4706 train_time:3162ms step_avg:158.10ms
step:31/1530 train_loss:5.4448 train_time:3322ms step_avg:158.17ms
step:32/1530 train_loss:5.2773 train_time:3483ms step_avg:158.33ms
step:33/1530 train_loss:5.5785 train_time:3644ms step_avg:158.42ms
step:34/1530 train_loss:5.4916 train_time:3802ms step_avg:158.44ms
step:35/1530 train_loss:5.6186 train_time:3963ms step_avg:158.54ms
step:36/1530 train_loss:5.5537 train_time:4124ms step_avg:158.60ms
step:37/1530 train_loss:5.4556 train_time:4283ms step_avg:158.64ms
step:38/1530 train_loss:5.2937 train_time:4444ms step_avg:158.72ms
step:39/1530 train_loss:5.3159 train_time:4604ms step_avg:158.75ms
step:40/1530 train_loss:5.2402 train_time:4763ms step_avg:158.78ms
step:41/1530 train_loss:5.2398 train_time:4923ms step_avg:158.79ms
step:42/1530 train_loss:5.1697 train_time:5083ms step_avg:158.86ms
step:43/1530 train_loss:5.2727 train_time:5243ms step_avg:158.88ms
step:44/1530 train_loss:5.2296 train_time:5403ms step_avg:158.90ms
step:45/1530 train_loss:5.3768 train_time:5563ms step_avg:158.94ms
step:46/1530 train_loss:5.1752 train_time:5723ms step_avg:158.96ms
step:47/1530 train_loss:5.0806 train_time:5883ms step_avg:159.00ms
step:48/1530 train_loss:5.2065 train_time:6042ms step_avg:158.99ms
step:49/1530 train_loss:5.1365 train_time:6203ms step_avg:159.04ms
step:50/1530 train_loss:5.2493 train_time:6363ms step_avg:159.08ms
step:51/1530 train_loss:5.1617 train_time:6523ms step_avg:159.10ms
step:52/1530 train_loss:5.0430 train_time:6685ms step_avg:159.16ms
step:53/1530 train_loss:5.1739 train_time:6845ms step_avg:159.18ms
step:54/1530 train_loss:5.0318 train_time:7005ms step_avg:159.20ms
step:55/1530 train_loss:5.4194 train_time:7164ms step_avg:159.20ms
step:56/1530 train_loss:5.0153 train_time:7324ms step_avg:159.21ms
step:57/1530 train_loss:4.8780 train_time:7484ms step_avg:159.23ms
step:58/1530 train_loss:5.0496 train_time:7644ms step_avg:159.24ms
step:59/1530 train_loss:5.0163 train_time:7803ms step_avg:159.25ms
step:60/1530 train_loss:5.1297 train_time:7963ms step_avg:159.27ms
step:61/1530 train_loss:4.8381 train_time:8124ms step_avg:159.28ms
step:62/1530 train_loss:4.9847 train_time:8283ms step_avg:159.28ms
step:63/1530 train_loss:4.9692 train_time:8443ms step_avg:159.31ms
step:64/1530 train_loss:4.9779 train_time:8604ms step_avg:159.33ms
step:65/1530 train_loss:4.7871 train_time:8764ms step_avg:159.34ms
step:66/1530 train_loss:4.9124 train_time:8924ms step_avg:159.35ms
step:67/1530 train_loss:4.8202 train_time:9084ms step_avg:159.37ms
step:68/1530 train_loss:5.1058 train_time:9244ms step_avg:159.38ms
step:69/1530 train_loss:4.7320 train_time:9403ms step_avg:159.37ms
step:70/1530 train_loss:4.8675 train_time:9564ms step_avg:159.39ms
step:71/1530 train_loss:4.9831 train_time:9724ms step_avg:159.41ms
step:72/1530 train_loss:4.8997 train_time:9884ms step_avg:159.42ms
step:73/1530 train_loss:4.7589 train_time:10043ms step_avg:159.42ms
step:74/1530 train_loss:4.9030 train_time:10203ms step_avg:159.42ms
step:75/1530 train_loss:4.8484 train_time:10364ms step_avg:159.44ms
step:76/1530 train_loss:4.7921 train_time:10523ms step_avg:159.44ms
step:77/1530 train_loss:4.9133 train_time:10684ms step_avg:159.47ms
step:78/1530 train_loss:5.1249 train_time:10845ms step_avg:159.48ms
step:79/1530 train_loss:4.8304 train_time:11003ms step_avg:159.47ms
step:80/1530 train_loss:4.8649 train_time:11164ms step_avg:159.49ms
step:81/1530 train_loss:4.6443 train_time:11324ms step_avg:159.49ms
step:82/1530 train_loss:4.8225 train_time:11484ms step_avg:159.50ms
step:83/1530 train_loss:4.7788 train_time:11644ms step_avg:159.51ms
step:84/1530 train_loss:4.7681 train_time:11804ms step_avg:159.52ms
step:85/1530 train_loss:4.6277 train_time:11965ms step_avg:159.53ms
step:86/1530 train_loss:4.8276 train_time:12126ms step_avg:159.55ms
step:87/1530 train_loss:4.7341 train_time:12286ms step_avg:159.55ms
step:88/1530 train_loss:4.7232 train_time:12446ms step_avg:159.56ms
step:89/1530 train_loss:4.7020 train_time:12606ms step_avg:159.56ms
step:90/1530 train_loss:4.6362 train_time:12765ms step_avg:159.56ms
step:91/1530 train_loss:4.6090 train_time:12925ms step_avg:159.57ms
step:92/1530 train_loss:4.7734 train_time:13085ms step_avg:159.58ms
step:93/1530 train_loss:4.6025 train_time:13245ms step_avg:159.58ms
step:94/1530 train_loss:4.6117 train_time:13405ms step_avg:159.59ms
step:95/1530 train_loss:4.6630 train_time:13566ms step_avg:159.60ms
step:96/1530 train_loss:4.5776 train_time:13726ms step_avg:159.60ms
step:97/1530 train_loss:4.6392 train_time:13888ms step_avg:159.63ms
step:98/1530 train_loss:4.5762 train_time:14048ms step_avg:159.64ms
step:99/1530 train_loss:4.6564 train_time:14208ms step_avg:159.64ms
step:100/1530 train_loss:4.6702 train_time:14370ms step_avg:159.66ms
step:101/1530 train_loss:4.5201 train_time:14530ms step_avg:159.67ms
step:102/1530 train_loss:4.6891 train_time:14691ms step_avg:159.68ms
step:103/1530 train_loss:4.5565 train_time:14852ms step_avg:159.70ms
step:104/1530 train_loss:4.5218 train_time:15013ms step_avg:159.71ms
step:105/1530 train_loss:4.5466 train_time:15174ms step_avg:159.73ms
step:106/1530 train_loss:4.5873 train_time:15333ms step_avg:159.72ms
step:107/1530 train_loss:4.4908 train_time:15494ms step_avg:159.74ms
step:108/1530 train_loss:4.3540 train_time:15655ms step_avg:159.75ms
step:109/1530 train_loss:4.4774 train_time:15816ms step_avg:159.75ms
step:110/1530 train_loss:4.4681 train_time:15977ms step_avg:159.77ms
step:111/1530 train_loss:4.4148 train_time:16137ms step_avg:159.77ms
step:112/1530 train_loss:4.5806 train_time:16298ms step_avg:159.78ms
step:113/1530 train_loss:4.4858 train_time:16459ms step_avg:159.79ms
step:114/1530 train_loss:4.3467 train_time:16618ms step_avg:159.79ms
step:115/1530 train_loss:4.4837 train_time:16783ms step_avg:159.83ms
step:116/1530 train_loss:4.4563 train_time:16945ms step_avg:159.86ms
step:117/1530 train_loss:4.3615 train_time:17109ms step_avg:159.90ms
step:118/1530 train_loss:4.5720 train_time:17275ms step_avg:159.95ms
step:119/1530 train_loss:4.4504 train_time:17439ms step_avg:159.99ms
step:120/1530 train_loss:4.3270 train_time:17603ms step_avg:160.03ms
step:121/1530 train_loss:4.2861 train_time:17767ms step_avg:160.06ms
step:122/1530 train_loss:4.4443 train_time:17930ms step_avg:160.09ms
step:123/1530 train_loss:4.2823 train_time:18095ms step_avg:160.13ms
step:124/1530 train_loss:4.5784 train_time:18258ms step_avg:160.16ms
step:125/1530 train_loss:4.4341 train_time:18422ms step_avg:160.19ms
step:125/1530 val_loss:4.3858 train_time:18469ms step_avg:160.60ms
step:126/1530 train_loss:4.3966 train_time:18586ms step_avg:160.22ms
step:127/1530 train_loss:4.4236 train_time:18752ms step_avg:160.27ms
step:128/1530 train_loss:4.3833 train_time:18916ms step_avg:160.30ms
step:129/1530 train_loss:4.6825 train_time:19079ms step_avg:160.33ms
step:130/1530 train_loss:4.3695 train_time:19243ms step_avg:160.36ms
step:131/1530 train_loss:4.3905 train_time:19407ms step_avg:160.39ms
step:132/1530 train_loss:4.3375 train_time:19570ms step_avg:160.41ms
step:133/1530 train_loss:4.4355 train_time:19734ms step_avg:160.44ms
step:134/1530 train_loss:4.2547 train_time:19897ms step_avg:160.46ms
step:135/1530 train_loss:4.4324 train_time:20061ms step_avg:160.48ms
step:136/1530 train_loss:4.2040 train_time:20225ms step_avg:160.52ms
step:137/1530 train_loss:4.3606 train_time:20390ms step_avg:160.55ms
step:138/1530 train_loss:4.2800 train_time:20554ms step_avg:160.58ms
step:139/1530 train_loss:4.3780 train_time:20719ms step_avg:160.61ms
step:140/1530 train_loss:4.4676 train_time:20883ms step_avg:160.63ms
step:141/1530 train_loss:4.3145 train_time:21045ms step_avg:160.65ms
step:142/1530 train_loss:4.2933 train_time:21209ms step_avg:160.68ms
step:143/1530 train_loss:4.2490 train_time:21373ms step_avg:160.70ms
step:144/1530 train_loss:4.3433 train_time:21537ms step_avg:160.72ms
step:145/1530 train_loss:4.2958 train_time:21702ms step_avg:160.75ms
step:146/1530 train_loss:4.1715 train_time:21866ms step_avg:160.78ms
step:147/1530 train_loss:4.3129 train_time:22029ms step_avg:160.80ms
step:148/1530 train_loss:4.3527 train_time:22193ms step_avg:160.82ms
step:149/1530 train_loss:4.2974 train_time:22357ms step_avg:160.84ms
step:150/1530 train_loss:4.4400 train_time:22520ms step_avg:160.86ms
step:151/1530 train_loss:4.2568 train_time:22685ms step_avg:160.89ms
step:152/1530 train_loss:4.2633 train_time:22851ms step_avg:160.92ms
step:153/1530 train_loss:4.3576 train_time:23015ms step_avg:160.95ms
step:154/1530 train_loss:4.3540 train_time:23179ms step_avg:160.96ms
step:155/1530 train_loss:4.2570 train_time:23342ms step_avg:160.98ms
step:156/1530 train_loss:4.3371 train_time:23507ms step_avg:161.01ms
step:157/1530 train_loss:4.4100 train_time:23671ms step_avg:161.03ms
step:158/1530 train_loss:4.2476 train_time:23836ms step_avg:161.06ms
step:159/1530 train_loss:4.2935 train_time:23999ms step_avg:161.07ms
step:160/1530 train_loss:4.1208 train_time:24165ms step_avg:161.10ms
step:161/1530 train_loss:4.3419 train_time:24329ms step_avg:161.12ms
step:162/1530 train_loss:4.3557 train_time:24493ms step_avg:161.13ms
step:163/1530 train_loss:4.3300 train_time:24655ms step_avg:161.15ms
step:164/1530 train_loss:4.1894 train_time:24819ms step_avg:161.16ms
step:165/1530 train_loss:4.2768 train_time:24983ms step_avg:161.18ms
step:166/1530 train_loss:4.3315 train_time:25146ms step_avg:161.19ms
step:167/1530 train_loss:4.1866 train_time:25310ms step_avg:161.21ms
step:168/1530 train_loss:4.2830 train_time:25474ms step_avg:161.23ms
step:169/1530 train_loss:4.1534 train_time:25637ms step_avg:161.24ms
step:170/1530 train_loss:4.0095 train_time:25801ms step_avg:161.26ms
step:171/1530 train_loss:4.1937 train_time:25964ms step_avg:161.27ms
step:172/1530 train_loss:4.2087 train_time:26127ms step_avg:161.28ms
step:173/1530 train_loss:4.2580 train_time:26290ms step_avg:161.29ms
step:174/1530 train_loss:4.4080 train_time:26453ms step_avg:161.30ms
step:175/1530 train_loss:4.2392 train_time:26615ms step_avg:161.31ms
step:176/1530 train_loss:4.0918 train_time:26778ms step_avg:161.31ms
step:177/1530 train_loss:4.0596 train_time:26940ms step_avg:161.32ms
step:178/1530 train_loss:4.1724 train_time:27103ms step_avg:161.33ms
step:179/1530 train_loss:4.1160 train_time:27265ms step_avg:161.33ms
step:180/1530 train_loss:4.1067 train_time:27428ms step_avg:161.34ms
step:181/1530 train_loss:4.2985 train_time:27590ms step_avg:161.35ms
step:182/1530 train_loss:4.1489 train_time:27754ms step_avg:161.36ms
step:183/1530 train_loss:4.1223 train_time:27917ms step_avg:161.37ms
step:184/1530 train_loss:4.1175 train_time:28080ms step_avg:161.38ms
step:185/1530 train_loss:4.1961 train_time:28242ms step_avg:161.38ms
step:186/1530 train_loss:4.1608 train_time:28405ms step_avg:161.39ms
step:187/1530 train_loss:4.2272 train_time:28568ms step_avg:161.40ms
step:188/1530 train_loss:4.1689 train_time:28863ms step_avg:162.15ms
step:189/1530 train_loss:4.1072 train_time:29195ms step_avg:163.10ms
step:190/1530 train_loss:4.2103 train_time:29359ms step_avg:163.10ms
step:191/1530 train_loss:4.0739 train_time:29522ms step_avg:163.10ms
step:192/1530 train_loss:4.0198 train_time:29685ms step_avg:163.10ms
step:193/1530 train_loss:4.2431 train_time:29849ms step_avg:163.11ms
step:194/1530 train_loss:4.1673 train_time:30011ms step_avg:163.10ms
step:195/1530 train_loss:4.3463 train_time:30173ms step_avg:163.10ms
step:196/1530 train_loss:4.1732 train_time:30336ms step_avg:163.10ms
step:197/1530 train_loss:4.0365 train_time:30499ms step_avg:163.10ms
step:198/1530 train_loss:4.1741 train_time:30662ms step_avg:163.10ms
step:199/1530 train_loss:4.0294 train_time:30825ms step_avg:163.09ms
step:200/1530 train_loss:4.1053 train_time:30988ms step_avg:163.09ms
step:201/1530 train_loss:3.9880 train_time:31151ms step_avg:163.09ms
step:202/1530 train_loss:4.2442 train_time:31313ms step_avg:163.09ms
step:203/1530 train_loss:4.0629 train_time:31475ms step_avg:163.09ms
step:204/1530 train_loss:4.1796 train_time:31639ms step_avg:163.09ms
step:205/1530 train_loss:4.2342 train_time:31801ms step_avg:163.08ms
step:206/1530 train_loss:3.9396 train_time:31963ms step_avg:163.08ms
step:207/1530 train_loss:4.0769 train_time:32126ms step_avg:163.08ms
step:208/1530 train_loss:4.0988 train_time:32290ms step_avg:163.08ms
step:209/1530 train_loss:4.2358 train_time:32451ms step_avg:163.07ms
step:210/1530 train_loss:4.1636 train_time:32615ms step_avg:163.07ms
step:211/1530 train_loss:4.0556 train_time:32777ms step_avg:163.07ms
step:212/1530 train_loss:4.1127 train_time:32940ms step_avg:163.07ms
step:213/1530 train_loss:4.0547 train_time:33103ms step_avg:163.07ms
step:214/1530 train_loss:4.1068 train_time:33266ms step_avg:163.07ms
step:215/1530 train_loss:3.9598 train_time:33428ms step_avg:163.06ms
step:216/1530 train_loss:3.9995 train_time:33591ms step_avg:163.06ms
step:217/1530 train_loss:4.0063 train_time:33754ms step_avg:163.06ms
step:218/1530 train_loss:4.0778 train_time:33916ms step_avg:163.06ms
step:219/1530 train_loss:4.0634 train_time:34079ms step_avg:163.06ms
step:220/1530 train_loss:4.0763 train_time:34243ms step_avg:163.06ms
step:221/1530 train_loss:4.0904 train_time:34405ms step_avg:163.06ms
step:222/1530 train_loss:3.9884 train_time:34568ms step_avg:163.06ms
