e3ffb000-a388-4bdc-b88f-2f85794c13c8.txt

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

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

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

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

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

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

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

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

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

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

    def step(self):

        for group in self.param_groups:

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

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

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

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

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

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

class CastedLinear(nn.Linear):

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

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

class Rotary(torch.nn.Module):

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

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

class CausalSelfAttention(nn.Module):

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

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

class MLP(nn.Module):

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

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

class Block(nn.Module):

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

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

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

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

class GPT(nn.Module):

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

        self.reset()

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

# -------------------------------------------------------------------------
# clean up nice
dist.destroy_process_group()
====================================================================================================
Running pytorch 2.6.0.dev20241203+cu124 compiled for CUDA 12.4
nvidia-smi:
Thu Dec  5 01:10:18 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.06             Driver Version: 535.183.06   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA H100 80GB HBM3          On  | 00000000:19:00.0 Off |                    0 |
| N/A   38C    P0              75W / 700W |      3MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          On  | 00000000:3B:00.0 Off |                    0 |
| N/A   30C    P0              93W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          On  | 00000000:4C:00.0 Off |                    0 |
| N/A   31C    P0             118W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          On  | 00000000:5D:00.0 Off |                    0 |
| N/A   38C    P0             118W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          On  | 00000000:9B:00.0 Off |                    0 |
| N/A   38C    P0             102W / 700W |     22MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          On  | 00000000:BB:00.0 Off |                    0 |
| N/A   29C    P0             106W / 700W |     35MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          On  | 00000000:CB:00.0 Off |                    0 |
| N/A   38C    P0             127W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          On  | 00000000:DB:00.0 Off |                    0 |
| N/A   30C    P0             118W / 700W |    529MiB / 81559MiB |      0%      Default |
|                                         |                      |             Disabled |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
+---------------------------------------------------------------------------------------+