step:223/1530 train_loss:3.9804 train_time:34731ms step_avg:163.06ms
step:224/1530 train_loss:4.2900 train_time:34894ms step_avg:163.06ms
step:225/1530 train_loss:3.9130 train_time:35058ms step_avg:163.06ms
step:226/1530 train_loss:3.9859 train_time:35221ms step_avg:163.06ms
step:227/1530 train_loss:3.9672 train_time:35384ms step_avg:163.06ms
step:228/1530 train_loss:4.1402 train_time:35548ms step_avg:163.07ms
step:229/1530 train_loss:3.9161 train_time:35715ms step_avg:163.08ms
step:230/1530 train_loss:4.0267 train_time:35880ms step_avg:163.09ms
step:231/1530 train_loss:3.8901 train_time:36046ms step_avg:163.11ms
step:232/1530 train_loss:3.9558 train_time:36212ms step_avg:163.12ms
step:233/1530 train_loss:4.0826 train_time:36377ms step_avg:163.12ms
step:234/1530 train_loss:4.0270 train_time:36544ms step_avg:163.14ms
step:235/1530 train_loss:3.8940 train_time:36711ms step_avg:163.16ms
step:236/1530 train_loss:4.0709 train_time:36877ms step_avg:163.17ms
step:237/1530 train_loss:4.0770 train_time:37043ms step_avg:163.19ms
step:238/1530 train_loss:3.9376 train_time:37209ms step_avg:163.20ms
step:239/1530 train_loss:4.0764 train_time:37375ms step_avg:163.21ms
step:240/1530 train_loss:4.1099 train_time:37542ms step_avg:163.22ms
step:241/1530 train_loss:3.9599 train_time:37707ms step_avg:163.23ms
step:242/1530 train_loss:4.1430 train_time:37873ms step_avg:163.24ms
step:243/1530 train_loss:4.0073 train_time:38038ms step_avg:163.25ms
step:244/1530 train_loss:4.0739 train_time:38204ms step_avg:163.27ms
step:245/1530 train_loss:4.1367 train_time:38370ms step_avg:163.28ms
step:246/1530 train_loss:4.0578 train_time:38536ms step_avg:163.29ms
step:247/1530 train_loss:4.0007 train_time:38701ms step_avg:163.30ms
step:248/1530 train_loss:4.0940 train_time:38867ms step_avg:163.31ms
step:249/1530 train_loss:3.9121 train_time:39033ms step_avg:163.32ms
step:250/1530 train_loss:3.9699 train_time:39199ms step_avg:163.33ms
step:250/1530 val_loss:4.0017 train_time:39247ms step_avg:163.53ms
step:251/1530 train_loss:4.0711 train_time:39366ms step_avg:163.34ms
step:252/1530 train_loss:4.1554 train_time:39535ms step_avg:163.37ms
step:253/1530 train_loss:3.9293 train_time:39701ms step_avg:163.38ms
step:254/1530 train_loss:3.8749 train_time:39867ms step_avg:163.39ms
step:255/1530 train_loss:4.0713 train_time:40033ms step_avg:163.40ms
step:256/1530 train_loss:3.9808 train_time:40199ms step_avg:163.41ms
step:257/1530 train_loss:3.9811 train_time:40364ms step_avg:163.42ms
step:258/1530 train_loss:3.9814 train_time:40531ms step_avg:163.43ms
step:259/1530 train_loss:4.0264 train_time:40696ms step_avg:163.44ms
step:260/1530 train_loss:4.0547 train_time:40862ms step_avg:163.45ms
step:261/1530 train_loss:4.0188 train_time:41030ms step_avg:163.47ms
step:262/1530 train_loss:3.9896 train_time:41196ms step_avg:163.48ms
step:263/1530 train_loss:3.8953 train_time:41361ms step_avg:163.48ms
step:264/1530 train_loss:3.9791 train_time:41529ms step_avg:163.50ms
step:265/1530 train_loss:3.8626 train_time:41696ms step_avg:163.51ms
step:266/1530 train_loss:3.9124 train_time:41861ms step_avg:163.52ms
step:267/1530 train_loss:3.9253 train_time:42027ms step_avg:163.53ms
step:268/1530 train_loss:3.9592 train_time:42192ms step_avg:163.54ms
step:269/1530 train_loss:3.8530 train_time:42357ms step_avg:163.54ms
step:270/1530 train_loss:4.0964 train_time:42523ms step_avg:163.55ms
step:271/1530 train_loss:3.9615 train_time:42689ms step_avg:163.56ms
step:272/1530 train_loss:3.9241 train_time:42855ms step_avg:163.57ms
step:273/1530 train_loss:3.9440 train_time:43021ms step_avg:163.58ms
step:274/1530 train_loss:4.0412 train_time:43188ms step_avg:163.59ms
step:275/1530 train_loss:4.0525 train_time:43353ms step_avg:163.60ms
step:276/1530 train_loss:4.2217 train_time:43520ms step_avg:163.61ms
step:277/1530 train_loss:4.0290 train_time:43686ms step_avg:163.62ms
step:278/1530 train_loss:4.0813 train_time:43851ms step_avg:163.62ms
step:279/1530 train_loss:3.9905 train_time:44017ms step_avg:163.63ms
step:280/1530 train_loss:4.1891 train_time:44184ms step_avg:163.64ms
step:281/1530 train_loss:3.9687 train_time:44350ms step_avg:163.65ms
step:282/1530 train_loss:3.9349 train_time:44516ms step_avg:163.66ms
step:283/1530 train_loss:3.9082 train_time:44682ms step_avg:163.67ms
step:284/1530 train_loss:4.0456 train_time:44849ms step_avg:163.68ms
step:285/1530 train_loss:4.0547 train_time:45014ms step_avg:163.69ms
step:286/1530 train_loss:4.0850 train_time:45179ms step_avg:163.69ms
step:287/1530 train_loss:3.8999 train_time:45344ms step_avg:163.70ms
step:288/1530 train_loss:4.0102 train_time:45510ms step_avg:163.70ms
step:289/1530 train_loss:3.8691 train_time:45675ms step_avg:163.71ms
step:290/1530 train_loss:3.8486 train_time:45839ms step_avg:163.71ms
step:291/1530 train_loss:3.9025 train_time:46007ms step_avg:163.72ms
step:292/1530 train_loss:3.8620 train_time:46171ms step_avg:163.73ms
step:293/1530 train_loss:3.8952 train_time:46336ms step_avg:163.73ms
step:294/1530 train_loss:3.9241 train_time:46502ms step_avg:163.74ms
step:295/1530 train_loss:3.8380 train_time:46667ms step_avg:163.74ms
step:296/1530 train_loss:3.8562 train_time:46833ms step_avg:163.75ms
step:297/1530 train_loss:3.8664 train_time:46998ms step_avg:163.75ms
step:298/1530 train_loss:3.9639 train_time:47163ms step_avg:163.76ms
step:299/1530 train_loss:3.8205 train_time:47328ms step_avg:163.76ms
step:300/1530 train_loss:3.9641 train_time:47493ms step_avg:163.77ms
step:301/1530 train_loss:3.9556 train_time:47658ms step_avg:163.77ms
step:302/1530 train_loss:3.9257 train_time:47823ms step_avg:163.78ms
step:303/1530 train_loss:3.9698 train_time:47989ms step_avg:163.78ms
step:304/1530 train_loss:3.9594 train_time:48153ms step_avg:163.79ms
step:305/1530 train_loss:4.4527 train_time:48319ms step_avg:163.79ms
step:306/1530 train_loss:3.9289 train_time:48484ms step_avg:163.80ms
step:307/1530 train_loss:3.8282 train_time:48650ms step_avg:163.80ms
step:308/1530 train_loss:3.9697 train_time:48814ms step_avg:163.81ms
step:309/1530 train_loss:3.8578 train_time:48979ms step_avg:163.81ms
step:310/1530 train_loss:4.0722 train_time:49145ms step_avg:163.82ms
step:311/1530 train_loss:3.9260 train_time:49310ms step_avg:163.82ms
step:312/1530 train_loss:3.8564 train_time:49475ms step_avg:163.82ms
step:313/1530 train_loss:3.9433 train_time:49640ms step_avg:163.83ms
step:314/1530 train_loss:4.0610 train_time:49807ms step_avg:163.84ms
step:315/1530 train_loss:3.9368 train_time:49972ms step_avg:163.84ms
step:316/1530 train_loss:3.7864 train_time:50137ms step_avg:163.85ms
step:317/1530 train_loss:3.8696 train_time:50303ms step_avg:163.85ms
step:318/1530 train_loss:3.9172 train_time:50469ms step_avg:163.86ms
step:319/1530 train_loss:3.8891 train_time:50634ms step_avg:163.86ms
step:320/1530 train_loss:4.0149 train_time:50799ms step_avg:163.87ms
step:321/1530 train_loss:3.9534 train_time:50964ms step_avg:163.87ms
step:322/1530 train_loss:3.9276 train_time:51130ms step_avg:163.88ms
step:323/1530 train_loss:4.0026 train_time:51295ms step_avg:163.88ms
step:324/1530 train_loss:3.9419 train_time:51460ms step_avg:163.88ms
step:325/1530 train_loss:4.0124 train_time:51625ms step_avg:163.89ms
step:326/1530 train_loss:3.8917 train_time:51791ms step_avg:163.90ms
step:327/1530 train_loss:4.3831 train_time:51956ms step_avg:163.90ms
step:328/1530 train_loss:4.0740 train_time:52121ms step_avg:163.90ms
step:329/1530 train_loss:3.7883 train_time:52285ms step_avg:163.90ms
step:330/1530 train_loss:3.7405 train_time:52451ms step_avg:163.91ms
step:331/1530 train_loss:3.9723 train_time:52616ms step_avg:163.91ms
step:332/1530 train_loss:3.9109 train_time:52781ms step_avg:163.92ms
step:333/1530 train_loss:3.8849 train_time:52947ms step_avg:163.92ms
step:334/1530 train_loss:3.8416 train_time:53112ms step_avg:163.93ms
step:335/1530 train_loss:4.0082 train_time:53277ms step_avg:163.93ms
step:336/1530 train_loss:3.9614 train_time:53442ms step_avg:163.93ms
step:337/1530 train_loss:4.4160 train_time:53608ms step_avg:163.94ms
step:338/1530 train_loss:3.9334 train_time:53773ms step_avg:163.94ms
step:339/1530 train_loss:3.8577 train_time:53938ms step_avg:163.95ms
step:340/1530 train_loss:3.9355 train_time:54104ms step_avg:163.95ms
step:341/1530 train_loss:3.8470 train_time:54271ms step_avg:163.96ms
step:342/1530 train_loss:3.8036 train_time:54439ms step_avg:163.97ms
step:343/1530 train_loss:3.8400 train_time:54609ms step_avg:163.99ms
step:344/1530 train_loss:3.9939 train_time:54776ms step_avg:164.00ms
step:345/1530 train_loss:3.8038 train_time:54945ms step_avg:164.01ms
step:346/1530 train_loss:3.7617 train_time:55114ms step_avg:164.03ms
step:347/1530 train_loss:3.7978 train_time:55282ms step_avg:164.04ms
step:348/1530 train_loss:3.8550 train_time:55450ms step_avg:164.05ms
step:349/1530 train_loss:3.8274 train_time:55617ms step_avg:164.06ms
step:350/1530 train_loss:3.5719 train_time:55786ms step_avg:164.08ms
step:351/1530 train_loss:3.8198 train_time:55953ms step_avg:164.09ms
step:352/1530 train_loss:4.1772 train_time:56121ms step_avg:164.10ms
step:353/1530 train_loss:3.6536 train_time:56290ms step_avg:164.11ms
step:354/1530 train_loss:3.9216 train_time:56456ms step_avg:164.12ms
step:355/1530 train_loss:3.7746 train_time:56626ms step_avg:164.13ms
step:356/1530 train_loss:3.8779 train_time:56795ms step_avg:164.15ms
step:357/1530 train_loss:3.7534 train_time:56963ms step_avg:164.16ms
step:358/1530 train_loss:3.8514 train_time:57131ms step_avg:164.17ms
step:359/1530 train_loss:3.7598 train_time:57301ms step_avg:164.19ms
step:360/1530 train_loss:3.4283 train_time:57471ms step_avg:164.20ms
step:361/1530 train_loss:4.0172 train_time:57640ms step_avg:164.22ms
step:362/1530 train_loss:3.9126 train_time:57810ms step_avg:164.23ms
step:363/1530 train_loss:3.8361 train_time:57976ms step_avg:164.24ms
step:364/1530 train_loss:3.7402 train_time:58145ms step_avg:164.25ms
step:365/1530 train_loss:3.9080 train_time:58313ms step_avg:164.26ms
step:366/1530 train_loss:3.8569 train_time:58480ms step_avg:164.27ms
step:367/1530 train_loss:3.8523 train_time:58648ms step_avg:164.28ms
step:368/1530 train_loss:3.8453 train_time:58815ms step_avg:164.29ms
step:369/1530 train_loss:3.7463 train_time:58983ms step_avg:164.30ms
step:370/1530 train_loss:3.8761 train_time:59150ms step_avg:164.31ms
step:371/1530 train_loss:3.7286 train_time:59318ms step_avg:164.32ms
step:372/1530 train_loss:3.6882 train_time:59488ms step_avg:164.33ms
step:373/1530 train_loss:3.9118 train_time:59654ms step_avg:164.34ms
step:374/1530 train_loss:3.8236 train_time:59822ms step_avg:164.35ms
step:375/1530 train_loss:3.7957 train_time:59991ms step_avg:164.36ms
step:375/1530 val_loss:3.8256 train_time:60038ms step_avg:164.49ms
step:376/1530 train_loss:3.8667 train_time:60160ms step_avg:164.37ms
step:377/1530 train_loss:3.7901 train_time:60461ms step_avg:164.75ms
step:378/1530 train_loss:3.8490 train_time:60640ms step_avg:164.78ms
step:379/1530 train_loss:3.8723 train_time:60958ms step_avg:165.20ms
step:380/1530 train_loss:3.9480 train_time:61128ms step_avg:165.21ms
step:381/1530 train_loss:3.8429 train_time:61295ms step_avg:165.22ms
step:382/1530 train_loss:3.7966 train_time:61463ms step_avg:165.22ms
step:383/1530 train_loss:3.7921 train_time:61631ms step_avg:165.23ms
step:384/1530 train_loss:3.8712 train_time:61798ms step_avg:165.23ms
step:385/1530 train_loss:3.7920 train_time:61965ms step_avg:165.24ms
step:386/1530 train_loss:3.8860 train_time:62132ms step_avg:165.25ms
step:387/1530 train_loss:4.0552 train_time:62299ms step_avg:165.25ms
step:388/1530 train_loss:3.7887 train_time:62466ms step_avg:165.26ms
step:389/1530 train_loss:3.7921 train_time:62636ms step_avg:165.27ms
step:390/1530 train_loss:3.8973 train_time:62803ms step_avg:165.27ms
step:391/1530 train_loss:3.8068 train_time:62972ms step_avg:165.28ms
step:392/1530 train_loss:3.9145 train_time:63139ms step_avg:165.29ms
step:393/1530 train_loss:3.7582 train_time:63307ms step_avg:165.29ms
step:394/1530 train_loss:3.8862 train_time:63475ms step_avg:165.30ms
step:395/1530 train_loss:3.6288 train_time:63642ms step_avg:165.30ms
step:396/1530 train_loss:3.8411 train_time:63810ms step_avg:165.31ms
step:397/1530 train_loss:3.8613 train_time:63977ms step_avg:165.31ms
step:398/1530 train_loss:3.8719 train_time:64145ms step_avg:165.32ms
step:399/1530 train_loss:3.7649 train_time:64311ms step_avg:165.32ms
step:400/1530 train_loss:3.8231 train_time:64479ms step_avg:165.33ms
step:401/1530 train_loss:3.9148 train_time:64646ms step_avg:165.33ms
step:402/1530 train_loss:3.8409 train_time:64814ms step_avg:165.34ms
step:403/1530 train_loss:3.9605 train_time:64981ms step_avg:165.35ms
step:404/1530 train_loss:3.6724 train_time:65148ms step_avg:165.35ms
step:405/1530 train_loss:3.7832 train_time:65315ms step_avg:165.35ms
step:406/1530 train_loss:4.0877 train_time:65482ms step_avg:165.36ms
step:407/1530 train_loss:3.7724 train_time:65651ms step_avg:165.37ms
step:408/1530 train_loss:3.8174 train_time:65817ms step_avg:165.37ms
step:409/1530 train_loss:3.8542 train_time:65984ms step_avg:165.37ms
step:410/1530 train_loss:3.7548 train_time:66151ms step_avg:165.38ms
step:411/1530 train_loss:3.7541 train_time:66318ms step_avg:165.38ms