====================================================================================================
Training DataLoader: total number of tokens: 1100000000 across 11 files
Validation DataLoader: total number of tokens: 100000000 across 1 files
====================================================================================================
step:0/1530 val_loss:10.8258 train_time:0ms step_avg:nanms
step:1/1530 train_loss:10.8258 train_time:31981ms step_avg:nanms
step:2/1530 train_loss:10.0771 train_time:32092ms step_avg:nanms
step:3/1530 train_loss:8.3927 train_time:32252ms step_avg:nanms
step:4/1530 train_loss:7.5369 train_time:32414ms step_avg:nanms
step:5/1530 train_loss:7.4896 train_time:32574ms step_avg:nanms
step:6/1530 train_loss:6.9861 train_time:32734ms step_avg:nanms
step:7/1530 train_loss:7.2268 train_time:32895ms step_avg:nanms
step:8/1530 train_loss:6.7293 train_time:33054ms step_avg:nanms
step:9/1530 train_loss:6.6153 train_time:33215ms step_avg:nanms
step:10/1530 train_loss:6.4809 train_time:33375ms step_avg:nanms
step:11/1530 train_loss:6.4004 train_time:114ms step_avg:nanms
step:12/1530 train_loss:6.3470 train_time:275ms step_avg:nanms
step:13/1530 train_loss:6.2470 train_time:435ms step_avg:145.09ms
step:14/1530 train_loss:6.2519 train_time:595ms step_avg:148.74ms
step:15/1530 train_loss:6.1752 train_time:755ms step_avg:151.05ms
step:16/1530 train_loss:6.1295 train_time:915ms step_avg:152.58ms
step:17/1530 train_loss:6.1590 train_time:1075ms step_avg:153.64ms
step:18/1530 train_loss:5.9676 train_time:1235ms step_avg:154.41ms
step:19/1530 train_loss:5.9639 train_time:1395ms step_avg:154.95ms
step:20/1530 train_loss:5.6878 train_time:1555ms step_avg:155.51ms
step:21/1530 train_loss:5.9665 train_time:1715ms step_avg:155.91ms
step:22/1530 train_loss:6.1679 train_time:1875ms step_avg:156.28ms
step:23/1530 train_loss:5.8430 train_time:2036ms step_avg:156.59ms
step:24/1530 train_loss:6.0268 train_time:2195ms step_avg:156.82ms
step:25/1530 train_loss:5.6730 train_time:2356ms step_avg:157.09ms
step:26/1530 train_loss:5.5929 train_time:2515ms step_avg:157.20ms
step:27/1530 train_loss:5.7641 train_time:2676ms step_avg:157.41ms
step:28/1530 train_loss:5.4154 train_time:2835ms step_avg:157.51ms
step:29/1530 train_loss:5.6704 train_time:2995ms step_avg:157.65ms
step:30/1530 train_loss:5.4718 train_time:3155ms step_avg:157.76ms
step:31/1530 train_loss:5.4350 train_time:3315ms step_avg:157.87ms
step:32/1530 train_loss:5.2877 train_time:3475ms step_avg:157.97ms
step:33/1530 train_loss:5.5826 train_time:3635ms step_avg:158.05ms
step:34/1530 train_loss:5.5012 train_time:3795ms step_avg:158.13ms
step:35/1530 train_loss:5.6052 train_time:3956ms step_avg:158.22ms
step:36/1530 train_loss:5.5461 train_time:4115ms step_avg:158.28ms
step:37/1530 train_loss:5.4676 train_time:4275ms step_avg:158.34ms
step:38/1530 train_loss:5.3340 train_time:4435ms step_avg:158.41ms
step:39/1530 train_loss:5.3285 train_time:4595ms step_avg:158.45ms
step:40/1530 train_loss:5.2440 train_time:4755ms step_avg:158.50ms
step:41/1530 train_loss:5.2207 train_time:4915ms step_avg:158.54ms
step:42/1530 train_loss:5.1657 train_time:5075ms step_avg:158.59ms
step:43/1530 train_loss:5.2599 train_time:5234ms step_avg:158.62ms
step:44/1530 train_loss:5.2330 train_time:5395ms step_avg:158.68ms
step:45/1530 train_loss:5.3750 train_time:5555ms step_avg:158.71ms
step:46/1530 train_loss:5.1621 train_time:5714ms step_avg:158.73ms
step:47/1530 train_loss:5.0555 train_time:5874ms step_avg:158.76ms
step:48/1530 train_loss:5.2207 train_time:6034ms step_avg:158.79ms
step:49/1530 train_loss:5.1396 train_time:6194ms step_avg:158.82ms
step:50/1530 train_loss:5.2508 train_time:6354ms step_avg:158.84ms
step:51/1530 train_loss:5.1477 train_time:6514ms step_avg:158.88ms
step:52/1530 train_loss:5.0263 train_time:6675ms step_avg:158.92ms
step:53/1530 train_loss:5.1506 train_time:6835ms step_avg:158.94ms
step:54/1530 train_loss:4.9954 train_time:6994ms step_avg:158.96ms
step:55/1530 train_loss:5.4079 train_time:7155ms step_avg:159.00ms
step:56/1530 train_loss:5.0331 train_time:7315ms step_avg:159.02ms
step:57/1530 train_loss:4.8812 train_time:7474ms step_avg:159.03ms
step:58/1530 train_loss:5.0330 train_time:7635ms step_avg:159.05ms
step:59/1530 train_loss:5.0160 train_time:7795ms step_avg:159.08ms
step:60/1530 train_loss:5.1310 train_time:7954ms step_avg:159.09ms
step:61/1530 train_loss:4.8581 train_time:8114ms step_avg:159.10ms
step:62/1530 train_loss:4.9959 train_time:8275ms step_avg:159.14ms
step:63/1530 train_loss:4.9767 train_time:8435ms step_avg:159.15ms
step:64/1530 train_loss:4.9138 train_time:8595ms step_avg:159.16ms
step:65/1530 train_loss:4.8184 train_time:8756ms step_avg:159.19ms
step:66/1530 train_loss:4.9251 train_time:8915ms step_avg:159.20ms
step:67/1530 train_loss:4.8355 train_time:9076ms step_avg:159.23ms
step:68/1530 train_loss:5.0816 train_time:9235ms step_avg:159.23ms
step:69/1530 train_loss:4.7244 train_time:9396ms step_avg:159.25ms
step:70/1530 train_loss:4.8450 train_time:9556ms step_avg:159.26ms
step:71/1530 train_loss:4.9632 train_time:9716ms step_avg:159.27ms
step:72/1530 train_loss:4.8883 train_time:9876ms step_avg:159.29ms
step:73/1530 train_loss:4.7736 train_time:10035ms step_avg:159.29ms
step:74/1530 train_loss:4.9084 train_time:10195ms step_avg:159.30ms
step:75/1530 train_loss:4.8638 train_time:10355ms step_avg:159.30ms
step:76/1530 train_loss:4.8103 train_time:10514ms step_avg:159.31ms
step:77/1530 train_loss:4.9044 train_time:10674ms step_avg:159.32ms
step:78/1530 train_loss:5.1141 train_time:10834ms step_avg:159.33ms
step:79/1530 train_loss:4.8234 train_time:10995ms step_avg:159.35ms
step:80/1530 train_loss:4.8678 train_time:11155ms step_avg:159.35ms
step:81/1530 train_loss:4.6525 train_time:11314ms step_avg:159.35ms
step:82/1530 train_loss:4.8188 train_time:11475ms step_avg:159.37ms
step:83/1530 train_loss:4.7837 train_time:11634ms step_avg:159.38ms
step:84/1530 train_loss:4.7661 train_time:11794ms step_avg:159.38ms
step:85/1530 train_loss:4.6197 train_time:11954ms step_avg:159.39ms
step:86/1530 train_loss:4.8295 train_time:12114ms step_avg:159.40ms
step:87/1530 train_loss:4.7413 train_time:12274ms step_avg:159.41ms
step:88/1530 train_loss:4.7428 train_time:12435ms step_avg:159.42ms
step:89/1530 train_loss:4.6960 train_time:12595ms step_avg:159.43ms
step:90/1530 train_loss:4.6357 train_time:12755ms step_avg:159.44ms
step:91/1530 train_loss:4.6301 train_time:12915ms step_avg:159.45ms
step:92/1530 train_loss:4.7844 train_time:13075ms step_avg:159.45ms
step:93/1530 train_loss:4.6136 train_time:13235ms step_avg:159.46ms
step:94/1530 train_loss:4.6393 train_time:13395ms step_avg:159.46ms
step:95/1530 train_loss:4.6795 train_time:13555ms step_avg:159.47ms
step:96/1530 train_loss:4.5871 train_time:13715ms step_avg:159.47ms
step:97/1530 train_loss:4.6613 train_time:13875ms step_avg:159.48ms
step:98/1530 train_loss:4.5870 train_time:14034ms step_avg:159.48ms
step:99/1530 train_loss:4.6634 train_time:14195ms step_avg:159.49ms
step:100/1530 train_loss:4.6855 train_time:14355ms step_avg:159.50ms
step:101/1530 train_loss:4.5272 train_time:14515ms step_avg:159.50ms
step:102/1530 train_loss:4.6914 train_time:14675ms step_avg:159.52ms
step:103/1530 train_loss:4.5701 train_time:14835ms step_avg:159.52ms
step:104/1530 train_loss:4.5405 train_time:14994ms step_avg:159.52ms
step:105/1530 train_loss:4.5592 train_time:15155ms step_avg:159.52ms
step:106/1530 train_loss:4.6133 train_time:15314ms step_avg:159.52ms
step:107/1530 train_loss:4.4952 train_time:15475ms step_avg:159.53ms
step:108/1530 train_loss:4.3617 train_time:15634ms step_avg:159.53ms
step:109/1530 train_loss:4.4887 train_time:15795ms step_avg:159.54ms
step:110/1530 train_loss:4.4909 train_time:15955ms step_avg:159.55ms
step:111/1530 train_loss:4.4353 train_time:16115ms step_avg:159.55ms
step:112/1530 train_loss:4.5828 train_time:16275ms step_avg:159.56ms
step:113/1530 train_loss:4.4801 train_time:16434ms step_avg:159.56ms
step:114/1530 train_loss:4.3640 train_time:16595ms step_avg:159.57ms
step:115/1530 train_loss:4.5045 train_time:16758ms step_avg:159.60ms
step:116/1530 train_loss:4.4688 train_time:16920ms step_avg:159.62ms
step:117/1530 train_loss:4.3517 train_time:17085ms step_avg:159.67ms
step:118/1530 train_loss:4.5853 train_time:17250ms step_avg:159.72ms
step:119/1530 train_loss:4.4544 train_time:17414ms step_avg:159.76ms
step:120/1530 train_loss:4.3338 train_time:17577ms step_avg:159.79ms
step:121/1530 train_loss:4.2971 train_time:17740ms step_avg:159.82ms
step:122/1530 train_loss:4.4437 train_time:17902ms step_avg:159.84ms
step:123/1530 train_loss:4.2749 train_time:18068ms step_avg:159.90ms
step:124/1530 train_loss:4.5823 train_time:18232ms step_avg:159.93ms
step:125/1530 train_loss:4.4413 train_time:18396ms step_avg:159.97ms
step:125/1530 val_loss:4.3898 train_time:18443ms step_avg:160.37ms
step:126/1530 train_loss:4.4030 train_time:18561ms step_avg:160.01ms
step:127/1530 train_loss:4.4256 train_time:18726ms step_avg:160.05ms
step:128/1530 train_loss:4.3737 train_time:18889ms step_avg:160.08ms
step:129/1530 train_loss:4.6909 train_time:19053ms step_avg:160.11ms
step:130/1530 train_loss:4.3811 train_time:19219ms step_avg:160.16ms
step:131/1530 train_loss:4.4027 train_time:19382ms step_avg:160.19ms
step:132/1530 train_loss:4.3468 train_time:19546ms step_avg:160.21ms
step:133/1530 train_loss:4.4471 train_time:19709ms step_avg:160.24ms
step:134/1530 train_loss:4.2768 train_time:19873ms step_avg:160.27ms
step:135/1530 train_loss:4.4460 train_time:20038ms step_avg:160.31ms
step:136/1530 train_loss:4.2152 train_time:20202ms step_avg:160.33ms
step:137/1530 train_loss:4.3706 train_time:20365ms step_avg:160.35ms
step:138/1530 train_loss:4.2799 train_time:20527ms step_avg:160.37ms
step:139/1530 train_loss:4.3767 train_time:20692ms step_avg:160.41ms
step:140/1530 train_loss:4.4787 train_time:20857ms step_avg:160.44ms
step:141/1530 train_loss:4.3116 train_time:21021ms step_avg:160.47ms
step:142/1530 train_loss:4.3055 train_time:21185ms step_avg:160.49ms
step:143/1530 train_loss:4.2524 train_time:21349ms step_avg:160.52ms
step:144/1530 train_loss:4.3510 train_time:21512ms step_avg:160.54ms
step:145/1530 train_loss:4.3013 train_time:21677ms step_avg:160.57ms
step:146/1530 train_loss:4.1694 train_time:21840ms step_avg:160.59ms
step:147/1530 train_loss:4.3236 train_time:22003ms step_avg:160.61ms
step:148/1530 train_loss:4.3595 train_time:22166ms step_avg:160.63ms
step:149/1530 train_loss:4.3045 train_time:22330ms step_avg:160.65ms
step:150/1530 train_loss:4.4486 train_time:22494ms step_avg:160.67ms
step:151/1530 train_loss:4.2674 train_time:22658ms step_avg:160.70ms
step:152/1530 train_loss:4.2700 train_time:22822ms step_avg:160.72ms
step:153/1530 train_loss:4.3550 train_time:22985ms step_avg:160.73ms
step:154/1530 train_loss:4.3664 train_time:23150ms step_avg:160.76ms
step:155/1530 train_loss:4.2766 train_time:23313ms step_avg:160.78ms
step:156/1530 train_loss:4.3524 train_time:23477ms step_avg:160.80ms
step:157/1530 train_loss:4.4067 train_time:23641ms step_avg:160.82ms
step:158/1530 train_loss:4.2387 train_time:23804ms step_avg:160.83ms
step:159/1530 train_loss:4.3014 train_time:23966ms step_avg:160.85ms
step:160/1530 train_loss:4.1300 train_time:24131ms step_avg:160.87ms
step:161/1530 train_loss:4.3449 train_time:24295ms step_avg:160.89ms
step:162/1530 train_loss:4.3544 train_time:24459ms step_avg:160.91ms
step:163/1530 train_loss:4.3413 train_time:24622ms step_avg:160.93ms
step:164/1530 train_loss:4.1948 train_time:24786ms step_avg:160.95ms
step:165/1530 train_loss:4.2940 train_time:24950ms step_avg:160.97ms
step:166/1530 train_loss:4.3715 train_time:25114ms step_avg:160.99ms
step:167/1530 train_loss:4.2291 train_time:25278ms step_avg:161.00ms
step:168/1530 train_loss:4.3001 train_time:25442ms step_avg:161.03ms
step:169/1530 train_loss:4.1631 train_time:25605ms step_avg:161.03ms
step:170/1530 train_loss:4.0222 train_time:25768ms step_avg:161.05ms
step:171/1530 train_loss:4.2107 train_time:25932ms step_avg:161.07ms
step:172/1530 train_loss:4.2141 train_time:26094ms step_avg:161.07ms
step:173/1530 train_loss:4.2686 train_time:26258ms step_avg:161.09ms
step:174/1530 train_loss:4.4268 train_time:26421ms step_avg:161.10ms
step:175/1530 train_loss:4.2492 train_time:26583ms step_avg:161.11ms
step:176/1530 train_loss:4.0938 train_time:26746ms step_avg:161.12ms
step:177/1530 train_loss:4.0607 train_time:26908ms step_avg:161.12ms
step:178/1530 train_loss:4.1830 train_time:27070ms step_avg:161.13ms
step:179/1530 train_loss:4.1277 train_time:27234ms step_avg:161.15ms
step:180/1530 train_loss:4.1101 train_time:27396ms step_avg:161.15ms
step:181/1530 train_loss:4.2975 train_time:27560ms step_avg:161.17ms
step:182/1530 train_loss:4.1578 train_time:27722ms step_avg:161.17ms
step:183/1530 train_loss:4.1398 train_time:27883ms step_avg:161.18ms
step:184/1530 train_loss:4.1309 train_time:28046ms step_avg:161.18ms
step:185/1530 train_loss:4.2176 train_time:28207ms step_avg:161.18ms
step:186/1530 train_loss:4.1796 train_time:28370ms step_avg:161.19ms
step:187/1530 train_loss:4.2313 train_time:28536ms step_avg:161.22ms
step:188/1530 train_loss:4.1666 train_time:28839ms step_avg:162.02ms
step:189/1530 train_loss:4.1092 train_time:29189ms step_avg:163.07ms
step:190/1530 train_loss:4.2087 train_time:29362ms step_avg:163.12ms
step:191/1530 train_loss:4.0865 train_time:29525ms step_avg:163.12ms
step:192/1530 train_loss:4.0342 train_time:29687ms step_avg:163.11ms
step:193/1530 train_loss:4.2573 train_time:29850ms step_avg:163.12ms
step:194/1530 train_loss:4.1811 train_time:30013ms step_avg:163.12ms
step:195/1530 train_loss:4.3551 train_time:30177ms step_avg:163.12ms
step:196/1530 train_loss:4.1860 train_time:30339ms step_avg:163.12ms
step:197/1530 train_loss:4.0530 train_time:30501ms step_avg:163.11ms
step:198/1530 train_loss:4.1832 train_time:30663ms step_avg:163.10ms
step:199/1530 train_loss:4.0387 train_time:30826ms step_avg:163.10ms
step:200/1530 train_loss:4.1168 train_time:30989ms step_avg:163.10ms
step:201/1530 train_loss:4.0088 train_time:31153ms step_avg:163.11ms
step:202/1530 train_loss:4.2559 train_time:31316ms step_avg:163.11ms
step:203/1530 train_loss:4.0672 train_time:31480ms step_avg:163.11ms
step:204/1530 train_loss:4.1955 train_time:31642ms step_avg:163.10ms
step:205/1530 train_loss:4.2527 train_time:31804ms step_avg:163.10ms
step:206/1530 train_loss:3.9500 train_time:31967ms step_avg:163.10ms
step:207/1530 train_loss:4.0832 train_time:32130ms step_avg:163.10ms
step:208/1530 train_loss:4.1040 train_time:32292ms step_avg:163.09ms
step:209/1530 train_loss:4.2440 train_time:32456ms step_avg:163.10ms
step:210/1530 train_loss:4.1810 train_time:32618ms step_avg:163.09ms
step:211/1530 train_loss:4.0635 train_time:32782ms step_avg:163.09ms
step:212/1530 train_loss:4.1226 train_time:32945ms step_avg:163.10ms
step:213/1530 train_loss:4.0634 train_time:33108ms step_avg:163.09ms
step:214/1530 train_loss:4.1194 train_time:33269ms step_avg:163.08ms
step:215/1530 train_loss:3.9741 train_time:33433ms step_avg:163.09ms
step:216/1530 train_loss:4.0125 train_time:33597ms step_avg:163.09ms
step:217/1530 train_loss:4.0161 train_time:33760ms step_avg:163.09ms
step:218/1530 train_loss:4.0866 train_time:33922ms step_avg:163.09ms
step:219/1530 train_loss:4.0764 train_time:34085ms step_avg:163.08ms
step:220/1530 train_loss:4.0923 train_time:34246ms step_avg:163.08ms
step:221/1530 train_loss:4.0980 train_time:34410ms step_avg:163.08ms
step:222/1530 train_loss:4.0001 train_time:34573ms step_avg:163.08ms
step:223/1530 train_loss:3.9842 train_time:34738ms step_avg:163.09ms
step:224/1530 train_loss:4.2972 train_time:34899ms step_avg:163.08ms
step:225/1530 train_loss:3.9167 train_time:35062ms step_avg:163.08ms
step:226/1530 train_loss:3.9960 train_time:35224ms step_avg:163.08ms
step:227/1530 train_loss:3.9731 train_time:35386ms step_avg:163.07ms
step:228/1530 train_loss:4.1408 train_time:35551ms step_avg:163.08ms
step:229/1530 train_loss:3.9304 train_time:35719ms step_avg:163.10ms
step:230/1530 train_loss:4.0460 train_time:35884ms step_avg:163.11ms
step:231/1530 train_loss:3.9262 train_time:36050ms step_avg:163.12ms
step:232/1530 train_loss:3.9839 train_time:36215ms step_avg:163.13ms
step:233/1530 train_loss:4.0882 train_time:36381ms step_avg:163.14ms
step:234/1530 train_loss:4.0348 train_time:36547ms step_avg:163.16ms
step:235/1530 train_loss:3.9122 train_time:36714ms step_avg:163.17ms
step:236/1530 train_loss:4.0864 train_time:36880ms step_avg:163.18ms
step:237/1530 train_loss:4.0905 train_time:37045ms step_avg:163.19ms
step:238/1530 train_loss:3.9481 train_time:37212ms step_avg:163.21ms
step:239/1530 train_loss:4.0864 train_time:37379ms step_avg:163.23ms
step:240/1530 train_loss:4.1153 train_time:37544ms step_avg:163.24ms
step:241/1530 train_loss:3.9739 train_time:37709ms step_avg:163.24ms
step:242/1530 train_loss:4.1614 train_time:37876ms step_avg:163.26ms
step:243/1530 train_loss:4.0250 train_time:38042ms step_avg:163.27ms
step:244/1530 train_loss:4.0817 train_time:38207ms step_avg:163.28ms
step:245/1530 train_loss:4.1416 train_time:38372ms step_avg:163.29ms
step:246/1530 train_loss:4.0591 train_time:38539ms step_avg:163.30ms
step:247/1530 train_loss:4.0024 train_time:38704ms step_avg:163.31ms
step:248/1530 train_loss:4.1099 train_time:38869ms step_avg:163.31ms
step:249/1530 train_loss:3.9263 train_time:39034ms step_avg:163.32ms
step:250/1530 train_loss:3.9812 train_time:39200ms step_avg:163.33ms
step:250/1530 val_loss:4.0113 train_time:39248ms step_avg:163.53ms
step:251/1530 train_loss:4.0851 train_time:39368ms step_avg:163.35ms
step:252/1530 train_loss:4.1709 train_time:39536ms step_avg:163.37ms
step:253/1530 train_loss:3.9384 train_time:39703ms step_avg:163.39ms
step:254/1530 train_loss:3.8865 train_time:39868ms step_avg:163.39ms
step:255/1530 train_loss:4.0848 train_time:40034ms step_avg:163.40ms
step:256/1530 train_loss:3.9904 train_time:40201ms step_avg:163.42ms
step:257/1530 train_loss:3.9932 train_time:40366ms step_avg:163.43ms
step:258/1530 train_loss:3.9951 train_time:40532ms step_avg:163.43ms
step:259/1530 train_loss:4.0437 train_time:40699ms step_avg:163.45ms
step:260/1530 train_loss:4.0702 train_time:40865ms step_avg:163.46ms
step:261/1530 train_loss:4.0324 train_time:41031ms step_avg:163.47ms
step:262/1530 train_loss:3.9943 train_time:41198ms step_avg:163.48ms
step:263/1530 train_loss:3.8971 train_time:41363ms step_avg:163.49ms
step:264/1530 train_loss:3.9915 train_time:41528ms step_avg:163.50ms
step:265/1530 train_loss:3.8829 train_time:41694ms step_avg:163.50ms
step:266/1530 train_loss:3.9286 train_time:41859ms step_avg:163.51ms
step:267/1530 train_loss:3.9306 train_time:42025ms step_avg:163.52ms
step:268/1530 train_loss:3.9683 train_time:42191ms step_avg:163.53ms
step:269/1530 train_loss:3.8590 train_time:42357ms step_avg:163.54ms
step:270/1530 train_loss:4.1072 train_time:42523ms step_avg:163.55ms
step:271/1530 train_loss:3.9709 train_time:42689ms step_avg:163.56ms
step:272/1530 train_loss:3.9330 train_time:42856ms step_avg:163.57ms
step:273/1530 train_loss:3.9470 train_time:43020ms step_avg:163.58ms
step:274/1530 train_loss:4.0473 train_time:43187ms step_avg:163.59ms
step:275/1530 train_loss:4.0675 train_time:43353ms step_avg:163.60ms
step:276/1530 train_loss:4.2301 train_time:43520ms step_avg:163.61ms
step:277/1530 train_loss:4.0400 train_time:43684ms step_avg:163.61ms
step:278/1530 train_loss:4.0990 train_time:43849ms step_avg:163.62ms
step:279/1530 train_loss:4.0060 train_time:44016ms step_avg:163.63ms
step:280/1530 train_loss:4.1898 train_time:44183ms step_avg:163.64ms
step:281/1530 train_loss:3.9776 train_time:44348ms step_avg:163.65ms
step:282/1530 train_loss:3.9506 train_time:44517ms step_avg:163.67ms
step:283/1530 train_loss:3.9170 train_time:44683ms step_avg:163.67ms
step:284/1530 train_loss:4.0500 train_time:44848ms step_avg:163.68ms
step:285/1530 train_loss:4.0650 train_time:45013ms step_avg:163.68ms
step:286/1530 train_loss:4.0918 train_time:45179ms step_avg:163.69ms
step:287/1530 train_loss:3.9127 train_time:45344ms step_avg:163.70ms
step:288/1530 train_loss:4.0088 train_time:45509ms step_avg:163.70ms
step:289/1530 train_loss:3.8773 train_time:45675ms step_avg:163.71ms
step:290/1530 train_loss:3.8711 train_time:45840ms step_avg:163.71ms
step:291/1530 train_loss:3.9104 train_time:46005ms step_avg:163.72ms
step:292/1530 train_loss:3.8697 train_time:46168ms step_avg:163.72ms
step:293/1530 train_loss:3.9146 train_time:46334ms step_avg:163.72ms
step:294/1530 train_loss:3.9387 train_time:46499ms step_avg:163.73ms
step:295/1530 train_loss:3.8394 train_time:46663ms step_avg:163.73ms
step:296/1530 train_loss:3.8680 train_time:46828ms step_avg:163.73ms
step:297/1530 train_loss:3.8685 train_time:46994ms step_avg:163.74ms
step:298/1530 train_loss:3.9705 train_time:47158ms step_avg:163.74ms
step:299/1530 train_loss:3.8261 train_time:47323ms step_avg:163.75ms
step:300/1530 train_loss:3.9652 train_time:47489ms step_avg:163.76ms
step:301/1530 train_loss:3.9646 train_time:47654ms step_avg:163.76ms
step:302/1530 train_loss:3.9399 train_time:47819ms step_avg:163.76ms
step:303/1530 train_loss:3.9778 train_time:47984ms step_avg:163.77ms
step:304/1530 train_loss:3.9708 train_time:48149ms step_avg:163.77ms
step:305/1530 train_loss:4.4602 train_time:48316ms step_avg:163.78ms
step:306/1530 train_loss:3.9480 train_time:48481ms step_avg:163.79ms
step:307/1530 train_loss:3.8462 train_time:48645ms step_avg:163.79ms
step:308/1530 train_loss:3.9833 train_time:48811ms step_avg:163.79ms
step:309/1530 train_loss:3.8766 train_time:48977ms step_avg:163.80ms
step:310/1530 train_loss:4.0922 train_time:49141ms step_avg:163.80ms
step:311/1530 train_loss:3.9341 train_time:49306ms step_avg:163.81ms
step:312/1530 train_loss:3.8692 train_time:49470ms step_avg:163.81ms
step:313/1530 train_loss:3.9488 train_time:49637ms step_avg:163.82ms
step:314/1530 train_loss:4.0735 train_time:49802ms step_avg:163.82ms
step:315/1530 train_loss:3.9474 train_time:49966ms step_avg:163.82ms
step:316/1530 train_loss:3.8016 train_time:50131ms step_avg:163.83ms
step:317/1530 train_loss:3.8841 train_time:50298ms step_avg:163.84ms
step:318/1530 train_loss:3.9286 train_time:50462ms step_avg:163.84ms
step:319/1530 train_loss:3.8970 train_time:50627ms step_avg:163.84ms
step:320/1530 train_loss:4.0174 train_time:50792ms step_avg:163.84ms
step:321/1530 train_loss:3.9572 train_time:50957ms step_avg:163.85ms
step:322/1530 train_loss:3.9377 train_time:51122ms step_avg:163.85ms
step:323/1530 train_loss:4.0153 train_time:51286ms step_avg:163.85ms
step:324/1530 train_loss:3.9595 train_time:51451ms step_avg:163.86ms
step:325/1530 train_loss:4.0218 train_time:51617ms step_avg:163.86ms
step:326/1530 train_loss:3.8971 train_time:51782ms step_avg:163.87ms
step:327/1530 train_loss:4.4087 train_time:51947ms step_avg:163.87ms
step:328/1530 train_loss:4.0763 train_time:52113ms step_avg:163.88ms
step:329/1530 train_loss:3.8012 train_time:52279ms step_avg:163.88ms
step:330/1530 train_loss:3.7564 train_time:52444ms step_avg:163.89ms
step:331/1530 train_loss:3.9843 train_time:52608ms step_avg:163.89ms
step:332/1530 train_loss:3.9217 train_time:52774ms step_avg:163.90ms
step:333/1530 train_loss:3.8954 train_time:52941ms step_avg:163.90ms
step:334/1530 train_loss:3.8440 train_time:53105ms step_avg:163.90ms
step:335/1530 train_loss:4.0157 train_time:53269ms step_avg:163.91ms
step:336/1530 train_loss:3.9681 train_time:53435ms step_avg:163.91ms
step:337/1530 train_loss:4.4354 train_time:53601ms step_avg:163.92ms
step:338/1530 train_loss:3.9442 train_time:53766ms step_avg:163.92ms
step:339/1530 train_loss:3.8727 train_time:53930ms step_avg:163.92ms
step:340/1530 train_loss:3.9423 train_time:54096ms step_avg:163.93ms
step:341/1530 train_loss:3.8604 train_time:54262ms step_avg:163.93ms
step:342/1530 train_loss:3.8206 train_time:54428ms step_avg:163.94ms
step:343/1530 train_loss:3.8439 train_time:54599ms step_avg:163.96ms
step:344/1530 train_loss:4.0049 train_time:54767ms step_avg:163.97ms
step:345/1530 train_loss:3.8247 train_time:54936ms step_avg:163.99ms
step:346/1530 train_loss:3.7734 train_time:55104ms step_avg:164.00ms
step:347/1530 train_loss:3.8086 train_time:55272ms step_avg:164.01ms
step:348/1530 train_loss:3.8623 train_time:55440ms step_avg:164.02ms
step:349/1530 train_loss:3.8365 train_time:55608ms step_avg:164.04ms
step:350/1530 train_loss:3.5736 train_time:55777ms step_avg:164.05ms
step:351/1530 train_loss:3.8239 train_time:55945ms step_avg:164.06ms
step:352/1530 train_loss:4.1948 train_time:56111ms step_avg:164.07ms
step:353/1530 train_loss:3.6642 train_time:56280ms step_avg:164.08ms
step:354/1530 train_loss:3.9276 train_time:56448ms step_avg:164.09ms
step:355/1530 train_loss:3.7925 train_time:56617ms step_avg:164.11ms
step:356/1530 train_loss:3.8862 train_time:56784ms step_avg:164.12ms
step:357/1530 train_loss:3.7750 train_time:56951ms step_avg:164.12ms
step:358/1530 train_loss:3.8726 train_time:57120ms step_avg:164.14ms
step:359/1530 train_loss:3.7830 train_time:57289ms step_avg:164.15ms
step:360/1530 train_loss:3.4465 train_time:57459ms step_avg:164.17ms
step:361/1530 train_loss:4.0204 train_time:57627ms step_avg:164.18ms
step:362/1530 train_loss:3.9263 train_time:57795ms step_avg:164.19ms
step:363/1530 train_loss:3.8426 train_time:57962ms step_avg:164.20ms
step:364/1530 train_loss:3.7481 train_time:58130ms step_avg:164.21ms
step:365/1530 train_loss:3.9228 train_time:58299ms step_avg:164.22ms
step:366/1530 train_loss:3.8703 train_time:58466ms step_avg:164.23ms
step:367/1530 train_loss:3.8645 train_time:58634ms step_avg:164.24ms
step:368/1530 train_loss:3.8594 train_time:58802ms step_avg:164.25ms
step:369/1530 train_loss:3.7524 train_time:58969ms step_avg:164.26ms
step:370/1530 train_loss:3.8827 train_time:59137ms step_avg:164.27ms
step:371/1530 train_loss:3.7369 train_time:59305ms step_avg:164.28ms
step:372/1530 train_loss:3.6993 train_time:59472ms step_avg:164.29ms
step:373/1530 train_loss:3.9138 train_time:59639ms step_avg:164.29ms
step:374/1530 train_loss:3.8379 train_time:59806ms step_avg:164.30ms
step:375/1530 train_loss:3.8007 train_time:59974ms step_avg:164.31ms
step:375/1530 val_loss:3.8318 train_time:60023ms step_avg:164.45ms
step:376/1530 train_loss:3.8709 train_time:60145ms step_avg:164.33ms
step:377/1530 train_loss:3.7976 train_time:60446ms step_avg:164.70ms
step:378/1530 train_loss:3.8590 train_time:60622ms step_avg:164.73ms
step:379/1530 train_loss:3.8751 train_time:60947ms step_avg:165.17ms
step:380/1530 train_loss:3.9566 train_time:61112ms step_avg:165.17ms
step:381/1530 train_loss:3.8423 train_time:61281ms step_avg:165.18ms
step:382/1530 train_loss:3.8143 train_time:61450ms step_avg:165.19ms
step:383/1530 train_loss:3.8055 train_time:61617ms step_avg:165.19ms
step:384/1530 train_loss:3.8801 train_time:61784ms step_avg:165.20ms
step:385/1530 train_loss:3.7964 train_time:61953ms step_avg:165.21ms
step:386/1530 train_loss:3.8934 train_time:62120ms step_avg:165.21ms
step:387/1530 train_loss:4.0637 train_time:62287ms step_avg:165.22ms
step:388/1530 train_loss:3.7981 train_time:62456ms step_avg:165.23ms
step:389/1530 train_loss:3.8019 train_time:62624ms step_avg:165.23ms
step:390/1530 train_loss:3.9023 train_time:62791ms step_avg:165.24ms
step:391/1530 train_loss:3.8235 train_time:62959ms step_avg:165.25ms
step:392/1530 train_loss:3.9318 train_time:63124ms step_avg:165.25ms
step:393/1530 train_loss:3.7749 train_time:63293ms step_avg:165.26ms
step:394/1530 train_loss:3.8879 train_time:63460ms step_avg:165.26ms
step:395/1530 train_loss:3.6378 train_time:63627ms step_avg:165.26ms
step:396/1530 train_loss:3.8452 train_time:63795ms step_avg:165.27ms
step:397/1530 train_loss:3.8686 train_time:63963ms step_avg:165.28ms
step:398/1530 train_loss:3.8893 train_time:64130ms step_avg:165.28ms
step:399/1530 train_loss:3.7773 train_time:64297ms step_avg:165.29ms
step:400/1530 train_loss:3.8356 train_time:64466ms step_avg:165.30ms
step:401/1530 train_loss:3.9251 train_time:64634ms step_avg:165.30ms
step:402/1530 train_loss:3.8540 train_time:64800ms step_avg:165.31ms
step:403/1530 train_loss:3.9635 train_time:64969ms step_avg:165.32ms
step:404/1530 train_loss:3.6849 train_time:65136ms step_avg:165.32ms
step:405/1530 train_loss:3.7892 train_time:65302ms step_avg:165.32ms
step:406/1530 train_loss:4.0996 train_time:65470ms step_avg:165.33ms
step:407/1530 train_loss:3.7835 train_time:65637ms step_avg:165.33ms
step:408/1530 train_loss:3.8215 train_time:65804ms step_avg:165.34ms
step:409/1530 train_loss:3.8645 train_time:65972ms step_avg:165.34ms
step:410/1530 train_loss:3.7647 train_time:66139ms step_avg:165.35ms
step:411/1530 train_loss:3.7649 train_time:66306ms step_avg:165.35ms
step:412/1530 train_loss:4.1925 train_time:66474ms step_avg:165.36ms
step:413/1530 train_loss:3.6552 train_time:66641ms step_avg:165.36ms
step:414/1530 train_loss:4.0168 train_time:66807ms step_avg:165.36ms
step:415/1530 train_loss:3.7621 train_time:66975ms step_avg:165.37ms
step:416/1530 train_loss:3.7726 train_time:67141ms step_avg:165.37ms
step:417/1530 train_loss:3.9646 train_time:67309ms step_avg:165.38ms
step:418/1530 train_loss:3.6966 train_time:67476ms step_avg:165.38ms