step:412/1530 train_loss:4.1799 train_time:66485ms step_avg:165.39ms
step:413/1530 train_loss:3.7891 train_time:66653ms step_avg:165.39ms
step:414/1530 train_loss:4.0080 train_time:66820ms step_avg:165.40ms
step:415/1530 train_loss:3.7502 train_time:66986ms step_avg:165.40ms
step:416/1530 train_loss:3.7572 train_time:67153ms step_avg:165.40ms
step:417/1530 train_loss:3.9515 train_time:67320ms step_avg:165.41ms
step:418/1530 train_loss:3.6889 train_time:67487ms step_avg:165.41ms
step:419/1530 train_loss:3.8033 train_time:67656ms step_avg:165.42ms
step:420/1530 train_loss:3.7011 train_time:67822ms step_avg:165.42ms
step:421/1530 train_loss:3.6465 train_time:67989ms step_avg:165.42ms
step:422/1530 train_loss:3.7790 train_time:68157ms step_avg:165.43ms
step:423/1530 train_loss:3.8657 train_time:68323ms step_avg:165.43ms
step:424/1530 train_loss:3.6118 train_time:68491ms step_avg:165.44ms
step:425/1530 train_loss:3.7938 train_time:68659ms step_avg:165.44ms
step:426/1530 train_loss:3.6614 train_time:68826ms step_avg:165.45ms
step:427/1530 train_loss:3.8906 train_time:68994ms step_avg:165.45ms
step:428/1530 train_loss:3.8064 train_time:69161ms step_avg:165.46ms
step:429/1530 train_loss:3.7596 train_time:69328ms step_avg:165.46ms
step:430/1530 train_loss:3.7040 train_time:69495ms step_avg:165.46ms
step:431/1530 train_loss:3.6304 train_time:69662ms step_avg:165.47ms
step:432/1530 train_loss:3.7631 train_time:69830ms step_avg:165.47ms
step:433/1530 train_loss:3.8107 train_time:69997ms step_avg:165.48ms
step:434/1530 train_loss:3.7709 train_time:70164ms step_avg:165.48ms
step:435/1530 train_loss:3.8017 train_time:70332ms step_avg:165.49ms
step:436/1530 train_loss:3.8269 train_time:70498ms step_avg:165.49ms
step:437/1530 train_loss:3.7109 train_time:70665ms step_avg:165.49ms
step:438/1530 train_loss:3.7011 train_time:70834ms step_avg:165.50ms
step:439/1530 train_loss:3.7053 train_time:71000ms step_avg:165.50ms
step:440/1530 train_loss:3.8890 train_time:71168ms step_avg:165.51ms
step:441/1530 train_loss:3.7577 train_time:71337ms step_avg:165.51ms
step:442/1530 train_loss:3.7376 train_time:71503ms step_avg:165.52ms
step:443/1530 train_loss:3.6177 train_time:71670ms step_avg:165.52ms
step:444/1530 train_loss:3.9219 train_time:71838ms step_avg:165.52ms
step:445/1530 train_loss:3.8395 train_time:72003ms step_avg:165.53ms
step:446/1530 train_loss:3.8295 train_time:72170ms step_avg:165.53ms
step:447/1530 train_loss:3.7519 train_time:72338ms step_avg:165.53ms
step:448/1530 train_loss:3.8462 train_time:72506ms step_avg:165.54ms
step:449/1530 train_loss:3.6872 train_time:72674ms step_avg:165.54ms
step:450/1530 train_loss:3.7078 train_time:72841ms step_avg:165.55ms
step:451/1530 train_loss:3.5815 train_time:73007ms step_avg:165.55ms
step:452/1530 train_loss:3.7123 train_time:73175ms step_avg:165.55ms
step:453/1530 train_loss:3.6683 train_time:73342ms step_avg:165.56ms
step:454/1530 train_loss:3.6286 train_time:73509ms step_avg:165.56ms
step:455/1530 train_loss:3.8342 train_time:73677ms step_avg:165.57ms
step:456/1530 train_loss:3.7221 train_time:73846ms step_avg:165.57ms
step:457/1530 train_loss:3.7830 train_time:74016ms step_avg:165.58ms
step:458/1530 train_loss:3.8284 train_time:74185ms step_avg:165.59ms
step:459/1530 train_loss:3.6244 train_time:74358ms step_avg:165.61ms
step:460/1530 train_loss:3.7926 train_time:74527ms step_avg:165.62ms
step:461/1530 train_loss:3.6850 train_time:74697ms step_avg:165.63ms
step:462/1530 train_loss:3.7321 train_time:74865ms step_avg:165.63ms
step:463/1530 train_loss:3.7753 train_time:75037ms step_avg:165.65ms
step:464/1530 train_loss:3.7086 train_time:75206ms step_avg:165.65ms
step:465/1530 train_loss:3.7163 train_time:75376ms step_avg:165.66ms
step:466/1530 train_loss:3.7993 train_time:75545ms step_avg:165.67ms
step:467/1530 train_loss:3.8221 train_time:75717ms step_avg:165.68ms
step:468/1530 train_loss:3.7899 train_time:75885ms step_avg:165.69ms
step:469/1530 train_loss:3.6811 train_time:76057ms step_avg:165.70ms
step:470/1530 train_loss:3.7662 train_time:76225ms step_avg:165.71ms
step:471/1530 train_loss:3.8074 train_time:76396ms step_avg:165.72ms
step:472/1530 train_loss:3.7847 train_time:76565ms step_avg:165.72ms
step:473/1530 train_loss:3.7098 train_time:76735ms step_avg:165.74ms
step:474/1530 train_loss:3.5909 train_time:76904ms step_avg:165.74ms
step:475/1530 train_loss:4.0098 train_time:77074ms step_avg:165.75ms
step:476/1530 train_loss:3.7531 train_time:77244ms step_avg:165.76ms
step:477/1530 train_loss:3.6002 train_time:77416ms step_avg:165.77ms
step:478/1530 train_loss:3.8205 train_time:77585ms step_avg:165.78ms
step:479/1530 train_loss:3.7697 train_time:77757ms step_avg:165.79ms
step:480/1530 train_loss:3.9213 train_time:77926ms step_avg:165.80ms
step:481/1530 train_loss:3.7236 train_time:78096ms step_avg:165.81ms
step:482/1530 train_loss:3.5242 train_time:78264ms step_avg:165.81ms
step:483/1530 train_loss:3.7967 train_time:78435ms step_avg:165.82ms
step:484/1530 train_loss:3.6616 train_time:78604ms step_avg:165.83ms
step:485/1530 train_loss:3.6546 train_time:78774ms step_avg:165.84ms
step:486/1530 train_loss:3.5665 train_time:78944ms step_avg:165.85ms
step:487/1530 train_loss:3.6795 train_time:79113ms step_avg:165.86ms
step:488/1530 train_loss:3.8754 train_time:79283ms step_avg:165.86ms
step:489/1530 train_loss:3.7070 train_time:79454ms step_avg:165.88ms
step:490/1530 train_loss:3.5876 train_time:79623ms step_avg:165.88ms
step:491/1530 train_loss:3.6158 train_time:79793ms step_avg:165.89ms
step:492/1530 train_loss:3.7315 train_time:79963ms step_avg:165.90ms
step:493/1530 train_loss:3.5716 train_time:80135ms step_avg:165.91ms
step:494/1530 train_loss:3.6938 train_time:80303ms step_avg:165.92ms
step:495/1530 train_loss:3.6590 train_time:80474ms step_avg:165.93ms
step:496/1530 train_loss:3.5102 train_time:80644ms step_avg:165.93ms
step:497/1530 train_loss:3.7349 train_time:80813ms step_avg:165.94ms
step:498/1530 train_loss:3.7842 train_time:80983ms step_avg:165.95ms
step:499/1530 train_loss:3.8191 train_time:81154ms step_avg:165.96ms
step:500/1530 train_loss:3.7254 train_time:81323ms step_avg:165.97ms
step:500/1530 val_loss:3.7022 train_time:81372ms step_avg:166.07ms
step:501/1530 train_loss:3.8026 train_time:81493ms step_avg:165.97ms
step:502/1530 train_loss:3.7428 train_time:81666ms step_avg:165.99ms
step:503/1530 train_loss:3.7734 train_time:81836ms step_avg:166.00ms
step:504/1530 train_loss:3.7140 train_time:82004ms step_avg:166.00ms
step:505/1530 train_loss:3.8016 train_time:82175ms step_avg:166.01ms
step:506/1530 train_loss:3.6413 train_time:82343ms step_avg:166.01ms
step:507/1530 train_loss:3.7575 train_time:82512ms step_avg:166.02ms
step:508/1530 train_loss:3.8224 train_time:82684ms step_avg:166.03ms
step:509/1530 train_loss:3.7666 train_time:82853ms step_avg:166.04ms
step:510/1530 train_loss:3.5781 train_time:83022ms step_avg:166.04ms
step:511/1530 train_loss:3.7753 train_time:83191ms step_avg:166.05ms
step:512/1530 train_loss:3.7180 train_time:83363ms step_avg:166.06ms
step:513/1530 train_loss:3.6657 train_time:83531ms step_avg:166.06ms
step:514/1530 train_loss:3.8516 train_time:83701ms step_avg:166.07ms
step:515/1530 train_loss:3.7326 train_time:83871ms step_avg:166.08ms
step:516/1530 train_loss:4.0685 train_time:84040ms step_avg:166.09ms
step:517/1530 train_loss:3.6851 train_time:84207ms step_avg:166.09ms
step:518/1530 train_loss:3.7663 train_time:84378ms step_avg:166.10ms
step:519/1530 train_loss:3.6554 train_time:84547ms step_avg:166.10ms
step:520/1530 train_loss:3.6808 train_time:84718ms step_avg:166.11ms
step:521/1530 train_loss:3.6588 train_time:84887ms step_avg:166.12ms
step:522/1530 train_loss:3.6542 train_time:85057ms step_avg:166.13ms
step:523/1530 train_loss:4.2795 train_time:85226ms step_avg:166.13ms
step:524/1530 train_loss:3.7348 train_time:85394ms step_avg:166.14ms
step:525/1530 train_loss:3.6757 train_time:85562ms step_avg:166.14ms
step:526/1530 train_loss:3.6929 train_time:85731ms step_avg:166.15ms
step:527/1530 train_loss:3.6562 train_time:85901ms step_avg:166.15ms
step:528/1530 train_loss:3.6224 train_time:86069ms step_avg:166.16ms
step:529/1530 train_loss:3.8495 train_time:86237ms step_avg:166.16ms
step:530/1530 train_loss:3.6456 train_time:86407ms step_avg:166.17ms
step:531/1530 train_loss:3.9156 train_time:86577ms step_avg:166.18ms
step:532/1530 train_loss:3.7320 train_time:86746ms step_avg:166.18ms
step:533/1530 train_loss:3.6514 train_time:86917ms step_avg:166.19ms
step:534/1530 train_loss:3.6643 train_time:87085ms step_avg:166.19ms
step:535/1530 train_loss:3.6085 train_time:87255ms step_avg:166.20ms
step:536/1530 train_loss:3.7487 train_time:87425ms step_avg:166.21ms
step:537/1530 train_loss:3.7241 train_time:87596ms step_avg:166.22ms
step:538/1530 train_loss:3.6212 train_time:87766ms step_avg:166.22ms
step:539/1530 train_loss:4.1069 train_time:87938ms step_avg:166.23ms
step:540/1530 train_loss:3.6720 train_time:88106ms step_avg:166.24ms
step:541/1530 train_loss:3.7812 train_time:88276ms step_avg:166.24ms
step:542/1530 train_loss:3.5822 train_time:88444ms step_avg:166.25ms
step:543/1530 train_loss:3.5848 train_time:88614ms step_avg:166.25ms
step:544/1530 train_loss:3.6369 train_time:88783ms step_avg:166.26ms
step:545/1530 train_loss:3.5876 train_time:88951ms step_avg:166.26ms
step:546/1530 train_loss:3.6210 train_time:89120ms step_avg:166.27ms
step:547/1530 train_loss:3.6361 train_time:89289ms step_avg:166.27ms
step:548/1530 train_loss:3.6007 train_time:89459ms step_avg:166.28ms
step:549/1530 train_loss:3.7222 train_time:89627ms step_avg:166.28ms
step:550/1530 train_loss:3.6137 train_time:89798ms step_avg:166.29ms
step:551/1530 train_loss:3.6261 train_time:89965ms step_avg:166.29ms
step:552/1530 train_loss:3.9316 train_time:90135ms step_avg:166.30ms
step:553/1530 train_loss:3.7596 train_time:90303ms step_avg:166.30ms
step:554/1530 train_loss:3.7083 train_time:90472ms step_avg:166.31ms
step:555/1530 train_loss:3.6303 train_time:90640ms step_avg:166.31ms
step:556/1530 train_loss:3.6970 train_time:90808ms step_avg:166.31ms
step:557/1530 train_loss:3.3049 train_time:90978ms step_avg:166.32ms
step:558/1530 train_loss:3.6109 train_time:91146ms step_avg:166.33ms
step:559/1530 train_loss:3.6476 train_time:91316ms step_avg:166.33ms
step:560/1530 train_loss:3.6899 train_time:91484ms step_avg:166.33ms
step:561/1530 train_loss:3.6113 train_time:91652ms step_avg:166.34ms
step:562/1530 train_loss:3.5550 train_time:91821ms step_avg:166.34ms
step:563/1530 train_loss:3.7546 train_time:91990ms step_avg:166.35ms
step:564/1530 train_loss:3.5693 train_time:92160ms step_avg:166.35ms
step:565/1530 train_loss:3.6759 train_time:92328ms step_avg:166.36ms
step:566/1530 train_loss:3.6174 train_time:92630ms step_avg:166.60ms
step:567/1530 train_loss:3.5968 train_time:92810ms step_avg:166.62ms
step:568/1530 train_loss:3.6873 train_time:92982ms step_avg:166.63ms
step:569/1530 train_loss:3.6450 train_time:93302ms step_avg:166.91ms
step:570/1530 train_loss:3.6860 train_time:93474ms step_avg:166.92ms
step:571/1530 train_loss:3.7574 train_time:93644ms step_avg:166.92ms
step:572/1530 train_loss:3.7257 train_time:93817ms step_avg:166.93ms
step:573/1530 train_loss:3.7332 train_time:93988ms step_avg:166.94ms
step:574/1530 train_loss:3.7786 train_time:94162ms step_avg:166.95ms
step:575/1530 train_loss:3.7258 train_time:94333ms step_avg:166.96ms
step:576/1530 train_loss:3.7552 train_time:94503ms step_avg:166.97ms
step:577/1530 train_loss:3.6646 train_time:94676ms step_avg:166.98ms
step:578/1530 train_loss:3.6713 train_time:94847ms step_avg:166.98ms
step:579/1530 train_loss:3.6667 train_time:95019ms step_avg:166.99ms
step:580/1530 train_loss:3.5838 train_time:95188ms step_avg:167.00ms
step:581/1530 train_loss:3.6355 train_time:95360ms step_avg:167.00ms
step:582/1530 train_loss:3.8501 train_time:95530ms step_avg:167.01ms
step:583/1530 train_loss:3.6253 train_time:95702ms step_avg:167.02ms
step:584/1530 train_loss:3.5861 train_time:95873ms step_avg:167.03ms
step:585/1530 train_loss:3.7863 train_time:96043ms step_avg:167.03ms
step:586/1530 train_loss:3.5195 train_time:96215ms step_avg:167.04ms
step:587/1530 train_loss:3.6699 train_time:96386ms step_avg:167.05ms
step:588/1530 train_loss:3.6337 train_time:96557ms step_avg:167.05ms
step:589/1530 train_loss:3.9954 train_time:96728ms step_avg:167.06ms
step:590/1530 train_loss:3.7790 train_time:96901ms step_avg:167.07ms
step:591/1530 train_loss:3.5079 train_time:97073ms step_avg:167.08ms
step:592/1530 train_loss:3.5305 train_time:97245ms step_avg:167.09ms
step:593/1530 train_loss:3.4944 train_time:97418ms step_avg:167.10ms
step:594/1530 train_loss:3.5493 train_time:97589ms step_avg:167.10ms
step:595/1530 train_loss:3.9134 train_time:97762ms step_avg:167.11ms
step:596/1530 train_loss:3.6508 train_time:97934ms step_avg:167.12ms
step:597/1530 train_loss:3.5843 train_time:98105ms step_avg:167.13ms
step:598/1530 train_loss:3.6535 train_time:98276ms step_avg:167.14ms
step:599/1530 train_loss:3.4777 train_time:98446ms step_avg:167.14ms
step:600/1530 train_loss:3.5904 train_time:98617ms step_avg:167.15ms
step:601/1530 train_loss:3.6402 train_time:98789ms step_avg:167.16ms
step:602/1530 train_loss:3.6667 train_time:98961ms step_avg:167.16ms