step:419/1530 train_loss:3.8125 train_time:67643ms step_avg:165.39ms
step:420/1530 train_loss:3.7111 train_time:67810ms step_avg:165.39ms
step:421/1530 train_loss:3.6564 train_time:67976ms step_avg:165.39ms
step:422/1530 train_loss:3.7883 train_time:68142ms step_avg:165.39ms
step:423/1530 train_loss:3.8782 train_time:68309ms step_avg:165.40ms
step:424/1530 train_loss:3.6221 train_time:68476ms step_avg:165.40ms
step:425/1530 train_loss:3.8017 train_time:68643ms step_avg:165.41ms
step:426/1530 train_loss:3.6577 train_time:68811ms step_avg:165.41ms
step:427/1530 train_loss:3.8947 train_time:68978ms step_avg:165.41ms
step:428/1530 train_loss:3.8168 train_time:69144ms step_avg:165.42ms
step:429/1530 train_loss:3.7619 train_time:69312ms step_avg:165.42ms
step:430/1530 train_loss:3.7070 train_time:69479ms step_avg:165.43ms
step:431/1530 train_loss:3.6376 train_time:69646ms step_avg:165.43ms
step:432/1530 train_loss:3.7676 train_time:69814ms step_avg:165.44ms
step:433/1530 train_loss:3.8264 train_time:69981ms step_avg:165.44ms
step:434/1530 train_loss:3.7785 train_time:70148ms step_avg:165.44ms
step:435/1530 train_loss:3.8150 train_time:70316ms step_avg:165.45ms
step:436/1530 train_loss:3.8335 train_time:70482ms step_avg:165.45ms
step:437/1530 train_loss:3.7205 train_time:70650ms step_avg:165.46ms
step:438/1530 train_loss:3.7074 train_time:70816ms step_avg:165.46ms
step:439/1530 train_loss:3.7159 train_time:70983ms step_avg:165.46ms
step:440/1530 train_loss:3.8933 train_time:71151ms step_avg:165.47ms
step:441/1530 train_loss:3.7642 train_time:71318ms step_avg:165.47ms
step:442/1530 train_loss:3.7441 train_time:71484ms step_avg:165.47ms
step:443/1530 train_loss:3.6315 train_time:71652ms step_avg:165.48ms
step:444/1530 train_loss:3.9373 train_time:71818ms step_avg:165.48ms
step:445/1530 train_loss:3.8489 train_time:71984ms step_avg:165.48ms
step:446/1530 train_loss:3.8426 train_time:72152ms step_avg:165.49ms
step:447/1530 train_loss:3.7570 train_time:72319ms step_avg:165.49ms
step:448/1530 train_loss:3.8548 train_time:72485ms step_avg:165.49ms
step:449/1530 train_loss:3.6967 train_time:72655ms step_avg:165.50ms
step:450/1530 train_loss:3.7271 train_time:72821ms step_avg:165.50ms
step:451/1530 train_loss:3.5866 train_time:72987ms step_avg:165.50ms
step:452/1530 train_loss:3.7272 train_time:73154ms step_avg:165.51ms
step:453/1530 train_loss:3.6846 train_time:73321ms step_avg:165.51ms
step:454/1530 train_loss:3.6461 train_time:73487ms step_avg:165.51ms
step:455/1530 train_loss:3.8484 train_time:73658ms step_avg:165.52ms
step:456/1530 train_loss:3.7330 train_time:73826ms step_avg:165.53ms
step:457/1530 train_loss:3.7907 train_time:73996ms step_avg:165.54ms
step:458/1530 train_loss:3.8326 train_time:74165ms step_avg:165.55ms
step:459/1530 train_loss:3.6411 train_time:74337ms step_avg:165.56ms
step:460/1530 train_loss:3.7984 train_time:74506ms step_avg:165.57ms
step:461/1530 train_loss:3.6974 train_time:74677ms step_avg:165.58ms
step:462/1530 train_loss:3.7433 train_time:74846ms step_avg:165.59ms
step:463/1530 train_loss:3.7839 train_time:75017ms step_avg:165.60ms
step:464/1530 train_loss:3.7225 train_time:75185ms step_avg:165.61ms
step:465/1530 train_loss:3.7242 train_time:75355ms step_avg:165.62ms
step:466/1530 train_loss:3.8056 train_time:75524ms step_avg:165.62ms
step:467/1530 train_loss:3.8312 train_time:75695ms step_avg:165.64ms
step:468/1530 train_loss:3.7985 train_time:75863ms step_avg:165.64ms
step:469/1530 train_loss:3.6935 train_time:76034ms step_avg:165.65ms
step:470/1530 train_loss:3.7685 train_time:76203ms step_avg:165.66ms
step:471/1530 train_loss:3.8187 train_time:76374ms step_avg:165.67ms
step:472/1530 train_loss:3.7881 train_time:76543ms step_avg:165.68ms
step:473/1530 train_loss:3.7187 train_time:76713ms step_avg:165.69ms
step:474/1530 train_loss:3.5992 train_time:76882ms step_avg:165.69ms
step:475/1530 train_loss:4.0258 train_time:77050ms step_avg:165.70ms
step:476/1530 train_loss:3.7612 train_time:77219ms step_avg:165.71ms
step:477/1530 train_loss:3.6051 train_time:77389ms step_avg:165.72ms
step:478/1530 train_loss:3.8256 train_time:77560ms step_avg:165.73ms
step:479/1530 train_loss:3.7779 train_time:77730ms step_avg:165.74ms
step:480/1530 train_loss:3.9259 train_time:77899ms step_avg:165.74ms
step:481/1530 train_loss:3.7249 train_time:78069ms step_avg:165.75ms
step:482/1530 train_loss:3.5369 train_time:78238ms step_avg:165.76ms
step:483/1530 train_loss:3.8116 train_time:78406ms step_avg:165.76ms
step:484/1530 train_loss:3.6688 train_time:78577ms step_avg:165.78ms
step:485/1530 train_loss:3.6621 train_time:78747ms step_avg:165.78ms
step:486/1530 train_loss:3.5735 train_time:78917ms step_avg:165.79ms
step:487/1530 train_loss:3.6906 train_time:79085ms step_avg:165.80ms
step:488/1530 train_loss:3.8878 train_time:79256ms step_avg:165.81ms
step:489/1530 train_loss:3.7185 train_time:79426ms step_avg:165.82ms
step:490/1530 train_loss:3.5997 train_time:79595ms step_avg:165.82ms
step:491/1530 train_loss:3.6204 train_time:79764ms step_avg:165.83ms
step:492/1530 train_loss:3.7358 train_time:79934ms step_avg:165.84ms
step:493/1530 train_loss:3.5769 train_time:80104ms step_avg:165.85ms
step:494/1530 train_loss:3.7072 train_time:80272ms step_avg:165.85ms
step:495/1530 train_loss:3.6684 train_time:80443ms step_avg:165.86ms
step:496/1530 train_loss:3.5264 train_time:80614ms step_avg:165.87ms
step:497/1530 train_loss:3.7389 train_time:80782ms step_avg:165.88ms
step:498/1530 train_loss:3.7972 train_time:80951ms step_avg:165.88ms
step:499/1530 train_loss:3.8326 train_time:81120ms step_avg:165.89ms
step:500/1530 train_loss:3.7407 train_time:81290ms step_avg:165.90ms
step:500/1530 val_loss:3.7125 train_time:81338ms step_avg:166.00ms
step:501/1530 train_loss:3.8073 train_time:81459ms step_avg:165.90ms
step:502/1530 train_loss:3.7581 train_time:81631ms step_avg:165.92ms
step:503/1530 train_loss:3.7841 train_time:81801ms step_avg:165.93ms
step:504/1530 train_loss:3.7258 train_time:81971ms step_avg:165.93ms
step:505/1530 train_loss:3.8179 train_time:82140ms step_avg:165.94ms
step:506/1530 train_loss:3.6587 train_time:82310ms step_avg:165.95ms
step:507/1530 train_loss:3.7659 train_time:82478ms step_avg:165.95ms
step:508/1530 train_loss:3.8337 train_time:82647ms step_avg:165.96ms
step:509/1530 train_loss:3.7821 train_time:82816ms step_avg:165.96ms
step:510/1530 train_loss:3.5841 train_time:82985ms step_avg:165.97ms
step:511/1530 train_loss:3.7789 train_time:83155ms step_avg:165.98ms
step:512/1530 train_loss:3.7284 train_time:83325ms step_avg:165.99ms
step:513/1530 train_loss:3.6696 train_time:83493ms step_avg:165.99ms
step:514/1530 train_loss:3.8998 train_time:83664ms step_avg:166.00ms
step:515/1530 train_loss:3.7455 train_time:83834ms step_avg:166.01ms
step:516/1530 train_loss:4.0856 train_time:84004ms step_avg:166.02ms
step:517/1530 train_loss:3.6988 train_time:84174ms step_avg:166.02ms
step:518/1530 train_loss:3.7795 train_time:84341ms step_avg:166.03ms
step:519/1530 train_loss:3.6607 train_time:84512ms step_avg:166.04ms
step:520/1530 train_loss:3.6923 train_time:84680ms step_avg:166.04ms
step:521/1530 train_loss:3.6674 train_time:84848ms step_avg:166.04ms
step:522/1530 train_loss:3.6684 train_time:85018ms step_avg:166.05ms
step:523/1530 train_loss:4.2918 train_time:85188ms step_avg:166.06ms
step:524/1530 train_loss:3.7442 train_time:85356ms step_avg:166.06ms
step:525/1530 train_loss:3.6864 train_time:85523ms step_avg:166.06ms
step:526/1530 train_loss:3.7001 train_time:85693ms step_avg:166.07ms
step:527/1530 train_loss:3.6611 train_time:85862ms step_avg:166.08ms
step:528/1530 train_loss:3.6319 train_time:86029ms step_avg:166.08ms
step:529/1530 train_loss:3.8531 train_time:86199ms step_avg:166.09ms
step:530/1530 train_loss:3.6525 train_time:86367ms step_avg:166.09ms
step:531/1530 train_loss:3.9291 train_time:86538ms step_avg:166.10ms
step:532/1530 train_loss:3.7399 train_time:86705ms step_avg:166.10ms
step:533/1530 train_loss:3.6604 train_time:86875ms step_avg:166.11ms
step:534/1530 train_loss:3.6775 train_time:87043ms step_avg:166.11ms
step:535/1530 train_loss:3.6140 train_time:87213ms step_avg:166.12ms
step:536/1530 train_loss:3.7559 train_time:87383ms step_avg:166.13ms
step:537/1530 train_loss:3.7296 train_time:87553ms step_avg:166.13ms
step:538/1530 train_loss:3.6293 train_time:87721ms step_avg:166.14ms
step:539/1530 train_loss:4.1194 train_time:87893ms step_avg:166.15ms
step:540/1530 train_loss:3.6816 train_time:88061ms step_avg:166.15ms
step:541/1530 train_loss:3.7946 train_time:88229ms step_avg:166.16ms
step:542/1530 train_loss:3.5935 train_time:88398ms step_avg:166.16ms
step:543/1530 train_loss:3.5930 train_time:88567ms step_avg:166.17ms
step:544/1530 train_loss:3.6485 train_time:88736ms step_avg:166.17ms
step:545/1530 train_loss:3.5968 train_time:88905ms step_avg:166.18ms
step:546/1530 train_loss:3.6320 train_time:89074ms step_avg:166.18ms
step:547/1530 train_loss:3.6445 train_time:89241ms step_avg:166.18ms
step:548/1530 train_loss:3.6136 train_time:89411ms step_avg:166.19ms
step:549/1530 train_loss:3.7263 train_time:89579ms step_avg:166.19ms
step:550/1530 train_loss:3.6253 train_time:89749ms step_avg:166.20ms
step:551/1530 train_loss:3.6395 train_time:89916ms step_avg:166.20ms
step:552/1530 train_loss:3.9329 train_time:90085ms step_avg:166.21ms
step:553/1530 train_loss:3.7587 train_time:90254ms step_avg:166.21ms
step:554/1530 train_loss:3.7195 train_time:90421ms step_avg:166.22ms
step:555/1530 train_loss:3.6308 train_time:90592ms step_avg:166.22ms
step:556/1530 train_loss:3.7018 train_time:90760ms step_avg:166.23ms
step:557/1530 train_loss:3.3237 train_time:90929ms step_avg:166.23ms
step:558/1530 train_loss:3.6192 train_time:91099ms step_avg:166.24ms
step:559/1530 train_loss:3.6547 train_time:91266ms step_avg:166.24ms
step:560/1530 train_loss:3.6968 train_time:91435ms step_avg:166.25ms
step:561/1530 train_loss:3.6171 train_time:91603ms step_avg:166.25ms
step:562/1530 train_loss:3.5604 train_time:91772ms step_avg:166.25ms
step:563/1530 train_loss:3.7654 train_time:91941ms step_avg:166.26ms
step:564/1530 train_loss:3.5799 train_time:92112ms step_avg:166.27ms
step:565/1530 train_loss:3.6846 train_time:92280ms step_avg:166.27ms
step:566/1530 train_loss:3.6244 train_time:92583ms step_avg:166.52ms
step:567/1530 train_loss:3.6063 train_time:92761ms step_avg:166.54ms
step:568/1530 train_loss:3.6926 train_time:92931ms step_avg:166.54ms
step:569/1530 train_loss:3.6562 train_time:93253ms step_avg:166.82ms
step:570/1530 train_loss:3.6939 train_time:93421ms step_avg:166.82ms
step:571/1530 train_loss:3.7619 train_time:93592ms step_avg:166.83ms
step:572/1530 train_loss:3.7323 train_time:93763ms step_avg:166.84ms
step:573/1530 train_loss:3.7441 train_time:93936ms step_avg:166.85ms
step:574/1530 train_loss:3.7858 train_time:94110ms step_avg:166.86ms
step:575/1530 train_loss:3.7403 train_time:94280ms step_avg:166.87ms
step:576/1530 train_loss:3.7598 train_time:94452ms step_avg:166.88ms
step:577/1530 train_loss:3.6837 train_time:94622ms step_avg:166.88ms
step:578/1530 train_loss:3.6766 train_time:94795ms step_avg:166.89ms
step:579/1530 train_loss:3.6778 train_time:94966ms step_avg:166.90ms
step:580/1530 train_loss:3.5900 train_time:95136ms step_avg:166.91ms
step:581/1530 train_loss:3.6476 train_time:95309ms step_avg:166.92ms
step:582/1530 train_loss:3.8586 train_time:95479ms step_avg:166.92ms
step:583/1530 train_loss:3.6354 train_time:95650ms step_avg:166.93ms
step:584/1530 train_loss:3.6033 train_time:95821ms step_avg:166.94ms
step:585/1530 train_loss:3.7906 train_time:95992ms step_avg:166.94ms
step:586/1530 train_loss:3.5273 train_time:96162ms step_avg:166.95ms
step:587/1530 train_loss:3.6706 train_time:96332ms step_avg:166.95ms
step:588/1530 train_loss:3.6463 train_time:96503ms step_avg:166.96ms
step:589/1530 train_loss:4.0039 train_time:96675ms step_avg:166.97ms
step:590/1530 train_loss:3.7835 train_time:96848ms step_avg:166.98ms
step:591/1530 train_loss:3.5111 train_time:97019ms step_avg:166.99ms
step:592/1530 train_loss:3.5415 train_time:97192ms step_avg:167.00ms
step:593/1530 train_loss:3.5124 train_time:97364ms step_avg:167.00ms
step:594/1530 train_loss:3.5587 train_time:97535ms step_avg:167.01ms
step:595/1530 train_loss:3.9167 train_time:97706ms step_avg:167.02ms
step:596/1530 train_loss:3.6552 train_time:97879ms step_avg:167.03ms
step:597/1530 train_loss:3.5944 train_time:98051ms step_avg:167.04ms
step:598/1530 train_loss:3.6658 train_time:98221ms step_avg:167.04ms
step:599/1530 train_loss:3.4840 train_time:98392ms step_avg:167.05ms
step:600/1530 train_loss:3.6020 train_time:98562ms step_avg:167.05ms
step:601/1530 train_loss:3.6542 train_time:98736ms step_avg:167.07ms
step:602/1530 train_loss:3.6723 train_time:98908ms step_avg:167.08ms
step:603/1530 train_loss:3.7919 train_time:99079ms step_avg:167.08ms
step:604/1530 train_loss:3.6148 train_time:99250ms step_avg:167.09ms
step:605/1530 train_loss:3.6175 train_time:99423ms step_avg:167.10ms
step:606/1530 train_loss:3.5778 train_time:99596ms step_avg:167.11ms
step:607/1530 train_loss:3.8404 train_time:99768ms step_avg:167.12ms
step:608/1530 train_loss:3.6422 train_time:99939ms step_avg:167.12ms
step:609/1530 train_loss:3.6249 train_time:100111ms step_avg:167.13ms
step:610/1530 train_loss:3.7101 train_time:100280ms step_avg:167.13ms
step:611/1530 train_loss:3.6041 train_time:100452ms step_avg:167.14ms
step:612/1530 train_loss:3.5752 train_time:100622ms step_avg:167.15ms
step:613/1530 train_loss:3.7711 train_time:100794ms step_avg:167.15ms
step:614/1530 train_loss:3.7023 train_time:100965ms step_avg:167.16ms
step:615/1530 train_loss:3.6955 train_time:101136ms step_avg:167.17ms
step:616/1530 train_loss:3.6355 train_time:101307ms step_avg:167.17ms
step:617/1530 train_loss:3.5615 train_time:101480ms step_avg:167.18ms
step:618/1530 train_loss:3.6946 train_time:101650ms step_avg:167.19ms
step:619/1530 train_loss:3.5550 train_time:101821ms step_avg:167.19ms
step:620/1530 train_loss:3.5944 train_time:101992ms step_avg:167.20ms
step:621/1530 train_loss:3.9312 train_time:102163ms step_avg:167.21ms
step:622/1530 train_loss:3.5837 train_time:102335ms step_avg:167.21ms
step:623/1530 train_loss:3.6075 train_time:102508ms step_avg:167.22ms
step:624/1530 train_loss:3.7066 train_time:102678ms step_avg:167.23ms
step:625/1530 train_loss:3.7105 train_time:102848ms step_avg:167.23ms
step:625/1530 val_loss:3.6293 train_time:102897ms step_avg:167.31ms
step:626/1530 train_loss:3.7434 train_time:103019ms step_avg:167.24ms
step:627/1530 train_loss:3.7140 train_time:103191ms step_avg:167.25ms
step:628/1530 train_loss:3.7703 train_time:103360ms step_avg:167.25ms
step:629/1530 train_loss:3.5990 train_time:103532ms step_avg:167.26ms
step:630/1530 train_loss:3.7311 train_time:103702ms step_avg:167.26ms
step:631/1530 train_loss:3.7441 train_time:103872ms step_avg:167.27ms
step:632/1530 train_loss:3.6510 train_time:104044ms step_avg:167.27ms
step:633/1530 train_loss:3.6143 train_time:104215ms step_avg:167.28ms
step:634/1530 train_loss:3.7031 train_time:104387ms step_avg:167.29ms
step:635/1530 train_loss:3.9604 train_time:104556ms step_avg:167.29ms
step:636/1530 train_loss:3.5556 train_time:104727ms step_avg:167.30ms
step:637/1530 train_loss:3.3630 train_time:104897ms step_avg:167.30ms
step:638/1530 train_loss:3.5987 train_time:105067ms step_avg:167.30ms
step:639/1530 train_loss:3.6397 train_time:105236ms step_avg:167.31ms
step:640/1530 train_loss:3.5769 train_time:105406ms step_avg:167.31ms
step:641/1530 train_loss:3.5907 train_time:105577ms step_avg:167.32ms
step:642/1530 train_loss:3.6347 train_time:105747ms step_avg:167.32ms
step:643/1530 train_loss:3.5961 train_time:105919ms step_avg:167.33ms
step:644/1530 train_loss:3.5748 train_time:106089ms step_avg:167.33ms
step:645/1530 train_loss:3.7806 train_time:106259ms step_avg:167.34ms
step:646/1530 train_loss:3.6871 train_time:106431ms step_avg:167.34ms
step:647/1530 train_loss:3.6691 train_time:106600ms step_avg:167.35ms
step:648/1530 train_loss:3.7201 train_time:106773ms step_avg:167.36ms
step:649/1530 train_loss:3.7746 train_time:106943ms step_avg:167.36ms
step:650/1530 train_loss:3.6228 train_time:107115ms step_avg:167.37ms
step:651/1530 train_loss:3.7754 train_time:107286ms step_avg:167.37ms
step:652/1530 train_loss:3.5937 train_time:107455ms step_avg:167.38ms
step:653/1530 train_loss:3.6641 train_time:107626ms step_avg:167.38ms
step:654/1530 train_loss:3.4359 train_time:107795ms step_avg:167.38ms
step:655/1530 train_loss:3.5878 train_time:107964ms step_avg:167.39ms
step:656/1530 train_loss:3.5860 train_time:108135ms step_avg:167.39ms
step:657/1530 train_loss:3.5056 train_time:108305ms step_avg:167.39ms
step:658/1530 train_loss:3.6932 train_time:108474ms step_avg:167.40ms
step:659/1530 train_loss:3.5908 train_time:108646ms step_avg:167.41ms
step:660/1530 train_loss:3.6943 train_time:108815ms step_avg:167.41ms
step:661/1530 train_loss:3.7598 train_time:108987ms step_avg:167.41ms
step:662/1530 train_loss:3.6758 train_time:109156ms step_avg:167.42ms
step:663/1530 train_loss:3.5628 train_time:109326ms step_avg:167.42ms
step:664/1530 train_loss:3.6195 train_time:109495ms step_avg:167.42ms
step:665/1530 train_loss:3.4994 train_time:109666ms step_avg:167.43ms
step:666/1530 train_loss:3.7874 train_time:109836ms step_avg:167.43ms
step:667/1530 train_loss:3.6152 train_time:110008ms step_avg:167.44ms
step:668/1530 train_loss:3.6502 train_time:110178ms step_avg:167.44ms
step:669/1530 train_loss:3.4932 train_time:110351ms step_avg:167.45ms
step:670/1530 train_loss:3.6054 train_time:110520ms step_avg:167.46ms
step:671/1530 train_loss:3.5662 train_time:110691ms step_avg:167.46ms
step:672/1530 train_loss:3.5748 train_time:110862ms step_avg:167.47ms
step:673/1530 train_loss:3.8543 train_time:111034ms step_avg:167.47ms
step:674/1530 train_loss:3.6317 train_time:111205ms step_avg:167.48ms
step:675/1530 train_loss:3.7221 train_time:111375ms step_avg:167.48ms
step:676/1530 train_loss:3.5015 train_time:111546ms step_avg:167.49ms
step:677/1530 train_loss:3.6073 train_time:111717ms step_avg:167.49ms
step:678/1530 train_loss:3.5632 train_time:111888ms step_avg:167.50ms
step:679/1530 train_loss:3.6863 train_time:112059ms step_avg:167.50ms
step:680/1530 train_loss:3.5912 train_time:112230ms step_avg:167.51ms
step:681/1530 train_loss:3.6225 train_time:112401ms step_avg:167.51ms
step:682/1530 train_loss:3.6686 train_time:112577ms step_avg:167.52ms
step:683/1530 train_loss:3.7484 train_time:112751ms step_avg:167.53ms
step:684/1530 train_loss:3.6519 train_time:112922ms step_avg:167.54ms
step:685/1530 train_loss:3.6890 train_time:113094ms step_avg:167.55ms
step:686/1530 train_loss:3.6497 train_time:113267ms step_avg:167.55ms
step:687/1530 train_loss:3.6706 train_time:113441ms step_avg:167.56ms
step:688/1530 train_loss:3.2156 train_time:113616ms step_avg:167.58ms
step:689/1530 train_loss:3.4090 train_time:113791ms step_avg:167.59ms
step:690/1530 train_loss:3.5487 train_time:113966ms step_avg:167.60ms
step:691/1530 train_loss:3.4147 train_time:114138ms step_avg:167.60ms
step:692/1530 train_loss:3.6354 train_time:114310ms step_avg:167.61ms
step:693/1530 train_loss:3.6572 train_time:114483ms step_avg:167.62ms
step:694/1530 train_loss:3.5611 train_time:114655ms step_avg:167.62ms
step:695/1530 train_loss:3.5376 train_time:114825ms step_avg:167.63ms
step:696/1530 train_loss:3.8576 train_time:114998ms step_avg:167.64ms
step:697/1530 train_loss:3.5926 train_time:115171ms step_avg:167.64ms
step:698/1530 train_loss:3.6477 train_time:115342ms step_avg:167.65ms
step:699/1530 train_loss:3.7866 train_time:115518ms step_avg:167.66ms
step:700/1530 train_loss:3.5724 train_time:115690ms step_avg:167.67ms
step:701/1530 train_loss:3.5448 train_time:115861ms step_avg:167.67ms
step:702/1530 train_loss:3.5229 train_time:116035ms step_avg:167.68ms
step:703/1530 train_loss:3.5035 train_time:116207ms step_avg:167.69ms
step:704/1530 train_loss:3.5781 train_time:116378ms step_avg:167.69ms
step:705/1530 train_loss:3.5674 train_time:116554ms step_avg:167.70ms
step:706/1530 train_loss:3.5857 train_time:116730ms step_avg:167.72ms
step:707/1530 train_loss:3.6521 train_time:116904ms step_avg:167.72ms
step:708/1530 train_loss:3.6073 train_time:117076ms step_avg:167.73ms
step:709/1530 train_loss:3.5905 train_time:117251ms step_avg:167.74ms
step:710/1530 train_loss:3.5465 train_time:117422ms step_avg:167.75ms
step:711/1530 train_loss:3.5963 train_time:117594ms step_avg:167.75ms
step:712/1530 train_loss:3.6527 train_time:117770ms step_avg:167.76ms
step:713/1530 train_loss:3.6620 train_time:117948ms step_avg:167.78ms
step:714/1530 train_loss:3.5677 train_time:118119ms step_avg:167.78ms
step:715/1530 train_loss:3.5734 train_time:118292ms step_avg:167.79ms
step:716/1530 train_loss:3.5997 train_time:118463ms step_avg:167.79ms
step:717/1530 train_loss:3.7121 train_time:118638ms step_avg:167.80ms
step:718/1530 train_loss:3.6038 train_time:118810ms step_avg:167.81ms
step:719/1530 train_loss:3.6825 train_time:118982ms step_avg:167.82ms
step:720/1530 train_loss:3.8558 train_time:119157ms step_avg:167.83ms
step:721/1530 train_loss:3.4730 train_time:119331ms step_avg:167.83ms
step:722/1530 train_loss:3.7423 train_time:119502ms step_avg:167.84ms
step:723/1530 train_loss:3.7724 train_time:119673ms step_avg:167.84ms
step:724/1530 train_loss:3.5729 train_time:119847ms step_avg:167.85ms
step:725/1530 train_loss:3.6582 train_time:120018ms step_avg:167.86ms
step:726/1530 train_loss:3.5415 train_time:120191ms step_avg:167.86ms
step:727/1530 train_loss:3.5876 train_time:120366ms step_avg:167.87ms
step:728/1530 train_loss:3.7463 train_time:120540ms step_avg:167.88ms
step:729/1530 train_loss:3.6794 train_time:120713ms step_avg:167.89ms
step:730/1530 train_loss:3.6739 train_time:120886ms step_avg:167.90ms
step:731/1530 train_loss:3.5689 train_time:121058ms step_avg:167.90ms
step:732/1530 train_loss:3.6007 train_time:121229ms step_avg:167.91ms
step:733/1530 train_loss:3.8384 train_time:121403ms step_avg:167.92ms
step:734/1530 train_loss:3.5714 train_time:121577ms step_avg:167.92ms
step:735/1530 train_loss:3.6244 train_time:121750ms step_avg:167.93ms
step:736/1530 train_loss:3.7400 train_time:121922ms step_avg:167.94ms
step:737/1530 train_loss:3.6864 train_time:122094ms step_avg:167.94ms
step:738/1530 train_loss:3.6097 train_time:122265ms step_avg:167.95ms
step:739/1530 train_loss:3.5118 train_time:122438ms step_avg:167.95ms
step:740/1530 train_loss:4.1232 train_time:122615ms step_avg:167.97ms
step:741/1530 train_loss:3.4924 train_time:122788ms step_avg:167.97ms
step:742/1530 train_loss:3.5594 train_time:122960ms step_avg:167.98ms
step:743/1530 train_loss:3.5838 train_time:123132ms step_avg:167.98ms
step:744/1530 train_loss:3.6545 train_time:123305ms step_avg:167.99ms
step:745/1530 train_loss:3.5937 train_time:123478ms step_avg:168.00ms
step:746/1530 train_loss:3.6002 train_time:123651ms step_avg:168.00ms
step:747/1530 train_loss:3.6501 train_time:123825ms step_avg:168.01ms
step:748/1530 train_loss:3.5724 train_time:124001ms step_avg:168.02ms
step:749/1530 train_loss:3.5679 train_time:124174ms step_avg:168.03ms
step:750/1530 train_loss:3.6046 train_time:124344ms step_avg:168.03ms
step:750/1530 val_loss:3.5730 train_time:124393ms step_avg:168.10ms
step:751/1530 train_loss:3.5749 train_time:124518ms step_avg:168.04ms
step:752/1530 train_loss:3.6228 train_time:124689ms step_avg:168.05ms
step:753/1530 train_loss:3.6281 train_time:124864ms step_avg:168.05ms
step:754/1530 train_loss:3.6039 train_time:125038ms step_avg:168.06ms
step:755/1530 train_loss:3.6944 train_time:125344ms step_avg:168.25ms
step:756/1530 train_loss:3.4666 train_time:125530ms step_avg:168.27ms
step:757/1530 train_loss:3.7361 train_time:125703ms step_avg:168.28ms
step:758/1530 train_loss:3.6556 train_time:125876ms step_avg:168.28ms
step:759/1530 train_loss:3.5971 train_time:126200ms step_avg:168.49ms
step:760/1530 train_loss:3.7132 train_time:126371ms step_avg:168.49ms
step:761/1530 train_loss:3.4066 train_time:126542ms step_avg:168.50ms
step:762/1530 train_loss:3.5542 train_time:126715ms step_avg:168.50ms
step:763/1530 train_loss:3.6697 train_time:126886ms step_avg:168.51ms
step:764/1530 train_loss:3.3252 train_time:127059ms step_avg:168.51ms
step:765/1530 train_loss:3.7377 train_time:127231ms step_avg:168.52ms
step:766/1530 train_loss:3.5804 train_time:127405ms step_avg:168.52ms
step:767/1530 train_loss:3.5667 train_time:127578ms step_avg:168.53ms
step:768/1530 train_loss:3.5756 train_time:127751ms step_avg:168.54ms
step:769/1530 train_loss:3.5919 train_time:127922ms step_avg:168.54ms
step:770/1530 train_loss:3.6471 train_time:128094ms step_avg:168.54ms
step:771/1530 train_loss:3.8928 train_time:128265ms step_avg:168.55ms
step:772/1530 train_loss:3.4611 train_time:128438ms step_avg:168.55ms
step:773/1530 train_loss:3.6395 train_time:128610ms step_avg:168.56ms
step:774/1530 train_loss:3.6499 train_time:128782ms step_avg:168.56ms
step:775/1530 train_loss:3.6138 train_time:128954ms step_avg:168.57ms
step:776/1530 train_loss:3.4110 train_time:129126ms step_avg:168.57ms
step:777/1530 train_loss:3.3910 train_time:129301ms step_avg:168.58ms
step:778/1530 train_loss:3.4955 train_time:129472ms step_avg:168.58ms
step:779/1530 train_loss:3.5886 train_time:129645ms step_avg:168.59ms
step:780/1530 train_loss:3.5949 train_time:129816ms step_avg:168.59ms
step:781/1530 train_loss:3.6757 train_time:129990ms step_avg:168.60ms
step:782/1530 train_loss:3.5923 train_time:130162ms step_avg:168.60ms
step:783/1530 train_loss:3.5741 train_time:130333ms step_avg:168.61ms
step:784/1530 train_loss:3.6057 train_time:130504ms step_avg:168.61ms
step:785/1530 train_loss:3.5704 train_time:130676ms step_avg:168.61ms
step:786/1530 train_loss:3.4481 train_time:130848ms step_avg:168.62ms
step:787/1530 train_loss:3.7848 train_time:131020ms step_avg:168.62ms
step:788/1530 train_loss:3.5109 train_time:131195ms step_avg:168.63ms
step:789/1530 train_loss:3.5518 train_time:131366ms step_avg:168.63ms
step:790/1530 train_loss:3.6323 train_time:131539ms step_avg:168.64ms
step:791/1530 train_loss:3.7801 train_time:131713ms step_avg:168.65ms
step:792/1530 train_loss:3.7650 train_time:131885ms step_avg:168.65ms
step:793/1530 train_loss:3.4462 train_time:132056ms step_avg:168.65ms
step:794/1530 train_loss:3.6015 train_time:132229ms step_avg:168.66ms
step:795/1530 train_loss:3.6837 train_time:132403ms step_avg:168.67ms
step:796/1530 train_loss:3.7911 train_time:132581ms step_avg:168.68ms
step:797/1530 train_loss:3.5297 train_time:132755ms step_avg:168.69ms
step:798/1530 train_loss:3.6619 train_time:132930ms step_avg:168.69ms
step:799/1530 train_loss:3.5425 train_time:133105ms step_avg:168.70ms
step:800/1530 train_loss:3.5379 train_time:133279ms step_avg:168.71ms
step:801/1530 train_loss:3.6323 train_time:133454ms step_avg:168.72ms
step:802/1530 train_loss:3.5038 train_time:133630ms step_avg:168.73ms
step:803/1530 train_loss:3.4848 train_time:133803ms step_avg:168.73ms
step:804/1530 train_loss:3.6283 train_time:133978ms step_avg:168.74ms
step:805/1530 train_loss:3.5232 train_time:134154ms step_avg:168.75ms
step:806/1530 train_loss:3.5695 train_time:134326ms step_avg:168.75ms
step:807/1530 train_loss:3.6483 train_time:134499ms step_avg:168.76ms
step:808/1530 train_loss:3.5506 train_time:134675ms step_avg:168.77ms
step:809/1530 train_loss:3.4997 train_time:134848ms step_avg:168.77ms
step:810/1530 train_loss:3.5665 train_time:135019ms step_avg:168.77ms
step:811/1530 train_loss:3.5890 train_time:135194ms step_avg:168.78ms
step:812/1530 train_loss:3.6125 train_time:135368ms step_avg:168.79ms
step:813/1530 train_loss:3.6359 train_time:135539ms step_avg:168.79ms
step:814/1530 train_loss:3.5732 train_time:135713ms step_avg:168.80ms
step:815/1530 train_loss:3.5698 train_time:135886ms step_avg:168.80ms
step:816/1530 train_loss:3.6898 train_time:136060ms step_avg:168.81ms
step:817/1530 train_loss:3.7710 train_time:136233ms step_avg:168.81ms
step:818/1530 train_loss:3.5302 train_time:136403ms step_avg:168.82ms
step:819/1530 train_loss:3.7296 train_time:136578ms step_avg:168.82ms
step:820/1530 train_loss:3.4993 train_time:136752ms step_avg:168.83ms
step:821/1530 train_loss:3.5669 train_time:136923ms step_avg:168.83ms
step:822/1530 train_loss:3.7018 train_time:137099ms step_avg:168.84ms
step:823/1530 train_loss:3.5815 train_time:137274ms step_avg:168.85ms
step:824/1530 train_loss:3.5191 train_time:137446ms step_avg:168.85ms
step:825/1530 train_loss:3.6231 train_time:137622ms step_avg:168.86ms
step:826/1530 train_loss:3.4915 train_time:137798ms step_avg:168.87ms
step:827/1530 train_loss:3.7393 train_time:137973ms step_avg:168.88ms
step:828/1530 train_loss:3.6297 train_time:138145ms step_avg:168.88ms
step:829/1530 train_loss:3.6399 train_time:138321ms step_avg:168.89ms
step:830/1530 train_loss:3.5434 train_time:138496ms step_avg:168.90ms
step:831/1530 train_loss:3.6015 train_time:138668ms step_avg:168.90ms
step:832/1530 train_loss:3.5202 train_time:138841ms step_avg:168.91ms
step:833/1530 train_loss:3.6601 train_time:139016ms step_avg:168.91ms
step:834/1530 train_loss:3.4807 train_time:139188ms step_avg:168.92ms
step:835/1530 train_loss:3.4627 train_time:139363ms step_avg:168.92ms
step:836/1530 train_loss:3.7206 train_time:139538ms step_avg:168.93ms