step:603/1530 train_loss:3.7849 train_time:99131ms step_avg:167.17ms
step:604/1530 train_loss:3.6079 train_time:99302ms step_avg:167.17ms
step:605/1530 train_loss:3.6124 train_time:99475ms step_avg:167.18ms
step:606/1530 train_loss:3.5702 train_time:99647ms step_avg:167.19ms
step:607/1530 train_loss:3.8307 train_time:99820ms step_avg:167.20ms
step:608/1530 train_loss:3.6316 train_time:99992ms step_avg:167.21ms
step:609/1530 train_loss:3.6184 train_time:100163ms step_avg:167.22ms
step:610/1530 train_loss:3.6990 train_time:100333ms step_avg:167.22ms
step:611/1530 train_loss:3.5967 train_time:100504ms step_avg:167.23ms
step:612/1530 train_loss:3.5703 train_time:100675ms step_avg:167.23ms
step:613/1530 train_loss:3.7634 train_time:100846ms step_avg:167.24ms
step:614/1530 train_loss:3.7017 train_time:101018ms step_avg:167.25ms
step:615/1530 train_loss:3.6968 train_time:101189ms step_avg:167.25ms
step:616/1530 train_loss:3.6319 train_time:101359ms step_avg:167.26ms
step:617/1530 train_loss:3.5544 train_time:101530ms step_avg:167.26ms
step:618/1530 train_loss:3.6837 train_time:101701ms step_avg:167.27ms
step:619/1530 train_loss:3.5467 train_time:101871ms step_avg:167.28ms
step:620/1530 train_loss:3.5841 train_time:102042ms step_avg:167.28ms
step:621/1530 train_loss:3.9224 train_time:102214ms step_avg:167.29ms
step:622/1530 train_loss:3.5662 train_time:102386ms step_avg:167.30ms
step:623/1530 train_loss:3.5955 train_time:102560ms step_avg:167.31ms
step:624/1530 train_loss:3.6938 train_time:102731ms step_avg:167.31ms
step:625/1530 train_loss:3.6985 train_time:102902ms step_avg:167.32ms
step:625/1530 val_loss:3.6198 train_time:102952ms step_avg:167.40ms
step:626/1530 train_loss:3.7318 train_time:103074ms step_avg:167.33ms
step:627/1530 train_loss:3.7079 train_time:103246ms step_avg:167.34ms
step:628/1530 train_loss:3.7603 train_time:103416ms step_avg:167.34ms
step:629/1530 train_loss:3.5888 train_time:103585ms step_avg:167.34ms
step:630/1530 train_loss:3.7207 train_time:103755ms step_avg:167.35ms
step:631/1530 train_loss:3.7374 train_time:103926ms step_avg:167.35ms
step:632/1530 train_loss:3.6450 train_time:104097ms step_avg:167.36ms
step:633/1530 train_loss:3.6044 train_time:104269ms step_avg:167.37ms
step:634/1530 train_loss:3.6927 train_time:104439ms step_avg:167.37ms
step:635/1530 train_loss:3.9474 train_time:104609ms step_avg:167.38ms
step:636/1530 train_loss:3.5479 train_time:104779ms step_avg:167.38ms
step:637/1530 train_loss:3.3494 train_time:104952ms step_avg:167.39ms
step:638/1530 train_loss:3.5957 train_time:105123ms step_avg:167.39ms
step:639/1530 train_loss:3.6282 train_time:105294ms step_avg:167.40ms
step:640/1530 train_loss:3.5636 train_time:105464ms step_avg:167.40ms
step:641/1530 train_loss:3.5819 train_time:105636ms step_avg:167.41ms
step:642/1530 train_loss:3.6261 train_time:105805ms step_avg:167.41ms
step:643/1530 train_loss:3.5913 train_time:105975ms step_avg:167.42ms
step:644/1530 train_loss:3.5584 train_time:106144ms step_avg:167.42ms
step:645/1530 train_loss:3.7742 train_time:106316ms step_avg:167.43ms
step:646/1530 train_loss:3.6695 train_time:106486ms step_avg:167.43ms
step:647/1530 train_loss:3.6574 train_time:106656ms step_avg:167.44ms
step:648/1530 train_loss:3.7062 train_time:106831ms step_avg:167.45ms
step:649/1530 train_loss:3.7632 train_time:107001ms step_avg:167.45ms
step:650/1530 train_loss:3.6211 train_time:107171ms step_avg:167.46ms
step:651/1530 train_loss:3.7627 train_time:107343ms step_avg:167.46ms
step:652/1530 train_loss:3.5892 train_time:107513ms step_avg:167.47ms
step:653/1530 train_loss:3.6604 train_time:107683ms step_avg:167.47ms
step:654/1530 train_loss:3.4270 train_time:107854ms step_avg:167.48ms
step:655/1530 train_loss:3.5763 train_time:108024ms step_avg:167.48ms
step:656/1530 train_loss:3.5755 train_time:108194ms step_avg:167.48ms
step:657/1530 train_loss:3.5065 train_time:108365ms step_avg:167.49ms
step:658/1530 train_loss:3.6837 train_time:108535ms step_avg:167.49ms
step:659/1530 train_loss:3.5827 train_time:108705ms step_avg:167.50ms
step:660/1530 train_loss:3.6816 train_time:108876ms step_avg:167.50ms
step:661/1530 train_loss:3.7488 train_time:109048ms step_avg:167.51ms
step:662/1530 train_loss:3.6682 train_time:109218ms step_avg:167.51ms
step:663/1530 train_loss:3.5499 train_time:109388ms step_avg:167.52ms
step:664/1530 train_loss:3.6037 train_time:109558ms step_avg:167.52ms
step:665/1530 train_loss:3.4925 train_time:109731ms step_avg:167.53ms
step:666/1530 train_loss:3.7765 train_time:109902ms step_avg:167.53ms
step:667/1530 train_loss:3.6001 train_time:110073ms step_avg:167.54ms
step:668/1530 train_loss:3.6397 train_time:110242ms step_avg:167.54ms
step:669/1530 train_loss:3.4836 train_time:110415ms step_avg:167.55ms
step:670/1530 train_loss:3.5981 train_time:110585ms step_avg:167.55ms
step:671/1530 train_loss:3.5573 train_time:110755ms step_avg:167.56ms
step:672/1530 train_loss:3.5626 train_time:110926ms step_avg:167.56ms
step:673/1530 train_loss:3.8481 train_time:111096ms step_avg:167.57ms
step:674/1530 train_loss:3.6222 train_time:111266ms step_avg:167.57ms
step:675/1530 train_loss:3.7066 train_time:111438ms step_avg:167.58ms
step:676/1530 train_loss:3.4882 train_time:111609ms step_avg:167.58ms
step:677/1530 train_loss:3.5981 train_time:111780ms step_avg:167.59ms
step:678/1530 train_loss:3.5573 train_time:111953ms step_avg:167.59ms
step:679/1530 train_loss:3.6740 train_time:112123ms step_avg:167.60ms
step:680/1530 train_loss:3.5857 train_time:112293ms step_avg:167.60ms
step:681/1530 train_loss:3.6145 train_time:112463ms step_avg:167.60ms
step:682/1530 train_loss:3.6653 train_time:112639ms step_avg:167.62ms
step:683/1530 train_loss:3.7372 train_time:112812ms step_avg:167.63ms
step:684/1530 train_loss:3.6454 train_time:112983ms step_avg:167.63ms
step:685/1530 train_loss:3.6848 train_time:113159ms step_avg:167.64ms
step:686/1530 train_loss:3.6357 train_time:113334ms step_avg:167.65ms
step:687/1530 train_loss:3.6614 train_time:113505ms step_avg:167.66ms
step:688/1530 train_loss:3.2123 train_time:113679ms step_avg:167.67ms
step:689/1530 train_loss:3.3947 train_time:113853ms step_avg:167.68ms
step:690/1530 train_loss:3.5357 train_time:114027ms step_avg:167.69ms
step:691/1530 train_loss:3.4076 train_time:114199ms step_avg:167.69ms
step:692/1530 train_loss:3.6295 train_time:114371ms step_avg:167.70ms
step:693/1530 train_loss:3.6420 train_time:114544ms step_avg:167.71ms
step:694/1530 train_loss:3.5529 train_time:114716ms step_avg:167.71ms
step:695/1530 train_loss:3.5350 train_time:114888ms step_avg:167.72ms
step:696/1530 train_loss:3.8566 train_time:115059ms step_avg:167.72ms
step:697/1530 train_loss:3.5845 train_time:115235ms step_avg:167.74ms
step:698/1530 train_loss:3.6394 train_time:115406ms step_avg:167.74ms
step:699/1530 train_loss:3.7637 train_time:115579ms step_avg:167.75ms
step:700/1530 train_loss:3.5687 train_time:115750ms step_avg:167.75ms
step:701/1530 train_loss:3.5404 train_time:115921ms step_avg:167.76ms
step:702/1530 train_loss:3.5073 train_time:116095ms step_avg:167.77ms
step:703/1530 train_loss:3.4993 train_time:116267ms step_avg:167.77ms
step:704/1530 train_loss:3.5666 train_time:116440ms step_avg:167.78ms
step:705/1530 train_loss:3.5588 train_time:116618ms step_avg:167.80ms
step:706/1530 train_loss:3.5747 train_time:116794ms step_avg:167.81ms
step:707/1530 train_loss:3.6393 train_time:116968ms step_avg:167.82ms
step:708/1530 train_loss:3.5969 train_time:117139ms step_avg:167.82ms
step:709/1530 train_loss:3.5787 train_time:117313ms step_avg:167.83ms
step:710/1530 train_loss:3.5402 train_time:117483ms step_avg:167.83ms
step:711/1530 train_loss:3.5883 train_time:117658ms step_avg:167.84ms
step:712/1530 train_loss:3.6441 train_time:117834ms step_avg:167.86ms
step:713/1530 train_loss:3.6465 train_time:118009ms step_avg:167.87ms
step:714/1530 train_loss:3.5582 train_time:118182ms step_avg:167.87ms
step:715/1530 train_loss:3.5689 train_time:118355ms step_avg:167.88ms
step:716/1530 train_loss:3.5841 train_time:118527ms step_avg:167.89ms
step:717/1530 train_loss:3.7011 train_time:118701ms step_avg:167.89ms
step:718/1530 train_loss:3.5938 train_time:118873ms step_avg:167.90ms
step:719/1530 train_loss:3.6771 train_time:119043ms step_avg:167.90ms
step:720/1530 train_loss:3.8437 train_time:119219ms step_avg:167.91ms
step:721/1530 train_loss:3.4637 train_time:119391ms step_avg:167.92ms
step:722/1530 train_loss:3.7351 train_time:119563ms step_avg:167.93ms
step:723/1530 train_loss:3.7647 train_time:119736ms step_avg:167.93ms
step:724/1530 train_loss:3.5660 train_time:119909ms step_avg:167.94ms
step:725/1530 train_loss:3.6550 train_time:120081ms step_avg:167.95ms
step:726/1530 train_loss:3.5267 train_time:120255ms step_avg:167.95ms
step:727/1530 train_loss:3.5788 train_time:120432ms step_avg:167.97ms
step:728/1530 train_loss:3.7276 train_time:120603ms step_avg:167.97ms
step:729/1530 train_loss:3.6666 train_time:120776ms step_avg:167.98ms
step:730/1530 train_loss:3.6583 train_time:120949ms step_avg:167.98ms
step:731/1530 train_loss:3.5509 train_time:121121ms step_avg:167.99ms
step:732/1530 train_loss:3.5939 train_time:121293ms step_avg:168.00ms
step:733/1530 train_loss:3.8237 train_time:121467ms step_avg:168.00ms
step:734/1530 train_loss:3.5602 train_time:121641ms step_avg:168.01ms
step:735/1530 train_loss:3.6118 train_time:121815ms step_avg:168.02ms
step:736/1530 train_loss:3.7327 train_time:121987ms step_avg:168.03ms
step:737/1530 train_loss:3.6750 train_time:122160ms step_avg:168.03ms
step:738/1530 train_loss:3.6038 train_time:122332ms step_avg:168.04ms
step:739/1530 train_loss:3.4967 train_time:122503ms step_avg:168.04ms
step:740/1530 train_loss:4.1112 train_time:122679ms step_avg:168.05ms
step:741/1530 train_loss:3.4882 train_time:122850ms step_avg:168.06ms
step:742/1530 train_loss:3.5513 train_time:123023ms step_avg:168.06ms
step:743/1530 train_loss:3.5792 train_time:123195ms step_avg:168.07ms
step:744/1530 train_loss:3.6457 train_time:123367ms step_avg:168.08ms
step:745/1530 train_loss:3.5715 train_time:123541ms step_avg:168.08ms
step:746/1530 train_loss:3.5925 train_time:123713ms step_avg:168.09ms
step:747/1530 train_loss:3.6444 train_time:123886ms step_avg:168.09ms
step:748/1530 train_loss:3.5606 train_time:124061ms step_avg:168.10ms
step:749/1530 train_loss:3.5579 train_time:124235ms step_avg:168.11ms
step:750/1530 train_loss:3.5904 train_time:124405ms step_avg:168.11ms
step:750/1530 val_loss:3.5631 train_time:124454ms step_avg:168.18ms
step:751/1530 train_loss:3.5711 train_time:124579ms step_avg:168.12ms
step:752/1530 train_loss:3.6144 train_time:124752ms step_avg:168.13ms
step:753/1530 train_loss:3.6123 train_time:124924ms step_avg:168.13ms
step:754/1530 train_loss:3.5930 train_time:125097ms step_avg:168.14ms
step:755/1530 train_loss:3.6819 train_time:125410ms step_avg:168.34ms
step:756/1530 train_loss:3.4563 train_time:125594ms step_avg:168.36ms
step:757/1530 train_loss:3.7232 train_time:125765ms step_avg:168.36ms
step:758/1530 train_loss:3.6481 train_time:125938ms step_avg:168.37ms
step:759/1530 train_loss:3.5863 train_time:126262ms step_avg:168.57ms
step:760/1530 train_loss:3.7035 train_time:126436ms step_avg:168.58ms
step:761/1530 train_loss:3.3922 train_time:126606ms step_avg:168.58ms
step:762/1530 train_loss:3.5438 train_time:126779ms step_avg:168.59ms
step:763/1530 train_loss:3.6557 train_time:126952ms step_avg:168.60ms
step:764/1530 train_loss:3.3185 train_time:127126ms step_avg:168.60ms
step:765/1530 train_loss:3.7324 train_time:127298ms step_avg:168.61ms
step:766/1530 train_loss:3.5737 train_time:127471ms step_avg:168.61ms
step:767/1530 train_loss:3.5684 train_time:127643ms step_avg:168.62ms
step:768/1530 train_loss:3.5667 train_time:127818ms step_avg:168.62ms
step:769/1530 train_loss:3.5779 train_time:127991ms step_avg:168.63ms
step:770/1530 train_loss:3.6394 train_time:128162ms step_avg:168.63ms
step:771/1530 train_loss:3.8803 train_time:128337ms step_avg:168.64ms
step:772/1530 train_loss:3.4464 train_time:128508ms step_avg:168.65ms
step:773/1530 train_loss:3.6257 train_time:128679ms step_avg:168.65ms
step:774/1530 train_loss:3.6379 train_time:128850ms step_avg:168.65ms
step:775/1530 train_loss:3.6055 train_time:129021ms step_avg:168.66ms
step:776/1530 train_loss:3.3964 train_time:129196ms step_avg:168.66ms
step:777/1530 train_loss:3.3844 train_time:129368ms step_avg:168.67ms
step:778/1530 train_loss:3.4909 train_time:129540ms step_avg:168.67ms
step:779/1530 train_loss:3.5771 train_time:129713ms step_avg:168.68ms
step:780/1530 train_loss:3.5836 train_time:129884ms step_avg:168.68ms
step:781/1530 train_loss:3.6668 train_time:130058ms step_avg:168.69ms
step:782/1530 train_loss:3.5889 train_time:130231ms step_avg:168.69ms
step:783/1530 train_loss:3.5691 train_time:130401ms step_avg:168.69ms
step:784/1530 train_loss:3.6091 train_time:130574ms step_avg:168.70ms
step:785/1530 train_loss:3.5576 train_time:130746ms step_avg:168.70ms
step:786/1530 train_loss:3.4348 train_time:130918ms step_avg:168.71ms
step:787/1530 train_loss:3.7665 train_time:131089ms step_avg:168.71ms
step:788/1530 train_loss:3.5038 train_time:131263ms step_avg:168.72ms
step:789/1530 train_loss:3.5439 train_time:131435ms step_avg:168.72ms