step:837/1530 train_loss:3.4096 train_time:139712ms step_avg:168.94ms
step:838/1530 train_loss:3.6018 train_time:139885ms step_avg:168.94ms
step:839/1530 train_loss:3.4332 train_time:140059ms step_avg:168.95ms
step:840/1530 train_loss:3.4793 train_time:140232ms step_avg:168.95ms
step:841/1530 train_loss:3.5748 train_time:140405ms step_avg:168.96ms
step:842/1530 train_loss:3.5866 train_time:140581ms step_avg:168.97ms
step:843/1530 train_loss:3.5643 train_time:140754ms step_avg:168.97ms
step:844/1530 train_loss:3.4344 train_time:140926ms step_avg:168.98ms
step:845/1530 train_loss:3.6688 train_time:141100ms step_avg:168.98ms
step:846/1530 train_loss:3.5235 train_time:141276ms step_avg:168.99ms
step:847/1530 train_loss:3.5019 train_time:141451ms step_avg:169.00ms
step:848/1530 train_loss:3.6484 train_time:141624ms step_avg:169.00ms
step:849/1530 train_loss:3.4992 train_time:141799ms step_avg:169.01ms
step:850/1530 train_loss:3.4505 train_time:141974ms step_avg:169.02ms
step:851/1530 train_loss:3.7465 train_time:142148ms step_avg:169.02ms
step:852/1530 train_loss:3.4482 train_time:142320ms step_avg:169.03ms
step:853/1530 train_loss:3.5695 train_time:142492ms step_avg:169.03ms
step:854/1530 train_loss:3.6562 train_time:142667ms step_avg:169.04ms
step:855/1530 train_loss:3.5259 train_time:142839ms step_avg:169.04ms
step:856/1530 train_loss:3.5530 train_time:143012ms step_avg:169.05ms
step:857/1530 train_loss:3.6083 train_time:143185ms step_avg:169.05ms
step:858/1530 train_loss:3.4720 train_time:143361ms step_avg:169.06ms
step:859/1530 train_loss:3.5671 train_time:143537ms step_avg:169.07ms
step:860/1530 train_loss:3.5959 train_time:143709ms step_avg:169.07ms
step:861/1530 train_loss:3.6398 train_time:143886ms step_avg:169.08ms
step:862/1530 train_loss:3.6101 train_time:144064ms step_avg:169.09ms
step:863/1530 train_loss:3.5745 train_time:144239ms step_avg:169.10ms
step:864/1530 train_loss:3.3860 train_time:144414ms step_avg:169.10ms
step:865/1530 train_loss:3.6060 train_time:144584ms step_avg:169.10ms
step:866/1530 train_loss:3.9134 train_time:144763ms step_avg:169.12ms
step:867/1530 train_loss:3.4615 train_time:144937ms step_avg:169.12ms
step:868/1530 train_loss:3.6474 train_time:145109ms step_avg:169.12ms
step:869/1530 train_loss:3.6212 train_time:145281ms step_avg:169.13ms
step:870/1530 train_loss:3.4549 train_time:145456ms step_avg:169.14ms
step:871/1530 train_loss:3.3967 train_time:145629ms step_avg:169.14ms
step:872/1530 train_loss:3.6581 train_time:145802ms step_avg:169.14ms
step:873/1530 train_loss:3.4680 train_time:145976ms step_avg:169.15ms
step:874/1530 train_loss:3.2263 train_time:146155ms step_avg:169.16ms
step:875/1530 train_loss:3.6373 train_time:146329ms step_avg:169.17ms
step:875/1530 val_loss:3.5274 train_time:146377ms step_avg:169.22ms
step:876/1530 train_loss:3.4472 train_time:146501ms step_avg:169.17ms
step:877/1530 train_loss:3.6259 train_time:146677ms step_avg:169.18ms
step:878/1530 train_loss:3.4833 train_time:146850ms step_avg:169.18ms
step:879/1530 train_loss:3.6587 train_time:147021ms step_avg:169.18ms
step:880/1530 train_loss:3.3206 train_time:147194ms step_avg:169.19ms
step:881/1530 train_loss:3.4799 train_time:147366ms step_avg:169.19ms
step:882/1530 train_loss:3.7091 train_time:147540ms step_avg:169.20ms
step:883/1530 train_loss:3.8490 train_time:147713ms step_avg:169.20ms
step:884/1530 train_loss:3.5737 train_time:147889ms step_avg:169.21ms
step:885/1530 train_loss:3.5069 train_time:148061ms step_avg:169.21ms
step:886/1530 train_loss:3.5762 train_time:148235ms step_avg:169.22ms
step:887/1530 train_loss:4.0856 train_time:148410ms step_avg:169.22ms
step:888/1530 train_loss:3.8386 train_time:148590ms step_avg:169.24ms
step:889/1530 train_loss:3.5259 train_time:148761ms step_avg:169.24ms
step:890/1530 train_loss:3.5373 train_time:148935ms step_avg:169.24ms
step:891/1530 train_loss:3.3696 train_time:149108ms step_avg:169.25ms
step:892/1530 train_loss:3.7213 train_time:149281ms step_avg:169.25ms
step:893/1530 train_loss:3.4264 train_time:149453ms step_avg:169.26ms
step:894/1530 train_loss:3.6490 train_time:149628ms step_avg:169.26ms
step:895/1530 train_loss:3.6861 train_time:149801ms step_avg:169.27ms
step:896/1530 train_loss:3.5076 train_time:149975ms step_avg:169.27ms
step:897/1530 train_loss:3.5478 train_time:150151ms step_avg:169.28ms
step:898/1530 train_loss:3.5985 train_time:150326ms step_avg:169.29ms
step:899/1530 train_loss:3.4812 train_time:150498ms step_avg:169.29ms
step:900/1530 train_loss:3.4296 train_time:150672ms step_avg:169.29ms
step:901/1530 train_loss:3.6248 train_time:150843ms step_avg:169.30ms
step:902/1530 train_loss:3.6403 train_time:151016ms step_avg:169.30ms
step:903/1530 train_loss:3.5484 train_time:151193ms step_avg:169.31ms
step:904/1530 train_loss:3.4982 train_time:151366ms step_avg:169.31ms
step:905/1530 train_loss:3.5056 train_time:151536ms step_avg:169.31ms
step:906/1530 train_loss:3.7129 train_time:151712ms step_avg:169.32ms
step:907/1530 train_loss:3.5234 train_time:151887ms step_avg:169.33ms
step:908/1530 train_loss:3.5694 train_time:152060ms step_avg:169.33ms
step:909/1530 train_loss:3.4611 train_time:152236ms step_avg:169.34ms
step:910/1530 train_loss:3.5377 train_time:152416ms step_avg:169.35ms
step:911/1530 train_loss:3.6494 train_time:152593ms step_avg:169.36ms
step:912/1530 train_loss:3.6097 train_time:152771ms step_avg:169.37ms
step:913/1530 train_loss:3.4727 train_time:152951ms step_avg:169.38ms
step:914/1530 train_loss:3.7552 train_time:153128ms step_avg:169.39ms
step:915/1530 train_loss:3.5376 train_time:153306ms step_avg:169.40ms
step:916/1530 train_loss:3.6276 train_time:153481ms step_avg:169.40ms
step:917/1530 train_loss:3.6119 train_time:153655ms step_avg:169.41ms
step:918/1530 train_loss:4.8389 train_time:153836ms step_avg:169.42ms
step:919/1530 train_loss:3.5019 train_time:154015ms step_avg:169.43ms
step:920/1530 train_loss:3.5969 train_time:154190ms step_avg:169.44ms
step:921/1530 train_loss:3.5556 train_time:154366ms step_avg:169.45ms
step:922/1530 train_loss:3.5873 train_time:154542ms step_avg:169.45ms
step:923/1530 train_loss:3.6154 train_time:154718ms step_avg:169.46ms
step:924/1530 train_loss:3.6910 train_time:154895ms step_avg:169.47ms
step:925/1530 train_loss:3.6539 train_time:155070ms step_avg:169.48ms
step:926/1530 train_loss:3.5633 train_time:155242ms step_avg:169.48ms
step:927/1530 train_loss:3.5609 train_time:155417ms step_avg:169.48ms
step:928/1530 train_loss:3.7879 train_time:155594ms step_avg:169.49ms
step:929/1530 train_loss:3.6212 train_time:155769ms step_avg:169.50ms
step:930/1530 train_loss:3.4114 train_time:155945ms step_avg:169.51ms
step:931/1530 train_loss:3.5018 train_time:156118ms step_avg:169.51ms
step:932/1530 train_loss:3.6556 train_time:156296ms step_avg:169.52ms
step:933/1530 train_loss:3.3708 train_time:156473ms step_avg:169.53ms
step:934/1530 train_loss:3.5919 train_time:156651ms step_avg:169.54ms
step:935/1530 train_loss:3.4488 train_time:156831ms step_avg:169.55ms
step:936/1530 train_loss:3.5203 train_time:157008ms step_avg:169.55ms
step:937/1530 train_loss:3.6306 train_time:157186ms step_avg:169.56ms
step:938/1530 train_loss:3.5470 train_time:157359ms step_avg:169.57ms
step:939/1530 train_loss:3.6811 train_time:157540ms step_avg:169.58ms
step:940/1530 train_loss:3.4895 train_time:157714ms step_avg:169.59ms
step:941/1530 train_loss:3.5560 train_time:157889ms step_avg:169.59ms
step:942/1530 train_loss:3.3646 train_time:158066ms step_avg:169.60ms
step:943/1530 train_loss:3.7122 train_time:158245ms step_avg:169.61ms
step:944/1530 train_loss:3.4077 train_time:158557ms step_avg:169.76ms
step:945/1530 train_loss:3.4349 train_time:158741ms step_avg:169.78ms
step:946/1530 train_loss:5.0842 train_time:158921ms step_avg:169.79ms
step:947/1530 train_loss:3.6064 train_time:159098ms step_avg:169.80ms
step:948/1530 train_loss:3.4923 train_time:159274ms step_avg:169.80ms
step:949/1530 train_loss:3.3846 train_time:159595ms step_avg:169.96ms
step:950/1530 train_loss:3.4492 train_time:159770ms step_avg:169.97ms
step:951/1530 train_loss:3.4153 train_time:159948ms step_avg:169.98ms
step:952/1530 train_loss:3.4843 train_time:160123ms step_avg:169.98ms
step:953/1530 train_loss:3.5771 train_time:160301ms step_avg:169.99ms
step:954/1530 train_loss:3.4538 train_time:160480ms step_avg:170.00ms
step:955/1530 train_loss:3.4834 train_time:160655ms step_avg:170.00ms
step:956/1530 train_loss:3.4432 train_time:160831ms step_avg:170.01ms
step:957/1530 train_loss:3.5001 train_time:161008ms step_avg:170.02ms
step:958/1530 train_loss:3.5089 train_time:161189ms step_avg:170.03ms
step:959/1530 train_loss:3.5122 train_time:161363ms step_avg:170.03ms
step:960/1530 train_loss:3.4088 train_time:161539ms step_avg:170.04ms
step:961/1530 train_loss:3.6546 train_time:161715ms step_avg:170.05ms
step:962/1530 train_loss:3.5961 train_time:161891ms step_avg:170.05ms
step:963/1530 train_loss:3.8134 train_time:162066ms step_avg:170.06ms
step:964/1530 train_loss:3.4343 train_time:162242ms step_avg:170.07ms
step:965/1530 train_loss:3.4855 train_time:162415ms step_avg:170.07ms
step:966/1530 train_loss:3.7114 train_time:162591ms step_avg:170.07ms
step:967/1530 train_loss:3.5289 train_time:162764ms step_avg:170.08ms
step:968/1530 train_loss:3.5249 train_time:162939ms step_avg:170.08ms
step:969/1530 train_loss:3.5877 train_time:163114ms step_avg:170.09ms
step:970/1530 train_loss:3.3838 train_time:163288ms step_avg:170.09ms
step:971/1530 train_loss:3.5378 train_time:163461ms step_avg:170.09ms
step:972/1530 train_loss:3.4836 train_time:163634ms step_avg:170.10ms
step:973/1530 train_loss:3.5477 train_time:163807ms step_avg:170.10ms
step:974/1530 train_loss:3.5987 train_time:163983ms step_avg:170.11ms
step:975/1530 train_loss:3.4726 train_time:164158ms step_avg:170.11ms
step:976/1530 train_loss:3.6812 train_time:164333ms step_avg:170.12ms
step:977/1530 train_loss:3.5782 train_time:164506ms step_avg:170.12ms
step:978/1530 train_loss:3.3689 train_time:164681ms step_avg:170.12ms
step:979/1530 train_loss:3.6386 train_time:164857ms step_avg:170.13ms
step:980/1530 train_loss:3.4196 train_time:165032ms step_avg:170.14ms
step:981/1530 train_loss:3.5833 train_time:165212ms step_avg:170.15ms
step:982/1530 train_loss:3.5562 train_time:165386ms step_avg:170.15ms
step:983/1530 train_loss:3.5225 train_time:165562ms step_avg:170.16ms
step:984/1530 train_loss:3.5015 train_time:165735ms step_avg:170.16ms
step:985/1530 train_loss:3.5816 train_time:165912ms step_avg:170.17ms
step:986/1530 train_loss:3.4156 train_time:166088ms step_avg:170.17ms
step:987/1530 train_loss:3.4963 train_time:166260ms step_avg:170.17ms
step:988/1530 train_loss:3.4763 train_time:166434ms step_avg:170.18ms
step:989/1530 train_loss:3.4299 train_time:166607ms step_avg:170.18ms
step:990/1530 train_loss:3.6601 train_time:166783ms step_avg:170.19ms
step:991/1530 train_loss:3.4711 train_time:166958ms step_avg:170.19ms
step:992/1530 train_loss:3.4481 train_time:167138ms step_avg:170.20ms
step:993/1530 train_loss:3.5071 train_time:167318ms step_avg:170.21ms
step:994/1530 train_loss:3.6024 train_time:167493ms step_avg:170.22ms
step:995/1530 train_loss:3.5357 train_time:167664ms step_avg:170.22ms
step:996/1530 train_loss:3.4618 train_time:167837ms step_avg:170.22ms
step:997/1530 train_loss:3.7618 train_time:168011ms step_avg:170.22ms
step:998/1530 train_loss:3.4466 train_time:168184ms step_avg:170.23ms
step:999/1530 train_loss:3.5911 train_time:168358ms step_avg:170.23ms
step:1000/1530 train_loss:3.4473 train_time:168535ms step_avg:170.24ms
step:1000/1530 val_loss:3.4743 train_time:168586ms step_avg:170.29ms
step:1001/1530 train_loss:3.5062 train_time:168711ms step_avg:170.24ms
step:1002/1530 train_loss:3.3793 train_time:168887ms step_avg:170.25ms
step:1003/1530 train_loss:3.5698 train_time:169064ms step_avg:170.26ms
step:1004/1530 train_loss:3.6092 train_time:169240ms step_avg:170.26ms
step:1005/1530 train_loss:3.3966 train_time:169414ms step_avg:170.26ms
step:1006/1530 train_loss:3.4713 train_time:169591ms step_avg:170.27ms
step:1007/1530 train_loss:3.4464 train_time:169766ms step_avg:170.28ms
step:1008/1530 train_loss:3.5670 train_time:169943ms step_avg:170.28ms
step:1009/1530 train_loss:3.6690 train_time:170122ms step_avg:170.29ms
step:1010/1530 train_loss:3.5677 train_time:170294ms step_avg:170.29ms
step:1011/1530 train_loss:3.5394 train_time:170468ms step_avg:170.30ms
step:1012/1530 train_loss:3.3962 train_time:170643ms step_avg:170.30ms
step:1013/1530 train_loss:3.5410 train_time:170819ms step_avg:170.31ms
step:1014/1530 train_loss:3.6283 train_time:170997ms step_avg:170.32ms
step:1015/1530 train_loss:3.3335 train_time:171174ms step_avg:170.32ms
step:1016/1530 train_loss:3.4107 train_time:171348ms step_avg:170.33ms
step:1017/1530 train_loss:3.4066 train_time:171525ms step_avg:170.33ms
step:1018/1530 train_loss:3.4021 train_time:171700ms step_avg:170.34ms
step:1019/1530 train_loss:3.5262 train_time:171875ms step_avg:170.34ms
step:1020/1530 train_loss:3.3835 train_time:172051ms step_avg:170.35ms
step:1021/1530 train_loss:3.3607 train_time:172227ms step_avg:170.35ms
step:1022/1530 train_loss:3.4867 train_time:172405ms step_avg:170.36ms
step:1023/1530 train_loss:3.5124 train_time:172581ms step_avg:170.37ms
step:1024/1530 train_loss:3.4875 train_time:172757ms step_avg:170.37ms
step:1025/1530 train_loss:3.4825 train_time:172935ms step_avg:170.38ms
step:1026/1530 train_loss:3.6271 train_time:173110ms step_avg:170.38ms
step:1027/1530 train_loss:3.3248 train_time:173287ms step_avg:170.39ms
step:1028/1530 train_loss:3.4011 train_time:173468ms step_avg:170.40ms
step:1029/1530 train_loss:3.3244 train_time:173648ms step_avg:170.41ms
step:1030/1530 train_loss:3.5433 train_time:173824ms step_avg:170.42ms
step:1031/1530 train_loss:3.5127 train_time:174002ms step_avg:170.42ms
step:1032/1530 train_loss:3.7001 train_time:174182ms step_avg:170.43ms
step:1033/1530 train_loss:3.4946 train_time:174357ms step_avg:170.44ms
step:1034/1530 train_loss:3.4070 train_time:174534ms step_avg:170.44ms
step:1035/1530 train_loss:3.4456 train_time:174712ms step_avg:170.45ms
step:1036/1530 train_loss:3.4867 train_time:174889ms step_avg:170.46ms
step:1037/1530 train_loss:3.7975 train_time:175066ms step_avg:170.46ms
step:1038/1530 train_loss:3.6258 train_time:175246ms step_avg:170.47ms
step:1039/1530 train_loss:3.5181 train_time:175428ms step_avg:170.48ms
step:1040/1530 train_loss:3.4168 train_time:175605ms step_avg:170.49ms
step:1041/1530 train_loss:3.4925 train_time:175781ms step_avg:170.50ms
step:1042/1530 train_loss:3.5257 train_time:175953ms step_avg:170.50ms