step:790/1530 train_loss:3.6227 train_time:131607ms step_avg:168.73ms
step:791/1530 train_loss:3.7699 train_time:131782ms step_avg:168.74ms
step:792/1530 train_loss:3.7611 train_time:131957ms step_avg:168.74ms
step:793/1530 train_loss:3.4455 train_time:132128ms step_avg:168.75ms
step:794/1530 train_loss:3.5884 train_time:132299ms step_avg:168.75ms
step:795/1530 train_loss:3.6688 train_time:132474ms step_avg:168.76ms
step:796/1530 train_loss:3.7288 train_time:132650ms step_avg:168.77ms
step:797/1530 train_loss:3.5246 train_time:132823ms step_avg:168.77ms
step:798/1530 train_loss:3.6424 train_time:132998ms step_avg:168.78ms
step:799/1530 train_loss:3.5336 train_time:133176ms step_avg:168.79ms
step:800/1530 train_loss:3.5259 train_time:133349ms step_avg:168.80ms
step:801/1530 train_loss:3.6250 train_time:133522ms step_avg:168.80ms
step:802/1530 train_loss:3.4972 train_time:133697ms step_avg:168.81ms
step:803/1530 train_loss:3.4919 train_time:133870ms step_avg:168.81ms
step:804/1530 train_loss:3.6178 train_time:134044ms step_avg:168.82ms
step:805/1530 train_loss:3.5127 train_time:134221ms step_avg:168.83ms
step:806/1530 train_loss:3.5589 train_time:134394ms step_avg:168.84ms
step:807/1530 train_loss:3.6354 train_time:134566ms step_avg:168.84ms
step:808/1530 train_loss:3.5418 train_time:134741ms step_avg:168.85ms
step:809/1530 train_loss:3.4878 train_time:134915ms step_avg:168.86ms
step:810/1530 train_loss:3.5576 train_time:135087ms step_avg:168.86ms
step:811/1530 train_loss:3.5759 train_time:135260ms step_avg:168.86ms
step:812/1530 train_loss:3.6044 train_time:135433ms step_avg:168.87ms
step:813/1530 train_loss:3.6211 train_time:135605ms step_avg:168.87ms
step:814/1530 train_loss:3.5604 train_time:135780ms step_avg:168.88ms
step:815/1530 train_loss:3.5626 train_time:135954ms step_avg:168.89ms
step:816/1530 train_loss:3.6896 train_time:136128ms step_avg:168.89ms
step:817/1530 train_loss:3.7640 train_time:136301ms step_avg:168.90ms
step:818/1530 train_loss:3.5181 train_time:136473ms step_avg:168.90ms
step:819/1530 train_loss:3.7133 train_time:136648ms step_avg:168.91ms
step:820/1530 train_loss:3.4927 train_time:136822ms step_avg:168.92ms
step:821/1530 train_loss:3.5598 train_time:136995ms step_avg:168.92ms
step:822/1530 train_loss:3.6946 train_time:137171ms step_avg:168.93ms
step:823/1530 train_loss:3.5750 train_time:137345ms step_avg:168.94ms
step:824/1530 train_loss:3.5089 train_time:137518ms step_avg:168.94ms
step:825/1530 train_loss:3.6174 train_time:137694ms step_avg:168.95ms
step:826/1530 train_loss:3.4780 train_time:137868ms step_avg:168.96ms
step:827/1530 train_loss:3.7267 train_time:138043ms step_avg:168.96ms
step:828/1530 train_loss:3.6154 train_time:138217ms step_avg:168.97ms
step:829/1530 train_loss:3.6261 train_time:138393ms step_avg:168.98ms
step:830/1530 train_loss:3.5354 train_time:138566ms step_avg:168.98ms
step:831/1530 train_loss:3.5967 train_time:138739ms step_avg:168.99ms
step:832/1530 train_loss:3.5113 train_time:138914ms step_avg:169.00ms
step:833/1530 train_loss:3.6472 train_time:139088ms step_avg:169.00ms
step:834/1530 train_loss:3.4688 train_time:139262ms step_avg:169.01ms
step:835/1530 train_loss:3.4517 train_time:139437ms step_avg:169.01ms
step:836/1530 train_loss:3.7154 train_time:139612ms step_avg:169.02ms
step:837/1530 train_loss:3.3975 train_time:139785ms step_avg:169.03ms
step:838/1530 train_loss:3.5953 train_time:139959ms step_avg:169.03ms
step:839/1530 train_loss:3.4155 train_time:140136ms step_avg:169.04ms
step:840/1530 train_loss:3.4682 train_time:140308ms step_avg:169.05ms
step:841/1530 train_loss:3.5649 train_time:140481ms step_avg:169.05ms
step:842/1530 train_loss:3.5767 train_time:140657ms step_avg:169.06ms
step:843/1530 train_loss:3.5595 train_time:140829ms step_avg:169.06ms
step:844/1530 train_loss:3.4272 train_time:141001ms step_avg:169.07ms
step:845/1530 train_loss:3.6604 train_time:141176ms step_avg:169.07ms
step:846/1530 train_loss:3.5123 train_time:141351ms step_avg:169.08ms
step:847/1530 train_loss:3.4911 train_time:141525ms step_avg:169.09ms
step:848/1530 train_loss:3.6356 train_time:141698ms step_avg:169.09ms
step:849/1530 train_loss:3.4843 train_time:141872ms step_avg:169.10ms
step:850/1530 train_loss:3.4368 train_time:142046ms step_avg:169.10ms
step:851/1530 train_loss:3.7271 train_time:142219ms step_avg:169.11ms
step:852/1530 train_loss:3.4351 train_time:142392ms step_avg:169.11ms
step:853/1530 train_loss:3.5670 train_time:142565ms step_avg:169.12ms
step:854/1530 train_loss:3.6497 train_time:142741ms step_avg:169.12ms
step:855/1530 train_loss:3.5176 train_time:142915ms step_avg:169.13ms
step:856/1530 train_loss:3.5418 train_time:143087ms step_avg:169.13ms
step:857/1530 train_loss:3.6004 train_time:143262ms step_avg:169.14ms
step:858/1530 train_loss:3.4711 train_time:143439ms step_avg:169.15ms
step:859/1530 train_loss:3.5534 train_time:143612ms step_avg:169.15ms
step:860/1530 train_loss:3.5831 train_time:143784ms step_avg:169.16ms
step:861/1530 train_loss:3.6318 train_time:143963ms step_avg:169.17ms
step:862/1530 train_loss:3.6055 train_time:144141ms step_avg:169.18ms
step:863/1530 train_loss:3.5687 train_time:144316ms step_avg:169.19ms
step:864/1530 train_loss:3.3806 train_time:144490ms step_avg:169.19ms
step:865/1530 train_loss:3.5975 train_time:144661ms step_avg:169.19ms
step:866/1530 train_loss:3.8976 train_time:144840ms step_avg:169.21ms
step:867/1530 train_loss:3.4513 train_time:145011ms step_avg:169.21ms
step:868/1530 train_loss:3.6402 train_time:145183ms step_avg:169.21ms
step:869/1530 train_loss:3.6152 train_time:145358ms step_avg:169.22ms
step:870/1530 train_loss:3.4476 train_time:145533ms step_avg:169.22ms
step:871/1530 train_loss:3.4064 train_time:145706ms step_avg:169.23ms
step:872/1530 train_loss:3.6501 train_time:145881ms step_avg:169.24ms
step:873/1530 train_loss:3.4552 train_time:146055ms step_avg:169.24ms
step:874/1530 train_loss:3.2182 train_time:146234ms step_avg:169.25ms
step:875/1530 train_loss:3.6300 train_time:146407ms step_avg:169.26ms
step:875/1530 val_loss:3.5161 train_time:146456ms step_avg:169.31ms
step:876/1530 train_loss:3.4364 train_time:146580ms step_avg:169.26ms
step:877/1530 train_loss:3.6110 train_time:146757ms step_avg:169.27ms
step:878/1530 train_loss:3.4628 train_time:146932ms step_avg:169.28ms
step:879/1530 train_loss:3.6421 train_time:147103ms step_avg:169.28ms
step:880/1530 train_loss:3.3082 train_time:147277ms step_avg:169.28ms
step:881/1530 train_loss:3.4714 train_time:147448ms step_avg:169.29ms
step:882/1530 train_loss:3.6937 train_time:147620ms step_avg:169.29ms
step:883/1530 train_loss:3.8366 train_time:147794ms step_avg:169.29ms
step:884/1530 train_loss:3.5637 train_time:147970ms step_avg:169.30ms
step:885/1530 train_loss:3.4944 train_time:148143ms step_avg:169.31ms
step:886/1530 train_loss:3.5685 train_time:148317ms step_avg:169.31ms
step:887/1530 train_loss:4.0799 train_time:148490ms step_avg:169.32ms
step:888/1530 train_loss:3.8278 train_time:148670ms step_avg:169.33ms
step:889/1530 train_loss:3.5160 train_time:148843ms step_avg:169.33ms
step:890/1530 train_loss:3.5323 train_time:149016ms step_avg:169.34ms
step:891/1530 train_loss:3.3577 train_time:149189ms step_avg:169.34ms
step:892/1530 train_loss:3.7138 train_time:149362ms step_avg:169.34ms
step:893/1530 train_loss:3.4180 train_time:149534ms step_avg:169.35ms
step:894/1530 train_loss:3.6423 train_time:149709ms step_avg:169.35ms
step:895/1530 train_loss:3.6753 train_time:149883ms step_avg:169.36ms
step:896/1530 train_loss:3.4982 train_time:150057ms step_avg:169.37ms
step:897/1530 train_loss:3.5431 train_time:150232ms step_avg:169.37ms
step:898/1530 train_loss:3.5870 train_time:150407ms step_avg:169.38ms
step:899/1530 train_loss:3.4717 train_time:150579ms step_avg:169.38ms
step:900/1530 train_loss:3.4292 train_time:150753ms step_avg:169.39ms
step:901/1530 train_loss:3.6144 train_time:150925ms step_avg:169.39ms
step:902/1530 train_loss:3.6302 train_time:151097ms step_avg:169.39ms
step:903/1530 train_loss:3.5399 train_time:151275ms step_avg:169.40ms
step:904/1530 train_loss:3.4849 train_time:151446ms step_avg:169.40ms
step:905/1530 train_loss:3.4973 train_time:151616ms step_avg:169.40ms
step:906/1530 train_loss:3.7063 train_time:151791ms step_avg:169.41ms
step:907/1530 train_loss:3.5204 train_time:151965ms step_avg:169.42ms
step:908/1530 train_loss:3.5632 train_time:152137ms step_avg:169.42ms
step:909/1530 train_loss:3.4542 train_time:152314ms step_avg:169.43ms
step:910/1530 train_loss:3.5232 train_time:152493ms step_avg:169.44ms
step:911/1530 train_loss:3.6402 train_time:152668ms step_avg:169.44ms
step:912/1530 train_loss:3.5937 train_time:152844ms step_avg:169.45ms
step:913/1530 train_loss:3.4565 train_time:153022ms step_avg:169.46ms
step:914/1530 train_loss:3.7413 train_time:153200ms step_avg:169.47ms
step:915/1530 train_loss:3.5284 train_time:153380ms step_avg:169.48ms
step:916/1530 train_loss:3.6132 train_time:153557ms step_avg:169.49ms
step:917/1530 train_loss:3.5987 train_time:153732ms step_avg:169.49ms
step:918/1530 train_loss:4.8190 train_time:153910ms step_avg:169.50ms
step:919/1530 train_loss:3.4963 train_time:154088ms step_avg:169.51ms
step:920/1530 train_loss:3.5876 train_time:154262ms step_avg:169.52ms
step:921/1530 train_loss:3.5495 train_time:154440ms step_avg:169.53ms
step:922/1530 train_loss:3.5861 train_time:154618ms step_avg:169.54ms
step:923/1530 train_loss:3.6125 train_time:154795ms step_avg:169.55ms
step:924/1530 train_loss:3.6757 train_time:154972ms step_avg:169.55ms
step:925/1530 train_loss:3.6422 train_time:155144ms step_avg:169.56ms
step:926/1530 train_loss:3.5535 train_time:155318ms step_avg:169.56ms
step:927/1530 train_loss:3.5529 train_time:155493ms step_avg:169.57ms
step:928/1530 train_loss:3.7757 train_time:155669ms step_avg:169.57ms
step:929/1530 train_loss:3.6078 train_time:155843ms step_avg:169.58ms
step:930/1530 train_loss:3.3994 train_time:156019ms step_avg:169.59ms
step:931/1530 train_loss:3.4907 train_time:156194ms step_avg:169.59ms
step:932/1530 train_loss:3.6426 train_time:156372ms step_avg:169.60ms
step:933/1530 train_loss:3.3642 train_time:156547ms step_avg:169.61ms
step:934/1530 train_loss:3.5794 train_time:156725ms step_avg:169.62ms
step:935/1530 train_loss:3.4364 train_time:156904ms step_avg:169.63ms
step:936/1530 train_loss:3.5155 train_time:157082ms step_avg:169.64ms
step:937/1530 train_loss:3.6188 train_time:157261ms step_avg:169.65ms
step:938/1530 train_loss:3.5379 train_time:157436ms step_avg:169.65ms
step:939/1530 train_loss:3.6765 train_time:157618ms step_avg:169.66ms
step:940/1530 train_loss:3.4782 train_time:157792ms step_avg:169.67ms
step:941/1530 train_loss:3.5462 train_time:157967ms step_avg:169.67ms
step:942/1530 train_loss:3.3565 train_time:158143ms step_avg:169.68ms
step:943/1530 train_loss:3.7093 train_time:158326ms step_avg:169.70ms
step:944/1530 train_loss:3.3993 train_time:158639ms step_avg:169.85ms
step:945/1530 train_loss:3.4239 train_time:158823ms step_avg:169.86ms
step:946/1530 train_loss:5.0689 train_time:159003ms step_avg:169.87ms
step:947/1530 train_loss:3.5963 train_time:159179ms step_avg:169.88ms
step:948/1530 train_loss:3.4902 train_time:159352ms step_avg:169.88ms
step:949/1530 train_loss:3.3714 train_time:159676ms step_avg:170.05ms
step:950/1530 train_loss:3.4367 train_time:159851ms step_avg:170.05ms
step:951/1530 train_loss:3.4042 train_time:160028ms step_avg:170.06ms
step:952/1530 train_loss:3.4754 train_time:160204ms step_avg:170.07ms
step:953/1530 train_loss:3.5674 train_time:160382ms step_avg:170.08ms
step:954/1530 train_loss:3.4431 train_time:160560ms step_avg:170.08ms
step:955/1530 train_loss:3.4722 train_time:160735ms step_avg:170.09ms
step:956/1530 train_loss:3.4419 train_time:160910ms step_avg:170.10ms
step:957/1530 train_loss:3.4934 train_time:161088ms step_avg:170.10ms
step:958/1530 train_loss:3.4990 train_time:161267ms step_avg:170.11ms
step:959/1530 train_loss:3.5122 train_time:161442ms step_avg:170.12ms
step:960/1530 train_loss:3.4004 train_time:161620ms step_avg:170.13ms
step:961/1530 train_loss:3.6437 train_time:161795ms step_avg:170.13ms
step:962/1530 train_loss:3.5917 train_time:161970ms step_avg:170.14ms
step:963/1530 train_loss:3.6115 train_time:162144ms step_avg:170.14ms
step:964/1530 train_loss:3.4267 train_time:162322ms step_avg:170.15ms
step:965/1530 train_loss:3.4766 train_time:162495ms step_avg:170.15ms
step:966/1530 train_loss:3.7062 train_time:162671ms step_avg:170.16ms
step:967/1530 train_loss:3.5167 train_time:162844ms step_avg:170.16ms
step:968/1530 train_loss:3.5066 train_time:163020ms step_avg:170.17ms
step:969/1530 train_loss:3.5839 train_time:163195ms step_avg:170.17ms
step:970/1530 train_loss:3.3722 train_time:163368ms step_avg:170.18ms
step:971/1530 train_loss:3.5282 train_time:163541ms step_avg:170.18ms
step:972/1530 train_loss:3.4713 train_time:163715ms step_avg:170.18ms
step:973/1530 train_loss:3.5352 train_time:163889ms step_avg:170.19ms
step:974/1530 train_loss:3.5853 train_time:164066ms step_avg:170.19ms
step:975/1530 train_loss:3.4653 train_time:164241ms step_avg:170.20ms
step:976/1530 train_loss:3.6738 train_time:164416ms step_avg:170.20ms
step:977/1530 train_loss:3.5700 train_time:164589ms step_avg:170.21ms
step:978/1530 train_loss:3.3543 train_time:164763ms step_avg:170.21ms
step:979/1530 train_loss:3.6230 train_time:164939ms step_avg:170.22ms
step:980/1530 train_loss:3.4124 train_time:165114ms step_avg:170.22ms
step:981/1530 train_loss:3.5722 train_time:165291ms step_avg:170.23ms
step:982/1530 train_loss:3.5425 train_time:165465ms step_avg:170.23ms
step:983/1530 train_loss:3.5109 train_time:165640ms step_avg:170.24ms
step:984/1530 train_loss:3.4922 train_time:165815ms step_avg:170.24ms
step:985/1530 train_loss:3.5726 train_time:165991ms step_avg:170.25ms
step:986/1530 train_loss:3.4087 train_time:166167ms step_avg:170.25ms
step:987/1530 train_loss:3.4882 train_time:166339ms step_avg:170.25ms
step:988/1530 train_loss:3.4690 train_time:166513ms step_avg:170.26ms
step:989/1530 train_loss:3.4142 train_time:166684ms step_avg:170.26ms
step:990/1530 train_loss:3.6528 train_time:166862ms step_avg:170.27ms
step:991/1530 train_loss:3.4659 train_time:167036ms step_avg:170.27ms
step:992/1530 train_loss:3.4415 train_time:167216ms step_avg:170.28ms
step:993/1530 train_loss:3.4953 train_time:167394ms step_avg:170.29ms
step:994/1530 train_loss:3.5927 train_time:167569ms step_avg:170.29ms
step:995/1530 train_loss:3.5258 train_time:167741ms step_avg:170.30ms
step:996/1530 train_loss:3.4558 train_time:167914ms step_avg:170.30ms
step:997/1530 train_loss:3.7483 train_time:168087ms step_avg:170.30ms
step:998/1530 train_loss:3.4350 train_time:168260ms step_avg:170.30ms
step:999/1530 train_loss:3.5867 train_time:168436ms step_avg:170.31ms
step:1000/1530 train_loss:3.4359 train_time:168613ms step_avg:170.32ms
step:1000/1530 val_loss:3.4637 train_time:168663ms step_avg:170.37ms
step:1001/1530 train_loss:3.4983 train_time:168789ms step_avg:170.32ms
step:1002/1530 train_loss:3.3737 train_time:168966ms step_avg:170.33ms
step:1003/1530 train_loss:3.5552 train_time:169143ms step_avg:170.34ms
step:1004/1530 train_loss:3.6017 train_time:169318ms step_avg:170.34ms
step:1005/1530 train_loss:3.3916 train_time:169493ms step_avg:170.35ms
step:1006/1530 train_loss:3.4648 train_time:169670ms step_avg:170.35ms
step:1007/1530 train_loss:3.4322 train_time:169846ms step_avg:170.36ms
step:1008/1530 train_loss:3.5553 train_time:170022ms step_avg:170.36ms
step:1009/1530 train_loss:3.6600 train_time:170199ms step_avg:170.37ms
step:1010/1530 train_loss:3.5601 train_time:170371ms step_avg:170.37ms
step:1011/1530 train_loss:3.5306 train_time:170545ms step_avg:170.37ms
step:1012/1530 train_loss:3.3935 train_time:170720ms step_avg:170.38ms
step:1013/1530 train_loss:3.5342 train_time:170895ms step_avg:170.38ms
step:1014/1530 train_loss:3.6182 train_time:171072ms step_avg:170.39ms
step:1015/1530 train_loss:3.3265 train_time:171248ms step_avg:170.40ms
step:1016/1530 train_loss:3.4080 train_time:171423ms step_avg:170.40ms
step:1017/1530 train_loss:3.4012 train_time:171601ms step_avg:170.41ms
step:1018/1530 train_loss:3.3931 train_time:171775ms step_avg:170.41ms
step:1019/1530 train_loss:3.5146 train_time:171952ms step_avg:170.42ms
step:1020/1530 train_loss:3.3749 train_time:172130ms step_avg:170.43ms
step:1021/1530 train_loss:3.3558 train_time:172306ms step_avg:170.43ms
step:1022/1530 train_loss:3.4743 train_time:172484ms step_avg:170.44ms
step:1023/1530 train_loss:3.5058 train_time:172659ms step_avg:170.44ms
step:1024/1530 train_loss:3.4785 train_time:172836ms step_avg:170.45ms
step:1025/1530 train_loss:3.4802 train_time:173015ms step_avg:170.46ms
step:1026/1530 train_loss:3.6117 train_time:173191ms step_avg:170.46ms
step:1027/1530 train_loss:3.3159 train_time:173367ms step_avg:170.47ms
step:1028/1530 train_loss:3.3939 train_time:173548ms step_avg:170.48ms
step:1029/1530 train_loss:3.3085 train_time:173727ms step_avg:170.49ms
step:1030/1530 train_loss:3.5356 train_time:173903ms step_avg:170.49ms
step:1031/1530 train_loss:3.5034 train_time:174080ms step_avg:170.50ms
step:1032/1530 train_loss:3.6949 train_time:174260ms step_avg:170.51ms
step:1033/1530 train_loss:3.4865 train_time:174435ms step_avg:170.51ms
step:1034/1530 train_loss:3.4045 train_time:174614ms step_avg:170.52ms
step:1035/1530 train_loss:3.4420 train_time:174793ms step_avg:170.53ms
step:1036/1530 train_loss:3.4803 train_time:174970ms step_avg:170.54ms
step:1037/1530 train_loss:3.7836 train_time:175148ms step_avg:170.54ms
step:1038/1530 train_loss:3.6163 train_time:175328ms step_avg:170.55ms
step:1039/1530 train_loss:3.5060 train_time:175510ms step_avg:170.56ms
step:1040/1530 train_loss:3.4141 train_time:175688ms step_avg:170.57ms
step:1041/1530 train_loss:3.4838 train_time:175867ms step_avg:170.58ms
step:1042/1530 train_loss:3.5204 train_time:176041ms step_avg:170.58ms
step:1043/1530 train_loss:3.4424 train_time:176215ms step_avg:170.59ms
step:1044/1530 train_loss:3.4546 train_time:176391ms step_avg:170.59ms
step:1045/1530 train_loss:3.5109 train_time:176569ms step_avg:170.60ms
step:1046/1530 train_loss:3.4231 train_time:176744ms step_avg:170.60ms
step:1047/1530 train_loss:3.6302 train_time:176922ms step_avg:170.61ms
step:1048/1530 train_loss:3.4943 train_time:177096ms step_avg:170.61ms
step:1049/1530 train_loss:3.4001 train_time:177272ms step_avg:170.62ms
step:1050/1530 train_loss:3.3898 train_time:177451ms step_avg:170.63ms
step:1051/1530 train_loss:3.4900 train_time:177628ms step_avg:170.63ms
step:1052/1530 train_loss:3.3599 train_time:177807ms step_avg:170.64ms
step:1053/1530 train_loss:3.6921 train_time:177985ms step_avg:170.65ms
step:1054/1530 train_loss:3.5393 train_time:178163ms step_avg:170.65ms
step:1055/1530 train_loss:3.3788 train_time:178339ms step_avg:170.66ms
step:1056/1530 train_loss:3.4950 train_time:178513ms step_avg:170.66ms
step:1057/1530 train_loss:3.5729 train_time:178690ms step_avg:170.67ms
step:1058/1530 train_loss:3.3001 train_time:178869ms step_avg:170.68ms
step:1059/1530 train_loss:3.3682 train_time:179050ms step_avg:170.69ms
step:1060/1530 train_loss:3.4383 train_time:179226ms step_avg:170.69ms
step:1061/1530 train_loss:3.4153 train_time:179398ms step_avg:170.69ms
step:1062/1530 train_loss:3.3796 train_time:179574ms step_avg:170.70ms
step:1063/1530 train_loss:3.4575 train_time:179749ms step_avg:170.70ms
step:1064/1530 train_loss:3.3798 train_time:179924ms step_avg:170.71ms
step:1065/1530 train_loss:3.3598 train_time:180101ms step_avg:170.71ms
step:1066/1530 train_loss:3.4111 train_time:180278ms step_avg:170.72ms
step:1067/1530 train_loss:3.2818 train_time:180454ms step_avg:170.72ms
step:1068/1530 train_loss:3.4299 train_time:180630ms step_avg:170.73ms
step:1069/1530 train_loss:3.2912 train_time:180810ms step_avg:170.74ms
step:1070/1530 train_loss:3.5627 train_time:180985ms step_avg:170.74ms
step:1071/1530 train_loss:3.5095 train_time:181165ms step_avg:170.75ms
step:1072/1530 train_loss:3.4355 train_time:181339ms step_avg:170.75ms
step:1073/1530 train_loss:3.5224 train_time:181512ms step_avg:170.75ms
step:1074/1530 train_loss:3.4231 train_time:181690ms step_avg:170.76ms
step:1075/1530 train_loss:3.3985 train_time:181868ms step_avg:170.77ms
step:1076/1530 train_loss:3.7978 train_time:182044ms step_avg:170.77ms
step:1077/1530 train_loss:3.4282 train_time:182219ms step_avg:170.78ms
step:1078/1530 train_loss:3.0806 train_time:182403ms step_avg:170.79ms
step:1079/1530 train_loss:3.5298 train_time:182580ms step_avg:170.80ms
step:1080/1530 train_loss:3.4287 train_time:182757ms step_avg:170.80ms
step:1081/1530 train_loss:3.4967 train_time:182930ms step_avg:170.80ms
step:1082/1530 train_loss:3.5857 train_time:183108ms step_avg:170.81ms
step:1083/1530 train_loss:3.4959 train_time:183284ms step_avg:170.81ms
step:1084/1530 train_loss:3.4618 train_time:183460ms step_avg:170.82ms
step:1085/1530 train_loss:3.4320 train_time:183633ms step_avg:170.82ms
step:1086/1530 train_loss:3.6229 train_time:183809ms step_avg:170.83ms
step:1087/1530 train_loss:3.4992 train_time:183985ms step_avg:170.83ms
step:1088/1530 train_loss:3.3678 train_time:184162ms step_avg:170.84ms
step:1089/1530 train_loss:3.3694 train_time:184342ms step_avg:170.85ms
step:1090/1530 train_loss:3.4731 train_time:184520ms step_avg:170.85ms
step:1091/1530 train_loss:3.2841 train_time:184697ms step_avg:170.86ms
step:1092/1530 train_loss:3.4877 train_time:184874ms step_avg:170.86ms
step:1093/1530 train_loss:3.5954 train_time:185051ms step_avg:170.87ms
step:1094/1530 train_loss:3.4411 train_time:185226ms step_avg:170.87ms
step:1095/1530 train_loss:3.4120 train_time:185401ms step_avg:170.88ms
step:1096/1530 train_loss:3.4234 train_time:185578ms step_avg:170.88ms
step:1097/1530 train_loss:3.4906 train_time:185756ms step_avg:170.89ms
step:1098/1530 train_loss:3.5616 train_time:185932ms step_avg:170.89ms
step:1099/1530 train_loss:3.5238 train_time:186111ms step_avg:170.90ms
step:1100/1530 train_loss:3.4257 train_time:186292ms step_avg:170.91ms
step:1101/1530 train_loss:3.2870 train_time:186471ms step_avg:170.92ms
step:1102/1530 train_loss:3.3085 train_time:186651ms step_avg:170.93ms
step:1103/1530 train_loss:3.4384 train_time:186832ms step_avg:170.93ms
step:1104/1530 train_loss:3.3174 train_time:187007ms step_avg:170.94ms
step:1105/1530 train_loss:4.0582 train_time:187187ms step_avg:170.95ms
step:1106/1530 train_loss:3.2246 train_time:187363ms step_avg:170.95ms
step:1107/1530 train_loss:3.5640 train_time:187539ms step_avg:170.96ms
step:1108/1530 train_loss:3.3452 train_time:187712ms step_avg:170.96ms
step:1109/1530 train_loss:3.4992 train_time:187888ms step_avg:170.96ms
step:1110/1530 train_loss:3.4250 train_time:188063ms step_avg:170.97ms
step:1111/1530 train_loss:3.4834 train_time:188236ms step_avg:170.97ms
step:1112/1530 train_loss:3.5539 train_time:188415ms step_avg:170.98ms
step:1113/1530 train_loss:3.4291 train_time:188597ms step_avg:170.99ms
step:1114/1530 train_loss:3.3689 train_time:188777ms step_avg:170.99ms
step:1115/1530 train_loss:3.2377 train_time:188954ms step_avg:171.00ms
step:1116/1530 train_loss:3.4276 train_time:189126ms step_avg:171.00ms
step:1117/1530 train_loss:3.5870 train_time:189305ms step_avg:171.01ms
step:1118/1530 train_loss:3.6217 train_time:189484ms step_avg:171.01ms
step:1119/1530 train_loss:3.4792 train_time:189658ms step_avg:171.02ms
step:1120/1530 train_loss:3.4891 train_time:189834ms step_avg:171.02ms
step:1121/1530 train_loss:3.3858 train_time:190012ms step_avg:171.03ms
step:1122/1530 train_loss:3.4567 train_time:190188ms step_avg:171.03ms
step:1123/1530 train_loss:3.5788 train_time:190365ms step_avg:171.04ms
step:1124/1530 train_loss:3.3341 train_time:190542ms step_avg:171.04ms
step:1125/1530 train_loss:3.2216 train_time:190718ms step_avg:171.05ms
step:1125/1530 val_loss:3.4063 train_time:190768ms step_avg:171.09ms
step:1126/1530 train_loss:3.4792 train_time:190893ms step_avg:171.05ms
step:1127/1530 train_loss:3.6719 train_time:191072ms step_avg:171.06ms
step:1128/1530 train_loss:3.2274 train_time:191250ms step_avg:171.06ms
step:1129/1530 train_loss:3.5512 train_time:191430ms step_avg:171.07ms
step:1130/1530 train_loss:3.3736 train_time:191610ms step_avg:171.08ms
step:1131/1530 train_loss:3.4020 train_time:191792ms step_avg:171.09ms
step:1132/1530 train_loss:3.3673 train_time:191966ms step_avg:171.09ms
step:1133/1530 train_loss:3.4864 train_time:192277ms step_avg:171.22ms
step:1134/1530 train_loss:3.4464 train_time:192462ms step_avg:171.23ms
step:1135/1530 train_loss:3.5191 train_time:192640ms step_avg:171.24ms
step:1136/1530 train_loss:3.5639 train_time:192816ms step_avg:171.24ms
step:1137/1530 train_loss:3.4528 train_time:192993ms step_avg:171.25ms
step:1138/1530 train_loss:3.3485 train_time:193172ms step_avg:171.25ms
step:1139/1530 train_loss:3.6547 train_time:193504ms step_avg:171.39ms
step:1140/1530 train_loss:3.4498 train_time:193683ms step_avg:171.40ms
step:1141/1530 train_loss:3.5914 train_time:193864ms step_avg:171.41ms
step:1142/1530 train_loss:3.4387 train_time:194040ms step_avg:171.41ms
step:1143/1530 train_loss:3.3652 train_time:194218ms step_avg:171.42ms
step:1144/1530 train_loss:3.4477 train_time:194394ms step_avg:171.42ms
step:1145/1530 train_loss:3.5905 train_time:194569ms step_avg:171.43ms
step:1146/1530 train_loss:3.5531 train_time:194751ms step_avg:171.44ms
step:1147/1530 train_loss:3.4832 train_time:194928ms step_avg:171.44ms
step:1148/1530 train_loss:3.4962 train_time:195108ms step_avg:171.45ms
step:1149/1530 train_loss:3.3231 train_time:195289ms step_avg:171.46ms
step:1150/1530 train_loss:3.3721 train_time:195465ms step_avg:171.46ms
step:1151/1530 train_loss:3.3192 train_time:195645ms step_avg:171.47ms
step:1152/1530 train_loss:3.3930 train_time:195827ms step_avg:171.48ms
step:1153/1530 train_loss:3.4292 train_time:196007ms step_avg:171.48ms
step:1154/1530 train_loss:3.5141 train_time:196182ms step_avg:171.49ms
step:1155/1530 train_loss:3.3162 train_time:196364ms step_avg:171.50ms
step:1156/1530 train_loss:3.5336 train_time:196545ms step_avg:171.51ms
step:1157/1530 train_loss:3.4894 train_time:196722ms step_avg:171.51ms
step:1158/1530 train_loss:3.2536 train_time:196899ms step_avg:171.51ms
step:1159/1530 train_loss:3.3456 train_time:197074ms step_avg:171.52ms
step:1160/1530 train_loss:3.3368 train_time:197248ms step_avg:171.52ms
step:1161/1530 train_loss:3.0825 train_time:197428ms step_avg:171.53ms