step:1043/1530 train_loss:3.4511 train_time:176128ms step_avg:170.50ms
step:1044/1530 train_loss:3.4633 train_time:176305ms step_avg:170.51ms
step:1045/1530 train_loss:3.5210 train_time:176484ms step_avg:170.52ms
step:1046/1530 train_loss:3.4288 train_time:176659ms step_avg:170.52ms
step:1047/1530 train_loss:3.6409 train_time:176833ms step_avg:170.52ms
step:1048/1530 train_loss:3.4996 train_time:177009ms step_avg:170.53ms
step:1049/1530 train_loss:3.4052 train_time:177185ms step_avg:170.53ms
step:1050/1530 train_loss:3.3980 train_time:177362ms step_avg:170.54ms
step:1051/1530 train_loss:3.5020 train_time:177539ms step_avg:170.55ms
step:1052/1530 train_loss:3.3676 train_time:177716ms step_avg:170.55ms
step:1053/1530 train_loss:3.6976 train_time:177893ms step_avg:170.56ms
step:1054/1530 train_loss:3.5413 train_time:178072ms step_avg:170.57ms
step:1055/1530 train_loss:3.3899 train_time:178247ms step_avg:170.57ms
step:1056/1530 train_loss:3.4999 train_time:178421ms step_avg:170.57ms
step:1057/1530 train_loss:3.5852 train_time:178600ms step_avg:170.58ms
step:1058/1530 train_loss:3.3104 train_time:178777ms step_avg:170.59ms
step:1059/1530 train_loss:3.3808 train_time:178957ms step_avg:170.60ms
step:1060/1530 train_loss:3.4415 train_time:179133ms step_avg:170.60ms
step:1061/1530 train_loss:3.4238 train_time:179308ms step_avg:170.61ms
step:1062/1530 train_loss:3.3847 train_time:179485ms step_avg:170.61ms
step:1063/1530 train_loss:3.4661 train_time:179659ms step_avg:170.62ms
step:1064/1530 train_loss:3.3894 train_time:179832ms step_avg:170.62ms
step:1065/1530 train_loss:3.3632 train_time:180011ms step_avg:170.63ms
step:1066/1530 train_loss:3.4190 train_time:180188ms step_avg:170.63ms
step:1067/1530 train_loss:3.2955 train_time:180367ms step_avg:170.64ms
step:1068/1530 train_loss:3.4399 train_time:180544ms step_avg:170.65ms
step:1069/1530 train_loss:3.3017 train_time:180723ms step_avg:170.65ms
step:1070/1530 train_loss:3.5750 train_time:180898ms step_avg:170.66ms
step:1071/1530 train_loss:3.5152 train_time:181076ms step_avg:170.67ms
step:1072/1530 train_loss:3.4442 train_time:181250ms step_avg:170.67ms
step:1073/1530 train_loss:3.5266 train_time:181424ms step_avg:170.67ms
step:1074/1530 train_loss:3.4380 train_time:181601ms step_avg:170.68ms
step:1075/1530 train_loss:3.4035 train_time:181777ms step_avg:170.68ms
step:1076/1530 train_loss:3.8023 train_time:181954ms step_avg:170.69ms
step:1077/1530 train_loss:3.4457 train_time:182129ms step_avg:170.69ms
step:1078/1530 train_loss:3.1028 train_time:182315ms step_avg:170.71ms
step:1079/1530 train_loss:3.5413 train_time:182491ms step_avg:170.71ms
step:1080/1530 train_loss:3.4327 train_time:182669ms step_avg:170.72ms
step:1081/1530 train_loss:3.5088 train_time:182843ms step_avg:170.72ms
step:1082/1530 train_loss:3.5942 train_time:183017ms step_avg:170.73ms
step:1083/1530 train_loss:3.5025 train_time:183192ms step_avg:170.73ms
step:1084/1530 train_loss:3.4722 train_time:183368ms step_avg:170.73ms
step:1085/1530 train_loss:3.4411 train_time:183544ms step_avg:170.74ms
step:1086/1530 train_loss:3.6362 train_time:183720ms step_avg:170.74ms
step:1087/1530 train_loss:3.5099 train_time:183896ms step_avg:170.75ms
step:1088/1530 train_loss:3.3772 train_time:184073ms step_avg:170.75ms
step:1089/1530 train_loss:3.3821 train_time:184251ms step_avg:170.76ms
step:1090/1530 train_loss:3.4902 train_time:184430ms step_avg:170.77ms
step:1091/1530 train_loss:3.2882 train_time:184609ms step_avg:170.78ms
step:1092/1530 train_loss:3.4892 train_time:184786ms step_avg:170.78ms
step:1093/1530 train_loss:3.6096 train_time:184963ms step_avg:170.79ms
step:1094/1530 train_loss:3.4504 train_time:185138ms step_avg:170.79ms
step:1095/1530 train_loss:3.4276 train_time:185312ms step_avg:170.79ms
step:1096/1530 train_loss:3.4324 train_time:185490ms step_avg:170.80ms
step:1097/1530 train_loss:3.4931 train_time:185669ms step_avg:170.81ms
step:1098/1530 train_loss:3.5651 train_time:185848ms step_avg:170.82ms
step:1099/1530 train_loss:3.5329 train_time:186026ms step_avg:170.82ms
step:1100/1530 train_loss:3.4328 train_time:186207ms step_avg:170.83ms
step:1101/1530 train_loss:3.2960 train_time:186386ms step_avg:170.84ms
step:1102/1530 train_loss:3.3180 train_time:186565ms step_avg:170.85ms
step:1103/1530 train_loss:3.4470 train_time:186746ms step_avg:170.86ms
step:1104/1530 train_loss:3.3274 train_time:186922ms step_avg:170.86ms
step:1105/1530 train_loss:4.0731 train_time:187099ms step_avg:170.87ms
step:1106/1530 train_loss:3.2326 train_time:187273ms step_avg:170.87ms
step:1107/1530 train_loss:3.5740 train_time:187448ms step_avg:170.87ms
step:1108/1530 train_loss:3.3555 train_time:187622ms step_avg:170.88ms
step:1109/1530 train_loss:3.5120 train_time:187796ms step_avg:170.88ms
step:1110/1530 train_loss:3.4342 train_time:187970ms step_avg:170.88ms
step:1111/1530 train_loss:3.4878 train_time:188145ms step_avg:170.89ms
step:1112/1530 train_loss:3.5686 train_time:188325ms step_avg:170.89ms
step:1113/1530 train_loss:3.4375 train_time:188509ms step_avg:170.91ms
step:1114/1530 train_loss:3.3741 train_time:188689ms step_avg:170.91ms
step:1115/1530 train_loss:3.2432 train_time:188868ms step_avg:170.92ms
step:1116/1530 train_loss:3.4303 train_time:189043ms step_avg:170.93ms
step:1117/1530 train_loss:3.5915 train_time:189222ms step_avg:170.93ms
step:1118/1530 train_loss:3.6295 train_time:189399ms step_avg:170.94ms
step:1119/1530 train_loss:3.4834 train_time:189574ms step_avg:170.94ms
step:1120/1530 train_loss:3.4970 train_time:189751ms step_avg:170.95ms
step:1121/1530 train_loss:3.3956 train_time:189928ms step_avg:170.95ms
step:1122/1530 train_loss:3.4644 train_time:190105ms step_avg:170.96ms
step:1123/1530 train_loss:3.5853 train_time:190283ms step_avg:170.96ms
step:1124/1530 train_loss:3.3450 train_time:190457ms step_avg:170.97ms
step:1125/1530 train_loss:3.2333 train_time:190634ms step_avg:170.97ms
step:1125/1530 val_loss:3.4147 train_time:190684ms step_avg:171.02ms
step:1126/1530 train_loss:3.4806 train_time:190813ms step_avg:170.98ms
step:1127/1530 train_loss:3.6812 train_time:190989ms step_avg:170.98ms
step:1128/1530 train_loss:3.2350 train_time:191168ms step_avg:170.99ms
step:1129/1530 train_loss:3.5578 train_time:191346ms step_avg:171.00ms
step:1130/1530 train_loss:3.3842 train_time:191524ms step_avg:171.00ms
step:1131/1530 train_loss:3.4067 train_time:191707ms step_avg:171.01ms
step:1132/1530 train_loss:3.3734 train_time:191880ms step_avg:171.02ms
step:1133/1530 train_loss:3.4968 train_time:192190ms step_avg:171.14ms
step:1134/1530 train_loss:3.4552 train_time:192377ms step_avg:171.15ms
step:1135/1530 train_loss:3.5275 train_time:192554ms step_avg:171.16ms
step:1136/1530 train_loss:3.5670 train_time:192734ms step_avg:171.17ms
step:1137/1530 train_loss:3.4682 train_time:192911ms step_avg:171.17ms
step:1138/1530 train_loss:3.3597 train_time:193089ms step_avg:171.18ms
step:1139/1530 train_loss:3.6565 train_time:193414ms step_avg:171.31ms
step:1140/1530 train_loss:3.4602 train_time:193589ms step_avg:171.32ms
step:1141/1530 train_loss:3.5989 train_time:193770ms step_avg:171.33ms
step:1142/1530 train_loss:3.4505 train_time:193946ms step_avg:171.33ms
step:1143/1530 train_loss:3.3729 train_time:194125ms step_avg:171.34ms
step:1144/1530 train_loss:3.4488 train_time:194302ms step_avg:171.34ms
step:1145/1530 train_loss:3.5923 train_time:194476ms step_avg:171.34ms
step:1146/1530 train_loss:3.5619 train_time:194658ms step_avg:171.35ms
step:1147/1530 train_loss:3.4942 train_time:194837ms step_avg:171.36ms
step:1148/1530 train_loss:3.5039 train_time:195016ms step_avg:171.37ms
step:1149/1530 train_loss:3.3294 train_time:195195ms step_avg:171.37ms
step:1150/1530 train_loss:3.3785 train_time:195371ms step_avg:171.38ms
step:1151/1530 train_loss:3.3265 train_time:195549ms step_avg:171.38ms
step:1152/1530 train_loss:3.4037 train_time:195731ms step_avg:171.39ms
step:1153/1530 train_loss:3.4386 train_time:195912ms step_avg:171.40ms
step:1154/1530 train_loss:3.5265 train_time:196087ms step_avg:171.40ms
step:1155/1530 train_loss:3.3244 train_time:196268ms step_avg:171.41ms
step:1156/1530 train_loss:3.5465 train_time:196452ms step_avg:171.42ms
step:1157/1530 train_loss:3.4965 train_time:196631ms step_avg:171.43ms
step:1158/1530 train_loss:3.2549 train_time:196809ms step_avg:171.44ms
step:1159/1530 train_loss:3.3541 train_time:196985ms step_avg:171.44ms
step:1160/1530 train_loss:3.3436 train_time:197159ms step_avg:171.44ms
step:1161/1530 train_loss:3.0955 train_time:197339ms step_avg:171.45ms
step:1162/1530 train_loss:3.4262 train_time:197517ms step_avg:171.46ms
step:1163/1530 train_loss:3.3943 train_time:197697ms step_avg:171.46ms
step:1164/1530 train_loss:3.2995 train_time:197875ms step_avg:171.47ms
step:1165/1530 train_loss:3.2507 train_time:198050ms step_avg:171.47ms
step:1166/1530 train_loss:3.3912 train_time:198230ms step_avg:171.48ms
step:1167/1530 train_loss:3.4174 train_time:198405ms step_avg:171.48ms
step:1168/1530 train_loss:3.7293 train_time:198581ms step_avg:171.49ms
step:1169/1530 train_loss:3.3844 train_time:198758ms step_avg:171.49ms
step:1170/1530 train_loss:3.3914 train_time:198936ms step_avg:171.50ms
step:1171/1530 train_loss:3.2934 train_time:199113ms step_avg:171.50ms
step:1172/1530 train_loss:3.4294 train_time:199288ms step_avg:171.50ms
step:1173/1530 train_loss:3.5409 train_time:199467ms step_avg:171.51ms
step:1174/1530 train_loss:3.3831 train_time:199654ms step_avg:171.52ms
step:1175/1530 train_loss:3.3760 train_time:199834ms step_avg:171.53ms
step:1176/1530 train_loss:3.4341 train_time:200016ms step_avg:171.54ms
step:1177/1530 train_loss:3.4544 train_time:200197ms step_avg:171.55ms
step:1178/1530 train_loss:3.5034 train_time:200373ms step_avg:171.55ms
step:1179/1530 train_loss:3.4077 train_time:200548ms step_avg:171.56ms
step:1180/1530 train_loss:3.3689 train_time:200737ms step_avg:171.57ms
step:1181/1530 train_loss:3.3403 train_time:200915ms step_avg:171.58ms
step:1182/1530 train_loss:3.3770 train_time:201094ms step_avg:171.58ms
step:1183/1530 train_loss:3.3399 train_time:201271ms step_avg:171.59ms
step:1184/1530 train_loss:3.5114 train_time:201448ms step_avg:171.59ms
step:1185/1530 train_loss:3.5496 train_time:201627ms step_avg:171.60ms
step:1186/1530 train_loss:3.3682 train_time:201806ms step_avg:171.60ms
step:1187/1530 train_loss:3.4149 train_time:201993ms step_avg:171.62ms
step:1188/1530 train_loss:3.4440 train_time:202169ms step_avg:171.62ms
step:1189/1530 train_loss:3.2862 train_time:202349ms step_avg:171.63ms
step:1190/1530 train_loss:3.4503 train_time:202528ms step_avg:171.63ms
step:1191/1530 train_loss:3.5838 train_time:202708ms step_avg:171.64ms
step:1192/1530 train_loss:3.3998 train_time:202882ms step_avg:171.64ms
step:1193/1530 train_loss:3.2822 train_time:203056ms step_avg:171.65ms
step:1194/1530 train_loss:3.5636 train_time:203234ms step_avg:171.65ms
step:1195/1530 train_loss:3.3762 train_time:203416ms step_avg:171.66ms
step:1196/1530 train_loss:3.3922 train_time:203603ms step_avg:171.67ms
step:1197/1530 train_loss:3.2967 train_time:203782ms step_avg:171.68ms
step:1198/1530 train_loss:3.3035 train_time:203965ms step_avg:171.69ms
step:1199/1530 train_loss:3.3449 train_time:204144ms step_avg:171.69ms
step:1200/1530 train_loss:3.4528 train_time:204321ms step_avg:171.70ms
step:1201/1530 train_loss:3.4796 train_time:204498ms step_avg:171.70ms
step:1202/1530 train_loss:3.6002 train_time:204688ms step_avg:171.72ms
step:1203/1530 train_loss:3.4107 train_time:204867ms step_avg:171.72ms
step:1204/1530 train_loss:3.3071 train_time:205048ms step_avg:171.73ms
step:1205/1530 train_loss:3.4412 train_time:205224ms step_avg:171.74ms
step:1206/1530 train_loss:3.4759 train_time:205400ms step_avg:171.74ms
step:1207/1530 train_loss:3.5261 train_time:205578ms step_avg:171.74ms
step:1208/1530 train_loss:3.3972 train_time:205754ms step_avg:171.75ms
step:1209/1530 train_loss:3.2482 train_time:205933ms step_avg:171.75ms
step:1210/1530 train_loss:3.3082 train_time:206114ms step_avg:171.76ms
step:1211/1530 train_loss:3.4023 train_time:206291ms step_avg:171.77ms
step:1212/1530 train_loss:3.4030 train_time:206467ms step_avg:171.77ms
step:1213/1530 train_loss:3.4139 train_time:206646ms step_avg:171.78ms
step:1214/1530 train_loss:3.2569 train_time:206826ms step_avg:171.78ms
step:1215/1530 train_loss:3.4011 train_time:207002ms step_avg:171.79ms
step:1216/1530 train_loss:3.3417 train_time:207180ms step_avg:171.79ms
step:1217/1530 train_loss:3.3298 train_time:207357ms step_avg:171.80ms
step:1218/1530 train_loss:3.4167 train_time:207536ms step_avg:171.80ms
step:1219/1530 train_loss:3.2573 train_time:207721ms step_avg:171.81ms
step:1220/1530 train_loss:3.4823 train_time:207898ms step_avg:171.82ms
step:1221/1530 train_loss:3.5073 train_time:208075ms step_avg:171.82ms
step:1222/1530 train_loss:3.4390 train_time:208251ms step_avg:171.82ms
step:1223/1530 train_loss:3.3000 train_time:208428ms step_avg:171.83ms
step:1224/1530 train_loss:3.2601 train_time:208611ms step_avg:171.84ms
step:1225/1530 train_loss:3.3726 train_time:208788ms step_avg:171.84ms
step:1226/1530 train_loss:3.3391 train_time:208969ms step_avg:171.85ms
step:1227/1530 train_loss:3.2848 train_time:209148ms step_avg:171.86ms
step:1228/1530 train_loss:3.4528 train_time:209322ms step_avg:171.86ms
step:1229/1530 train_loss:3.3748 train_time:209502ms step_avg:171.86ms
step:1230/1530 train_loss:3.4072 train_time:209683ms step_avg:171.87ms
step:1231/1530 train_loss:3.5862 train_time:209861ms step_avg:171.88ms
step:1232/1530 train_loss:3.5053 train_time:210041ms step_avg:171.88ms
step:1233/1530 train_loss:3.4317 train_time:210218ms step_avg:171.89ms
step:1234/1530 train_loss:3.5890 train_time:210397ms step_avg:171.89ms
step:1235/1530 train_loss:3.3300 train_time:210577ms step_avg:171.90ms
step:1236/1530 train_loss:3.2949 train_time:210752ms step_avg:171.90ms
step:1237/1530 train_loss:3.2758 train_time:210929ms step_avg:171.91ms
step:1238/1530 train_loss:3.2827 train_time:211113ms step_avg:171.92ms
step:1239/1530 train_loss:3.3385 train_time:211290ms step_avg:171.92ms
step:1240/1530 train_loss:3.3909 train_time:211470ms step_avg:171.93ms
step:1241/1530 train_loss:3.4336 train_time:211650ms step_avg:171.93ms
step:1242/1530 train_loss:3.3039 train_time:211827ms step_avg:171.94ms
step:1243/1530 train_loss:3.4111 train_time:212004ms step_avg:171.94ms
step:1244/1530 train_loss:3.4156 train_time:212178ms step_avg:171.94ms
step:1245/1530 train_loss:3.4171 train_time:212355ms step_avg:171.95ms
step:1246/1530 train_loss:3.2467 train_time:212534ms step_avg:171.95ms
step:1247/1530 train_loss:3.3787 train_time:212710ms step_avg:171.96ms