step:1162/1530 train_loss:3.4167 train_time:197606ms step_avg:171.53ms
step:1163/1530 train_loss:3.3878 train_time:197785ms step_avg:171.54ms
step:1164/1530 train_loss:3.2901 train_time:197964ms step_avg:171.55ms
step:1165/1530 train_loss:3.2465 train_time:198139ms step_avg:171.55ms
step:1166/1530 train_loss:3.3866 train_time:198317ms step_avg:171.55ms
step:1167/1530 train_loss:3.4084 train_time:198493ms step_avg:171.56ms
step:1168/1530 train_loss:3.7265 train_time:198668ms step_avg:171.56ms
step:1169/1530 train_loss:3.3753 train_time:198846ms step_avg:171.57ms
step:1170/1530 train_loss:3.3863 train_time:199022ms step_avg:171.57ms
step:1171/1530 train_loss:3.2812 train_time:199197ms step_avg:171.57ms
step:1172/1530 train_loss:3.4182 train_time:199372ms step_avg:171.58ms
step:1173/1530 train_loss:3.5394 train_time:199552ms step_avg:171.58ms
step:1174/1530 train_loss:3.3813 train_time:199736ms step_avg:171.59ms
step:1175/1530 train_loss:3.3673 train_time:199917ms step_avg:171.60ms
step:1176/1530 train_loss:3.4202 train_time:200096ms step_avg:171.61ms
step:1177/1530 train_loss:3.4462 train_time:200279ms step_avg:171.62ms
step:1178/1530 train_loss:3.4954 train_time:200454ms step_avg:171.62ms
step:1179/1530 train_loss:3.3936 train_time:200628ms step_avg:171.62ms
step:1180/1530 train_loss:3.3586 train_time:200815ms step_avg:171.64ms
step:1181/1530 train_loss:3.3364 train_time:200992ms step_avg:171.64ms
step:1182/1530 train_loss:3.3745 train_time:201170ms step_avg:171.65ms
step:1183/1530 train_loss:3.3321 train_time:201347ms step_avg:171.65ms
step:1184/1530 train_loss:3.5064 train_time:201524ms step_avg:171.66ms
step:1185/1530 train_loss:3.5437 train_time:201707ms step_avg:171.67ms
step:1186/1530 train_loss:3.3631 train_time:201887ms step_avg:171.67ms
step:1187/1530 train_loss:3.4130 train_time:202071ms step_avg:171.68ms
step:1188/1530 train_loss:3.4410 train_time:202247ms step_avg:171.69ms
step:1189/1530 train_loss:3.2727 train_time:202428ms step_avg:171.69ms
step:1190/1530 train_loss:3.4444 train_time:202608ms step_avg:171.70ms
step:1191/1530 train_loss:3.5761 train_time:202790ms step_avg:171.71ms
step:1192/1530 train_loss:3.3901 train_time:202966ms step_avg:171.71ms
step:1193/1530 train_loss:3.2679 train_time:203143ms step_avg:171.72ms
step:1194/1530 train_loss:3.5560 train_time:203320ms step_avg:171.72ms
step:1195/1530 train_loss:3.3690 train_time:203501ms step_avg:171.73ms
step:1196/1530 train_loss:3.3763 train_time:203687ms step_avg:171.74ms
step:1197/1530 train_loss:3.2919 train_time:203867ms step_avg:171.75ms
step:1198/1530 train_loss:3.3022 train_time:204052ms step_avg:171.76ms
step:1199/1530 train_loss:3.3410 train_time:204233ms step_avg:171.77ms
step:1200/1530 train_loss:3.4442 train_time:204411ms step_avg:171.77ms
step:1201/1530 train_loss:3.4786 train_time:204591ms step_avg:171.78ms
step:1202/1530 train_loss:3.5896 train_time:204779ms step_avg:171.79ms
step:1203/1530 train_loss:3.4049 train_time:204960ms step_avg:171.80ms
step:1204/1530 train_loss:3.3037 train_time:205140ms step_avg:171.81ms
step:1205/1530 train_loss:3.4350 train_time:205317ms step_avg:171.81ms
step:1206/1530 train_loss:3.4753 train_time:205493ms step_avg:171.82ms
step:1207/1530 train_loss:3.5123 train_time:205671ms step_avg:171.82ms
step:1208/1530 train_loss:3.3946 train_time:205845ms step_avg:171.82ms
step:1209/1530 train_loss:3.2455 train_time:206025ms step_avg:171.83ms
step:1210/1530 train_loss:3.3063 train_time:206203ms step_avg:171.84ms
step:1211/1530 train_loss:3.3926 train_time:206381ms step_avg:171.84ms
step:1212/1530 train_loss:3.3926 train_time:206558ms step_avg:171.84ms
step:1213/1530 train_loss:3.4098 train_time:206735ms step_avg:171.85ms
step:1214/1530 train_loss:3.2543 train_time:206916ms step_avg:171.86ms
step:1215/1530 train_loss:3.3912 train_time:207094ms step_avg:171.86ms
step:1216/1530 train_loss:3.3280 train_time:207273ms step_avg:171.87ms
step:1217/1530 train_loss:3.3221 train_time:207450ms step_avg:171.87ms
step:1218/1530 train_loss:3.4052 train_time:207629ms step_avg:171.88ms
step:1219/1530 train_loss:3.2489 train_time:207814ms step_avg:171.89ms
step:1220/1530 train_loss:3.4732 train_time:207990ms step_avg:171.89ms
step:1221/1530 train_loss:3.5064 train_time:208167ms step_avg:171.90ms
step:1222/1530 train_loss:3.4248 train_time:208343ms step_avg:171.90ms
step:1223/1530 train_loss:3.2998 train_time:208521ms step_avg:171.91ms
step:1224/1530 train_loss:3.2538 train_time:208701ms step_avg:171.91ms
step:1225/1530 train_loss:3.3642 train_time:208878ms step_avg:171.92ms
step:1226/1530 train_loss:3.3314 train_time:209058ms step_avg:171.92ms
step:1227/1530 train_loss:3.2785 train_time:209237ms step_avg:171.93ms
step:1228/1530 train_loss:3.4418 train_time:209412ms step_avg:171.93ms
step:1229/1530 train_loss:3.3645 train_time:209592ms step_avg:171.94ms
step:1230/1530 train_loss:3.3921 train_time:209775ms step_avg:171.95ms
step:1231/1530 train_loss:3.5813 train_time:209954ms step_avg:171.95ms
step:1232/1530 train_loss:3.4977 train_time:210136ms step_avg:171.96ms
step:1233/1530 train_loss:3.4230 train_time:210312ms step_avg:171.96ms
step:1234/1530 train_loss:3.5838 train_time:210491ms step_avg:171.97ms
step:1235/1530 train_loss:3.3211 train_time:210673ms step_avg:171.98ms
step:1236/1530 train_loss:3.2914 train_time:210849ms step_avg:171.98ms
step:1237/1530 train_loss:3.2704 train_time:211027ms step_avg:171.99ms
step:1238/1530 train_loss:3.2766 train_time:211210ms step_avg:171.99ms
step:1239/1530 train_loss:3.3323 train_time:211390ms step_avg:172.00ms
step:1240/1530 train_loss:3.3850 train_time:211568ms step_avg:172.01ms
step:1241/1530 train_loss:3.4209 train_time:211747ms step_avg:172.01ms
step:1242/1530 train_loss:3.2969 train_time:211924ms step_avg:172.02ms
step:1243/1530 train_loss:3.4038 train_time:212103ms step_avg:172.02ms
step:1244/1530 train_loss:3.4044 train_time:212276ms step_avg:172.02ms
step:1245/1530 train_loss:3.4072 train_time:212452ms step_avg:172.03ms
step:1246/1530 train_loss:3.2414 train_time:212630ms step_avg:172.03ms
step:1247/1530 train_loss:3.3691 train_time:212806ms step_avg:172.03ms
step:1248/1530 train_loss:3.4273 train_time:212983ms step_avg:172.04ms
step:1249/1530 train_loss:3.4277 train_time:213161ms step_avg:172.04ms
step:1250/1530 train_loss:3.3028 train_time:213339ms step_avg:172.05ms
step:1250/1530 val_loss:3.3533 train_time:213393ms step_avg:172.09ms
step:1251/1530 train_loss:3.4882 train_time:213523ms step_avg:172.06ms
step:1252/1530 train_loss:3.3628 train_time:213699ms step_avg:172.06ms
step:1253/1530 train_loss:3.3105 train_time:213877ms step_avg:172.07ms
step:1254/1530 train_loss:3.4128 train_time:214057ms step_avg:172.07ms
step:1255/1530 train_loss:3.5166 train_time:214246ms step_avg:172.09ms
step:1256/1530 train_loss:3.3082 train_time:214428ms step_avg:172.09ms
step:1257/1530 train_loss:3.3747 train_time:214606ms step_avg:172.10ms
step:1258/1530 train_loss:3.3638 train_time:214790ms step_avg:172.11ms
step:1259/1530 train_loss:3.3243 train_time:214968ms step_avg:172.11ms
step:1260/1530 train_loss:3.2078 train_time:215145ms step_avg:172.12ms
step:1261/1530 train_loss:3.2996 train_time:215326ms step_avg:172.12ms
step:1262/1530 train_loss:3.3246 train_time:215507ms step_avg:172.13ms
step:1263/1530 train_loss:3.2385 train_time:215689ms step_avg:172.14ms
step:1264/1530 train_loss:3.4426 train_time:215865ms step_avg:172.14ms
step:1265/1530 train_loss:3.4282 train_time:216041ms step_avg:172.14ms
step:1266/1530 train_loss:3.4369 train_time:216220ms step_avg:172.15ms
step:1267/1530 train_loss:3.3696 train_time:216400ms step_avg:172.16ms
step:1268/1530 train_loss:3.4074 train_time:216581ms step_avg:172.16ms
step:1269/1530 train_loss:3.2524 train_time:216766ms step_avg:172.17ms
step:1270/1530 train_loss:3.1029 train_time:216943ms step_avg:172.18ms
step:1271/1530 train_loss:3.4028 train_time:217122ms step_avg:172.18ms
step:1272/1530 train_loss:3.3531 train_time:217298ms step_avg:172.19ms
step:1273/1530 train_loss:3.3735 train_time:217481ms step_avg:172.19ms
step:1274/1530 train_loss:3.3528 train_time:217662ms step_avg:172.20ms
step:1275/1530 train_loss:3.4309 train_time:217840ms step_avg:172.21ms
step:1276/1530 train_loss:3.4666 train_time:218013ms step_avg:172.21ms
step:1277/1530 train_loss:3.4107 train_time:218192ms step_avg:172.21ms
step:1278/1530 train_loss:3.4061 train_time:218368ms step_avg:172.21ms
step:1279/1530 train_loss:3.2645 train_time:218549ms step_avg:172.22ms
step:1280/1530 train_loss:3.3696 train_time:218734ms step_avg:172.23ms
step:1281/1530 train_loss:3.4275 train_time:218911ms step_avg:172.23ms
step:1282/1530 train_loss:3.4694 train_time:219085ms step_avg:172.24ms
step:1283/1530 train_loss:3.3337 train_time:219266ms step_avg:172.24ms
step:1284/1530 train_loss:3.3665 train_time:219444ms step_avg:172.25ms
step:1285/1530 train_loss:3.3618 train_time:219624ms step_avg:172.25ms
step:1286/1530 train_loss:3.3339 train_time:219800ms step_avg:172.26ms
step:1287/1530 train_loss:3.4839 train_time:219979ms step_avg:172.26ms
step:1288/1530 train_loss:3.2907 train_time:220161ms step_avg:172.27ms
step:1289/1530 train_loss:3.3837 train_time:220347ms step_avg:172.28ms
step:1290/1530 train_loss:3.4597 train_time:220530ms step_avg:172.29ms
step:1291/1530 train_loss:3.3831 train_time:220710ms step_avg:172.30ms
step:1292/1530 train_loss:3.4781 train_time:220891ms step_avg:172.30ms
step:1293/1530 train_loss:3.5149 train_time:221073ms step_avg:172.31ms
step:1294/1530 train_loss:3.4580 train_time:221251ms step_avg:172.31ms
step:1295/1530 train_loss:3.2840 train_time:221431ms step_avg:172.32ms
step:1296/1530 train_loss:3.3725 train_time:221612ms step_avg:172.33ms
step:1297/1530 train_loss:3.2733 train_time:221792ms step_avg:172.33ms
step:1298/1530 train_loss:3.2728 train_time:221974ms step_avg:172.34ms
step:1299/1530 train_loss:3.3977 train_time:222152ms step_avg:172.34ms
step:1300/1530 train_loss:3.3986 train_time:222328ms step_avg:172.35ms
step:1301/1530 train_loss:3.4059 train_time:222505ms step_avg:172.35ms
step:1302/1530 train_loss:3.5734 train_time:222687ms step_avg:172.36ms
step:1303/1530 train_loss:3.3041 train_time:222870ms step_avg:172.37ms
step:1304/1530 train_loss:3.5121 train_time:223052ms step_avg:172.37ms
step:1305/1530 train_loss:3.2607 train_time:223228ms step_avg:172.38ms
step:1306/1530 train_loss:3.4495 train_time:223408ms step_avg:172.38ms
step:1307/1530 train_loss:3.4558 train_time:223583ms step_avg:172.38ms
step:1308/1530 train_loss:3.2858 train_time:223762ms step_avg:172.39ms
step:1309/1530 train_loss:3.3088 train_time:223943ms step_avg:172.40ms
step:1310/1530 train_loss:3.2881 train_time:224120ms step_avg:172.40ms
step:1311/1530 train_loss:3.2965 train_time:224300ms step_avg:172.41ms
step:1312/1530 train_loss:3.3728 train_time:224480ms step_avg:172.41ms
step:1313/1530 train_loss:3.3411 train_time:224657ms step_avg:172.42ms
step:1314/1530 train_loss:3.0447 train_time:224840ms step_avg:172.42ms
step:1315/1530 train_loss:3.2765 train_time:225017ms step_avg:172.43ms
step:1316/1530 train_loss:3.3958 train_time:225192ms step_avg:172.43ms
step:1317/1530 train_loss:3.4243 train_time:225370ms step_avg:172.43ms
step:1318/1530 train_loss:3.3018 train_time:225557ms step_avg:172.44ms
step:1319/1530 train_loss:3.4253 train_time:225737ms step_avg:172.45ms
step:1320/1530 train_loss:3.4632 train_time:225918ms step_avg:172.46ms
step:1321/1530 train_loss:3.3645 train_time:226097ms step_avg:172.46ms
step:1322/1530 train_loss:3.3252 train_time:226409ms step_avg:172.57ms
step:1323/1530 train_loss:3.3194 train_time:226598ms step_avg:172.58ms
step:1324/1530 train_loss:3.4391 train_time:226779ms step_avg:172.59ms
step:1325/1530 train_loss:3.4908 train_time:226963ms step_avg:172.60ms
step:1326/1530 train_loss:3.2143 train_time:227144ms step_avg:172.60ms
step:1327/1530 train_loss:3.1643 train_time:227319ms step_avg:172.60ms
step:1328/1530 train_loss:3.4904 train_time:227500ms step_avg:172.61ms
step:1329/1530 train_loss:3.2930 train_time:227844ms step_avg:172.74ms
step:1330/1530 train_loss:3.4320 train_time:228024ms step_avg:172.75ms
step:1331/1530 train_loss:3.3294 train_time:228199ms step_avg:172.75ms
step:1332/1530 train_loss:3.7366 train_time:228381ms step_avg:172.75ms
step:1333/1530 train_loss:3.4749 train_time:228562ms step_avg:172.76ms
step:1334/1530 train_loss:3.3709 train_time:228741ms step_avg:172.77ms
step:1335/1530 train_loss:3.2901 train_time:228920ms step_avg:172.77ms
step:1336/1530 train_loss:3.2972 train_time:229104ms step_avg:172.78ms
step:1337/1530 train_loss:3.5531 train_time:229284ms step_avg:172.78ms
step:1338/1530 train_loss:3.5193 train_time:229464ms step_avg:172.79ms
step:1339/1530 train_loss:3.3345 train_time:229644ms step_avg:172.79ms
step:1340/1530 train_loss:3.2849 train_time:229822ms step_avg:172.80ms
step:1341/1530 train_loss:3.5900 train_time:230000ms step_avg:172.80ms
step:1342/1530 train_loss:3.3572 train_time:230181ms step_avg:172.81ms
step:1343/1530 train_loss:3.3617 train_time:230357ms step_avg:172.81ms
step:1344/1530 train_loss:3.4140 train_time:230536ms step_avg:172.82ms
step:1345/1530 train_loss:3.3841 train_time:230718ms step_avg:172.82ms
step:1346/1530 train_loss:3.3012 train_time:230894ms step_avg:172.82ms
step:1347/1530 train_loss:3.2794 train_time:231070ms step_avg:172.83ms
step:1348/1530 train_loss:3.3467 train_time:231247ms step_avg:172.83ms
step:1349/1530 train_loss:3.2757 train_time:231423ms step_avg:172.83ms
step:1350/1530 train_loss:3.3896 train_time:231603ms step_avg:172.84ms
step:1351/1530 train_loss:3.2438 train_time:231781ms step_avg:172.84ms
step:1352/1530 train_loss:3.3071 train_time:231959ms step_avg:172.85ms
step:1353/1530 train_loss:3.4061 train_time:232138ms step_avg:172.85ms
step:1354/1530 train_loss:3.2599 train_time:232314ms step_avg:172.85ms
step:1355/1530 train_loss:3.1885 train_time:232491ms step_avg:172.86ms
step:1356/1530 train_loss:3.5086 train_time:232671ms step_avg:172.86ms
step:1357/1530 train_loss:3.4234 train_time:232850ms step_avg:172.87ms
step:1358/1530 train_loss:3.1857 train_time:233027ms step_avg:172.87ms
step:1359/1530 train_loss:3.4447 train_time:233207ms step_avg:172.87ms
step:1360/1530 train_loss:3.3500 train_time:233386ms step_avg:172.88ms
step:1361/1530 train_loss:3.1251 train_time:233573ms step_avg:172.89ms
step:1362/1530 train_loss:3.3943 train_time:233753ms step_avg:172.89ms
step:1363/1530 train_loss:3.2874 train_time:233939ms step_avg:172.90ms
step:1364/1530 train_loss:3.3014 train_time:234115ms step_avg:172.91ms
step:1365/1530 train_loss:3.3147 train_time:234292ms step_avg:172.91ms
step:1366/1530 train_loss:3.4197 train_time:234473ms step_avg:172.92ms
step:1367/1530 train_loss:3.3977 train_time:234650ms step_avg:172.92ms
step:1368/1530 train_loss:3.3446 train_time:234829ms step_avg:172.92ms
step:1369/1530 train_loss:3.2785 train_time:235018ms step_avg:172.93ms
step:1370/1530 train_loss:3.6060 train_time:235197ms step_avg:172.94ms
step:1371/1530 train_loss:3.3150 train_time:235380ms step_avg:172.95ms
step:1372/1530 train_loss:3.3698 train_time:235564ms step_avg:172.95ms
step:1373/1530 train_loss:3.3634 train_time:235744ms step_avg:172.96ms
step:1374/1530 train_loss:3.1518 train_time:235925ms step_avg:172.97ms
step:1375/1530 train_loss:3.5335 train_time:236105ms step_avg:172.97ms
step:1375/1530 val_loss:3.3112 train_time:236155ms step_avg:173.01ms
step:1376/1530 train_loss:3.3472 train_time:236285ms step_avg:172.98ms
step:1377/1530 train_loss:3.4790 train_time:236465ms step_avg:172.98ms
step:1378/1530 train_loss:3.4669 train_time:236642ms step_avg:172.98ms
step:1379/1530 train_loss:3.1100 train_time:236823ms step_avg:172.99ms
step:1380/1530 train_loss:3.3090 train_time:237004ms step_avg:173.00ms
step:1381/1530 train_loss:3.6911 train_time:237189ms step_avg:173.00ms
step:1382/1530 train_loss:3.2079 train_time:237367ms step_avg:173.01ms
step:1383/1530 train_loss:3.3924 train_time:237549ms step_avg:173.01ms
step:1384/1530 train_loss:3.4754 train_time:237730ms step_avg:173.02ms
step:1385/1530 train_loss:3.4048 train_time:237905ms step_avg:173.02ms
step:1386/1530 train_loss:3.3425 train_time:238085ms step_avg:173.03ms
step:1387/1530 train_loss:3.1946 train_time:238265ms step_avg:173.03ms
step:1388/1530 train_loss:3.3441 train_time:238440ms step_avg:173.03ms
step:1389/1530 train_loss:3.3141 train_time:238623ms step_avg:173.04ms
step:1390/1530 train_loss:3.5668 train_time:238801ms step_avg:173.04ms
step:1391/1530 train_loss:3.2958 train_time:238976ms step_avg:173.05ms
step:1392/1530 train_loss:3.2866 train_time:239156ms step_avg:173.05ms
step:1393/1530 train_loss:3.2382 train_time:239335ms step_avg:173.06ms
step:1394/1530 train_loss:3.4988 train_time:239513ms step_avg:173.06ms
step:1395/1530 train_loss:3.3898 train_time:239691ms step_avg:173.06ms
step:1396/1530 train_loss:3.4054 train_time:239870ms step_avg:173.07ms
step:1397/1530 train_loss:3.3093 train_time:240046ms step_avg:173.07ms
step:1398/1530 train_loss:3.2558 train_time:240222ms step_avg:173.07ms
step:1399/1530 train_loss:3.3156 train_time:240399ms step_avg:173.07ms
step:1400/1530 train_loss:3.3195 train_time:240581ms step_avg:173.08ms
step:1401/1530 train_loss:3.3490 train_time:240758ms step_avg:173.08ms
step:1402/1530 train_loss:3.3017 train_time:240935ms step_avg:173.09ms
step:1403/1530 train_loss:3.4995 train_time:241120ms step_avg:173.09ms
step:1404/1530 train_loss:3.2811 train_time:241296ms step_avg:173.10ms
step:1405/1530 train_loss:3.3164 train_time:241478ms step_avg:173.10ms
step:1406/1530 train_loss:3.3149 train_time:241660ms step_avg:173.11ms
step:1407/1530 train_loss:3.1761 train_time:241836ms step_avg:173.11ms
step:1408/1530 train_loss:3.3076 train_time:242016ms step_avg:173.12ms
step:1409/1530 train_loss:3.3021 train_time:242203ms step_avg:173.13ms
step:1410/1530 train_loss:3.2871 train_time:242381ms step_avg:173.13ms
step:1411/1530 train_loss:3.3637 train_time:242556ms step_avg:173.13ms
step:1412/1530 train_loss:3.3353 train_time:242734ms step_avg:173.13ms
step:1413/1530 train_loss:3.3590 train_time:242913ms step_avg:173.14ms
step:1414/1530 train_loss:3.3301 train_time:243094ms step_avg:173.14ms
step:1415/1530 train_loss:3.4073 train_time:243280ms step_avg:173.15ms
step:1416/1530 train_loss:3.2285 train_time:243468ms step_avg:173.16ms
step:1417/1530 train_loss:3.2804 train_time:243653ms step_avg:173.17ms
step:1418/1530 train_loss:3.3877 train_time:243833ms step_avg:173.18ms
step:1419/1530 train_loss:3.3462 train_time:244015ms step_avg:173.18ms
step:1420/1530 train_loss:3.3654 train_time:244197ms step_avg:173.19ms
step:1421/1530 train_loss:3.3727 train_time:244376ms step_avg:173.19ms
step:1422/1530 train_loss:3.3317 train_time:244553ms step_avg:173.20ms
step:1423/1530 train_loss:3.3140 train_time:244733ms step_avg:173.20ms
step:1424/1530 train_loss:3.3312 train_time:244916ms step_avg:173.21ms
step:1425/1530 train_loss:3.1906 train_time:245103ms step_avg:173.22ms
step:1426/1530 train_loss:3.3218 train_time:245280ms step_avg:173.22ms
step:1427/1530 train_loss:3.2821 train_time:245463ms step_avg:173.23ms
step:1428/1530 train_loss:3.3774 train_time:245639ms step_avg:173.23ms
step:1429/1530 train_loss:3.3536 train_time:245815ms step_avg:173.23ms
step:1430/1530 train_loss:3.2580 train_time:245995ms step_avg:173.24ms
step:1431/1530 train_loss:3.3244 train_time:246177ms step_avg:173.24ms
step:1432/1530 train_loss:3.3369 train_time:246359ms step_avg:173.25ms
step:1433/1530 train_loss:3.1280 train_time:246541ms step_avg:173.25ms
step:1434/1530 train_loss:3.2868 train_time:246726ms step_avg:173.26ms
step:1435/1530 train_loss:3.1180 train_time:246906ms step_avg:173.27ms
step:1436/1530 train_loss:3.2319 train_time:247086ms step_avg:173.27ms
step:1437/1530 train_loss:3.4020 train_time:247264ms step_avg:173.28ms
step:1438/1530 train_loss:3.3817 train_time:247441ms step_avg:173.28ms
step:1439/1530 train_loss:3.3148 train_time:247622ms step_avg:173.28ms
step:1440/1530 train_loss:3.1897 train_time:247796ms step_avg:173.28ms
step:1441/1530 train_loss:3.3386 train_time:247975ms step_avg:173.29ms
step:1442/1530 train_loss:3.3874 train_time:248160ms step_avg:173.30ms
step:1443/1530 train_loss:3.4941 train_time:248349ms step_avg:173.31ms
step:1444/1530 train_loss:3.4484 train_time:248526ms step_avg:173.31ms
step:1445/1530 train_loss:3.3382 train_time:248704ms step_avg:173.31ms
step:1446/1530 train_loss:3.1991 train_time:248884ms step_avg:173.32ms
step:1447/1530 train_loss:3.2974 train_time:249064ms step_avg:173.32ms
step:1448/1530 train_loss:3.2968 train_time:249244ms step_avg:173.33ms
step:1449/1530 train_loss:3.3943 train_time:249422ms step_avg:173.33ms
step:1450/1530 train_loss:3.3914 train_time:249601ms step_avg:173.33ms
step:1451/1530 train_loss:3.2012 train_time:249779ms step_avg:173.34ms
step:1452/1530 train_loss:3.3272 train_time:249959ms step_avg:173.34ms
step:1453/1530 train_loss:3.2540 train_time:250134ms step_avg:173.34ms
step:1454/1530 train_loss:3.2873 train_time:250312ms step_avg:173.35ms
step:1455/1530 train_loss:3.3247 train_time:250496ms step_avg:173.35ms
step:1456/1530 train_loss:3.2835 train_time:250673ms step_avg:173.36ms
step:1457/1530 train_loss:3.1548 train_time:250851ms step_avg:173.36ms
step:1458/1530 train_loss:3.4220 train_time:251030ms step_avg:173.36ms
step:1459/1530 train_loss:3.2727 train_time:251214ms step_avg:173.37ms
step:1460/1530 train_loss:3.3158 train_time:251394ms step_avg:173.37ms
step:1461/1530 train_loss:3.4269 train_time:251573ms step_avg:173.38ms
step:1462/1530 train_loss:3.2591 train_time:251750ms step_avg:173.38ms
step:1463/1530 train_loss:3.4648 train_time:251934ms step_avg:173.39ms
step:1464/1530 train_loss:3.3612 train_time:252113ms step_avg:173.39ms
step:1465/1530 train_loss:3.3545 train_time:252294ms step_avg:173.40ms
step:1466/1530 train_loss:3.2849 train_time:252473ms step_avg:173.40ms
step:1467/1530 train_loss:3.3940 train_time:252654ms step_avg:173.41ms
step:1468/1530 train_loss:3.2859 train_time:252831ms step_avg:173.41ms
step:1469/1530 train_loss:3.2717 train_time:253011ms step_avg:173.41ms
step:1470/1530 train_loss:3.3308 train_time:253193ms step_avg:173.42ms
step:1471/1530 train_loss:3.2581 train_time:253381ms step_avg:173.43ms
step:1472/1530 train_loss:3.2446 train_time:253566ms step_avg:173.44ms
step:1473/1530 train_loss:3.4444 train_time:253744ms step_avg:173.44ms
step:1474/1530 train_loss:3.3142 train_time:253927ms step_avg:173.45ms
step:1475/1530 train_loss:3.1506 train_time:254112ms step_avg:173.46ms
step:1476/1530 train_loss:3.2660 train_time:254290ms step_avg:173.46ms
step:1477/1530 train_loss:3.2383 train_time:254476ms step_avg:173.47ms
step:1478/1530 train_loss:3.3118 train_time:254662ms step_avg:173.48ms
step:1479/1530 train_loss:3.4022 train_time:254846ms step_avg:173.48ms
step:1480/1530 train_loss:3.2673 train_time:255023ms step_avg:173.48ms
step:1481/1530 train_loss:3.4524 train_time:255203ms step_avg:173.49ms
step:1482/1530 train_loss:3.3670 train_time:255390ms step_avg:173.50ms
step:1483/1530 train_loss:3.2772 train_time:255580ms step_avg:173.51ms
step:1484/1530 train_loss:3.2649 train_time:255768ms step_avg:173.52ms
step:1485/1530 train_loss:3.2810 train_time:255948ms step_avg:173.52ms
step:1486/1530 train_loss:3.2252 train_time:256133ms step_avg:173.53ms
step:1487/1530 train_loss:3.3400 train_time:256315ms step_avg:173.54ms
step:1488/1530 train_loss:3.2425 train_time:256498ms step_avg:173.54ms
step:1489/1530 train_loss:3.3145 train_time:256678ms step_avg:173.55ms
step:1490/1530 train_loss:3.2524 train_time:256859ms step_avg:173.55ms
step:1491/1530 train_loss:3.1608 train_time:257039ms step_avg:173.56ms
step:1492/1530 train_loss:3.2668 train_time:257218ms step_avg:173.56ms
step:1493/1530 train_loss:3.4351 train_time:257397ms step_avg:173.57ms
step:1494/1530 train_loss:3.2982 train_time:257575ms step_avg:173.57ms
step:1495/1530 train_loss:3.0341 train_time:257760ms step_avg:173.58ms
step:1496/1530 train_loss:3.3616 train_time:257942ms step_avg:173.58ms
step:1497/1530 train_loss:3.3110 train_time:258127ms step_avg:173.59ms
step:1498/1530 train_loss:3.3462 train_time:258312ms step_avg:173.60ms
step:1499/1530 train_loss:3.3143 train_time:258501ms step_avg:173.61ms
step:1500/1530 train_loss:3.2958 train_time:258694ms step_avg:173.62ms
step:1500/1530 val_loss:3.2803 train_time:258750ms step_avg:173.66ms
step:1501/1530 train_loss:3.0916 train_time:258884ms step_avg:173.63ms
step:1502/1530 train_loss:3.3605 train_time:259076ms step_avg:173.64ms
step:1503/1530 train_loss:3.2411 train_time:259256ms step_avg:173.65ms
step:1504/1530 train_loss:3.2485 train_time:259438ms step_avg:173.65ms
step:1505/1530 train_loss:3.2096 train_time:259617ms step_avg:173.66ms
step:1506/1530 train_loss:3.2770 train_time:259799ms step_avg:173.66ms
step:1507/1530 train_loss:3.1794 train_time:259993ms step_avg:173.68ms
step:1508/1530 train_loss:3.4845 train_time:260176ms step_avg:173.68ms
step:1509/1530 train_loss:3.2800 train_time:260354ms step_avg:173.68ms
step:1510/1530 train_loss:3.2732 train_time:260534ms step_avg:173.69ms
step:1511/1530 train_loss:3.4148 train_time:260854ms step_avg:173.79ms
step:1512/1530 train_loss:3.4215 train_time:261040ms step_avg:173.79ms
step:1513/1530 train_loss:3.2733 train_time:261225ms step_avg:173.80ms
step:1514/1530 train_loss:3.0824 train_time:261408ms step_avg:173.81ms
step:1515/1530 train_loss:3.2396 train_time:261589ms step_avg:173.81ms
step:1516/1530 train_loss:3.2560 train_time:261776ms step_avg:173.82ms
step:1517/1530 train_loss:3.3005 train_time:261957ms step_avg:173.83ms
step:1518/1530 train_loss:3.2061 train_time:262142ms step_avg:173.83ms
step:1519/1530 train_loss:3.5107 train_time:262475ms step_avg:173.94ms
step:1520/1530 train_loss:3.1263 train_time:262660ms step_avg:173.95ms
step:1521/1530 train_loss:3.2018 train_time:262836ms step_avg:173.95ms
step:1522/1530 train_loss:3.3525 train_time:263021ms step_avg:173.96ms
step:1523/1530 train_loss:3.2290 train_time:263198ms step_avg:173.96ms
step:1524/1530 train_loss:3.3507 train_time:263377ms step_avg:173.96ms
step:1525/1530 train_loss:3.3378 train_time:263564ms step_avg:173.97ms
step:1526/1530 train_loss:3.2807 train_time:263757ms step_avg:173.98ms
step:1527/1530 train_loss:3.2910 train_time:263938ms step_avg:173.99ms
step:1528/1530 train_loss:3.4052 train_time:264117ms step_avg:173.99ms
step:1529/1530 train_loss:3.4076 train_time:264296ms step_avg:173.99ms
step:1530/1530 train_loss:3.2364 train_time:264474ms step_avg:174.00ms
step:1530/1530 val_loss:3.2779 train_time:264528ms step_avg:174.03ms