step:1248/1530 train_loss:3.4317 train_time:212885ms step_avg:171.96ms
step:1249/1530 train_loss:3.4270 train_time:213062ms step_avg:171.96ms
step:1250/1530 train_loss:3.3176 train_time:213241ms step_avg:171.97ms
step:1250/1530 val_loss:3.3614 train_time:213295ms step_avg:172.01ms
step:1251/1530 train_loss:3.4986 train_time:213428ms step_avg:171.98ms
step:1252/1530 train_loss:3.3652 train_time:213604ms step_avg:171.98ms
step:1253/1530 train_loss:3.3179 train_time:213780ms step_avg:171.99ms
step:1254/1530 train_loss:3.4216 train_time:213959ms step_avg:171.99ms
step:1255/1530 train_loss:3.5228 train_time:214149ms step_avg:172.01ms
step:1256/1530 train_loss:3.3170 train_time:214331ms step_avg:172.02ms
step:1257/1530 train_loss:3.3781 train_time:214509ms step_avg:172.02ms
step:1258/1530 train_loss:3.3667 train_time:214692ms step_avg:172.03ms
step:1259/1530 train_loss:3.3360 train_time:214871ms step_avg:172.03ms
step:1260/1530 train_loss:3.2165 train_time:215046ms step_avg:172.04ms
step:1261/1530 train_loss:3.3073 train_time:215226ms step_avg:172.04ms
step:1262/1530 train_loss:3.3342 train_time:215409ms step_avg:172.05ms
step:1263/1530 train_loss:3.2455 train_time:215590ms step_avg:172.06ms
step:1264/1530 train_loss:3.4479 train_time:215767ms step_avg:172.06ms
step:1265/1530 train_loss:3.4318 train_time:215943ms step_avg:172.07ms
step:1266/1530 train_loss:3.4493 train_time:216122ms step_avg:172.07ms
step:1267/1530 train_loss:3.3726 train_time:216301ms step_avg:172.08ms
step:1268/1530 train_loss:3.4207 train_time:216480ms step_avg:172.08ms
step:1269/1530 train_loss:3.2574 train_time:216663ms step_avg:172.09ms
step:1270/1530 train_loss:3.1114 train_time:216839ms step_avg:172.09ms
step:1271/1530 train_loss:3.4067 train_time:217017ms step_avg:172.10ms
step:1272/1530 train_loss:3.3573 train_time:217194ms step_avg:172.10ms
step:1273/1530 train_loss:3.3836 train_time:217376ms step_avg:172.11ms
step:1274/1530 train_loss:3.3638 train_time:217555ms step_avg:172.12ms
step:1275/1530 train_loss:3.4434 train_time:217732ms step_avg:172.12ms
step:1276/1530 train_loss:3.4746 train_time:217907ms step_avg:172.12ms
step:1277/1530 train_loss:3.4139 train_time:218087ms step_avg:172.13ms
step:1278/1530 train_loss:3.4120 train_time:218262ms step_avg:172.13ms
step:1279/1530 train_loss:3.2690 train_time:218444ms step_avg:172.14ms
step:1280/1530 train_loss:3.3701 train_time:218630ms step_avg:172.15ms
step:1281/1530 train_loss:3.4303 train_time:218807ms step_avg:172.15ms
step:1282/1530 train_loss:3.4721 train_time:218983ms step_avg:172.16ms
step:1283/1530 train_loss:3.3448 train_time:219162ms step_avg:172.16ms
step:1284/1530 train_loss:3.3728 train_time:219340ms step_avg:172.17ms
step:1285/1530 train_loss:3.3659 train_time:219518ms step_avg:172.17ms
step:1286/1530 train_loss:3.3403 train_time:219694ms step_avg:172.17ms
step:1287/1530 train_loss:3.4944 train_time:219871ms step_avg:172.18ms
step:1288/1530 train_loss:3.3008 train_time:220051ms step_avg:172.18ms
step:1289/1530 train_loss:3.3843 train_time:220237ms step_avg:172.19ms
step:1290/1530 train_loss:3.4630 train_time:220420ms step_avg:172.20ms
step:1291/1530 train_loss:3.3893 train_time:220599ms step_avg:172.21ms
step:1292/1530 train_loss:3.4854 train_time:220780ms step_avg:172.22ms
step:1293/1530 train_loss:3.5251 train_time:220959ms step_avg:172.22ms
step:1294/1530 train_loss:3.4641 train_time:221138ms step_avg:172.23ms
step:1295/1530 train_loss:3.2896 train_time:221317ms step_avg:172.23ms
step:1296/1530 train_loss:3.3818 train_time:221497ms step_avg:172.24ms
step:1297/1530 train_loss:3.2832 train_time:221675ms step_avg:172.24ms
step:1298/1530 train_loss:3.2818 train_time:221854ms step_avg:172.25ms
step:1299/1530 train_loss:3.4035 train_time:222032ms step_avg:172.25ms
step:1300/1530 train_loss:3.4088 train_time:222209ms step_avg:172.26ms
step:1301/1530 train_loss:3.4085 train_time:222386ms step_avg:172.26ms
step:1302/1530 train_loss:3.5831 train_time:222570ms step_avg:172.27ms
step:1303/1530 train_loss:3.3136 train_time:222753ms step_avg:172.28ms
step:1304/1530 train_loss:3.5268 train_time:222933ms step_avg:172.28ms
step:1305/1530 train_loss:3.2714 train_time:223111ms step_avg:172.29ms
step:1306/1530 train_loss:3.4565 train_time:223292ms step_avg:172.29ms
step:1307/1530 train_loss:3.4607 train_time:223467ms step_avg:172.30ms
step:1308/1530 train_loss:3.2940 train_time:223645ms step_avg:172.30ms
step:1309/1530 train_loss:3.3169 train_time:223823ms step_avg:172.30ms
step:1310/1530 train_loss:3.2899 train_time:224001ms step_avg:172.31ms
step:1311/1530 train_loss:3.3020 train_time:224177ms step_avg:172.31ms
step:1312/1530 train_loss:3.3836 train_time:224357ms step_avg:172.32ms
step:1313/1530 train_loss:3.3465 train_time:224533ms step_avg:172.32ms
step:1314/1530 train_loss:3.0497 train_time:224716ms step_avg:172.33ms
step:1315/1530 train_loss:3.2852 train_time:224893ms step_avg:172.33ms
step:1316/1530 train_loss:3.4074 train_time:225069ms step_avg:172.33ms
step:1317/1530 train_loss:3.4247 train_time:225246ms step_avg:172.34ms
step:1318/1530 train_loss:3.3081 train_time:225433ms step_avg:172.35ms
step:1319/1530 train_loss:3.4341 train_time:225613ms step_avg:172.36ms
step:1320/1530 train_loss:3.4676 train_time:225796ms step_avg:172.36ms
step:1321/1530 train_loss:3.3714 train_time:225975ms step_avg:172.37ms
step:1322/1530 train_loss:3.3259 train_time:226287ms step_avg:172.47ms
step:1323/1530 train_loss:3.3279 train_time:226476ms step_avg:172.49ms
step:1324/1530 train_loss:3.4406 train_time:226655ms step_avg:172.49ms
step:1325/1530 train_loss:3.4997 train_time:226839ms step_avg:172.50ms
step:1326/1530 train_loss:3.2208 train_time:227018ms step_avg:172.51ms
step:1327/1530 train_loss:3.1732 train_time:227195ms step_avg:172.51ms
step:1328/1530 train_loss:3.4962 train_time:227375ms step_avg:172.51ms
step:1329/1530 train_loss:3.3077 train_time:227710ms step_avg:172.64ms
step:1330/1530 train_loss:3.4335 train_time:227892ms step_avg:172.65ms
step:1331/1530 train_loss:3.3370 train_time:228068ms step_avg:172.65ms
step:1332/1530 train_loss:3.7445 train_time:228250ms step_avg:172.66ms
step:1333/1530 train_loss:3.4843 train_time:228431ms step_avg:172.66ms
step:1334/1530 train_loss:3.3747 train_time:228610ms step_avg:172.67ms
step:1335/1530 train_loss:3.2985 train_time:228791ms step_avg:172.67ms
step:1336/1530 train_loss:3.3009 train_time:228974ms step_avg:172.68ms
step:1337/1530 train_loss:3.5516 train_time:229153ms step_avg:172.68ms
step:1338/1530 train_loss:3.5307 train_time:229333ms step_avg:172.69ms
step:1339/1530 train_loss:3.3446 train_time:229512ms step_avg:172.70ms
step:1340/1530 train_loss:3.2942 train_time:229690ms step_avg:172.70ms
step:1341/1530 train_loss:3.5924 train_time:229868ms step_avg:172.70ms
step:1342/1530 train_loss:3.3628 train_time:230048ms step_avg:172.71ms
step:1343/1530 train_loss:3.3657 train_time:230226ms step_avg:172.71ms
step:1344/1530 train_loss:3.4178 train_time:230406ms step_avg:172.72ms
step:1345/1530 train_loss:3.3900 train_time:230588ms step_avg:172.73ms
step:1346/1530 train_loss:3.3004 train_time:230765ms step_avg:172.73ms
step:1347/1530 train_loss:3.2821 train_time:230941ms step_avg:172.73ms
step:1348/1530 train_loss:3.3539 train_time:231117ms step_avg:172.73ms
step:1349/1530 train_loss:3.2759 train_time:231294ms step_avg:172.74ms
step:1350/1530 train_loss:3.3980 train_time:231475ms step_avg:172.74ms
step:1351/1530 train_loss:3.2510 train_time:231651ms step_avg:172.74ms
step:1352/1530 train_loss:3.3177 train_time:231829ms step_avg:172.75ms
step:1353/1530 train_loss:3.4108 train_time:232010ms step_avg:172.76ms
step:1354/1530 train_loss:3.2664 train_time:232188ms step_avg:172.76ms
step:1355/1530 train_loss:3.1947 train_time:232363ms step_avg:172.76ms
step:1356/1530 train_loss:3.5205 train_time:232541ms step_avg:172.76ms
step:1357/1530 train_loss:3.4275 train_time:232720ms step_avg:172.77ms
step:1358/1530 train_loss:3.1919 train_time:232897ms step_avg:172.77ms
step:1359/1530 train_loss:3.4463 train_time:233076ms step_avg:172.78ms
step:1360/1530 train_loss:3.3548 train_time:233254ms step_avg:172.78ms
step:1361/1530 train_loss:3.1402 train_time:233439ms step_avg:172.79ms
step:1362/1530 train_loss:3.3985 train_time:233620ms step_avg:172.80ms
step:1363/1530 train_loss:3.2839 train_time:233807ms step_avg:172.81ms
step:1364/1530 train_loss:3.3099 train_time:233986ms step_avg:172.81ms
step:1365/1530 train_loss:3.3194 train_time:234163ms step_avg:172.81ms
step:1366/1530 train_loss:3.4279 train_time:234344ms step_avg:172.82ms
step:1367/1530 train_loss:3.4043 train_time:234522ms step_avg:172.82ms
step:1368/1530 train_loss:3.3497 train_time:234701ms step_avg:172.83ms
step:1369/1530 train_loss:3.2790 train_time:234890ms step_avg:172.84ms
step:1370/1530 train_loss:3.6100 train_time:235070ms step_avg:172.85ms
step:1371/1530 train_loss:3.3175 train_time:235250ms step_avg:172.85ms
step:1372/1530 train_loss:3.3764 train_time:235433ms step_avg:172.86ms
step:1373/1530 train_loss:3.3781 train_time:235612ms step_avg:172.86ms
step:1374/1530 train_loss:3.1561 train_time:235793ms step_avg:172.87ms
step:1375/1530 train_loss:3.5425 train_time:235972ms step_avg:172.87ms
step:1375/1530 val_loss:3.3189 train_time:236023ms step_avg:172.91ms
step:1376/1530 train_loss:3.3540 train_time:236153ms step_avg:172.88ms
step:1377/1530 train_loss:3.4872 train_time:236329ms step_avg:172.88ms
step:1378/1530 train_loss:3.4753 train_time:236507ms step_avg:172.89ms
step:1379/1530 train_loss:3.1260 train_time:236689ms step_avg:172.89ms
step:1380/1530 train_loss:3.3188 train_time:236868ms step_avg:172.90ms
step:1381/1530 train_loss:3.7171 train_time:237053ms step_avg:172.90ms
step:1382/1530 train_loss:3.2200 train_time:237230ms step_avg:172.91ms
step:1383/1530 train_loss:3.4007 train_time:237413ms step_avg:172.92ms
step:1384/1530 train_loss:3.4816 train_time:237598ms step_avg:172.92ms
step:1385/1530 train_loss:3.4152 train_time:237772ms step_avg:172.93ms
step:1386/1530 train_loss:3.3450 train_time:237950ms step_avg:172.93ms
step:1387/1530 train_loss:3.2079 train_time:238128ms step_avg:172.93ms
step:1388/1530 train_loss:3.3519 train_time:238307ms step_avg:172.94ms
step:1389/1530 train_loss:3.3220 train_time:238490ms step_avg:172.94ms
step:1390/1530 train_loss:3.5752 train_time:238666ms step_avg:172.95ms
step:1391/1530 train_loss:3.2970 train_time:238843ms step_avg:172.95ms
step:1392/1530 train_loss:3.2949 train_time:239021ms step_avg:172.95ms
step:1393/1530 train_loss:3.2429 train_time:239202ms step_avg:172.96ms
step:1394/1530 train_loss:3.5045 train_time:239380ms step_avg:172.96ms
step:1395/1530 train_loss:3.4010 train_time:239557ms step_avg:172.97ms
step:1396/1530 train_loss:3.4089 train_time:239735ms step_avg:172.97ms
step:1397/1530 train_loss:3.3172 train_time:239911ms step_avg:172.97ms
step:1398/1530 train_loss:3.2618 train_time:240087ms step_avg:172.97ms
step:1399/1530 train_loss:3.3269 train_time:240264ms step_avg:172.98ms
step:1400/1530 train_loss:3.3279 train_time:240446ms step_avg:172.98ms
step:1401/1530 train_loss:3.3541 train_time:240622ms step_avg:172.98ms
step:1402/1530 train_loss:3.3075 train_time:240800ms step_avg:172.99ms
step:1403/1530 train_loss:3.4968 train_time:240984ms step_avg:173.00ms
step:1404/1530 train_loss:3.2865 train_time:241160ms step_avg:173.00ms
step:1405/1530 train_loss:3.3187 train_time:241341ms step_avg:173.00ms
step:1406/1530 train_loss:3.3179 train_time:241521ms step_avg:173.01ms
step:1407/1530 train_loss:3.1789 train_time:241698ms step_avg:173.01ms
step:1408/1530 train_loss:3.3214 train_time:241877ms step_avg:173.02ms
step:1409/1530 train_loss:3.3066 train_time:242064ms step_avg:173.03ms
step:1410/1530 train_loss:3.2991 train_time:242241ms step_avg:173.03ms
step:1411/1530 train_loss:3.3711 train_time:242416ms step_avg:173.03ms
step:1412/1530 train_loss:3.3366 train_time:242595ms step_avg:173.03ms
step:1413/1530 train_loss:3.3726 train_time:242772ms step_avg:173.04ms
step:1414/1530 train_loss:3.3368 train_time:242952ms step_avg:173.04ms
step:1415/1530 train_loss:3.4125 train_time:243137ms step_avg:173.05ms
step:1416/1530 train_loss:3.2372 train_time:243324ms step_avg:173.06ms
step:1417/1530 train_loss:3.2893 train_time:243508ms step_avg:173.07ms
step:1418/1530 train_loss:3.3987 train_time:243688ms step_avg:173.07ms
step:1419/1530 train_loss:3.3560 train_time:243871ms step_avg:173.08ms
step:1420/1530 train_loss:3.3770 train_time:244051ms step_avg:173.09ms
step:1421/1530 train_loss:3.3812 train_time:244229ms step_avg:173.09ms
step:1422/1530 train_loss:3.3393 train_time:244407ms step_avg:173.09ms
step:1423/1530 train_loss:3.3243 train_time:244585ms step_avg:173.10ms
step:1424/1530 train_loss:3.3373 train_time:244769ms step_avg:173.10ms
step:1425/1530 train_loss:3.1971 train_time:244957ms step_avg:173.11ms
step:1426/1530 train_loss:3.3285 train_time:245135ms step_avg:173.12ms
step:1427/1530 train_loss:3.2924 train_time:245318ms step_avg:173.13ms
step:1428/1530 train_loss:3.3819 train_time:245498ms step_avg:173.13ms
step:1429/1530 train_loss:3.3617 train_time:245677ms step_avg:173.13ms
step:1430/1530 train_loss:3.2650 train_time:245856ms step_avg:173.14ms
step:1431/1530 train_loss:3.3311 train_time:246039ms step_avg:173.15ms
step:1432/1530 train_loss:3.3415 train_time:246221ms step_avg:173.15ms
step:1433/1530 train_loss:3.1433 train_time:246405ms step_avg:173.16ms
step:1434/1530 train_loss:3.2979 train_time:246590ms step_avg:173.17ms
step:1435/1530 train_loss:3.1235 train_time:246769ms step_avg:173.17ms
step:1436/1530 train_loss:3.2379 train_time:246947ms step_avg:173.17ms
step:1437/1530 train_loss:3.4134 train_time:247123ms step_avg:173.18ms
step:1438/1530 train_loss:3.3883 train_time:247302ms step_avg:173.18ms
step:1439/1530 train_loss:3.3190 train_time:247483ms step_avg:173.19ms
step:1440/1530 train_loss:3.1976 train_time:247657ms step_avg:173.19ms
step:1441/1530 train_loss:3.3438 train_time:247837ms step_avg:173.19ms
step:1442/1530 train_loss:3.3864 train_time:248021ms step_avg:173.20ms
step:1443/1530 train_loss:3.4972 train_time:248207ms step_avg:173.21ms
step:1444/1530 train_loss:3.4501 train_time:248384ms step_avg:173.21ms
step:1445/1530 train_loss:3.3442 train_time:248561ms step_avg:173.21ms
step:1446/1530 train_loss:3.2068 train_time:248741ms step_avg:173.22ms
step:1447/1530 train_loss:3.3050 train_time:248922ms step_avg:173.22ms
step:1448/1530 train_loss:3.3043 train_time:249100ms step_avg:173.23ms
step:1449/1530 train_loss:3.4030 train_time:249278ms step_avg:173.23ms
step:1450/1530 train_loss:3.3965 train_time:249459ms step_avg:173.24ms
step:1451/1530 train_loss:3.2076 train_time:249637ms step_avg:173.24ms
step:1452/1530 train_loss:3.3307 train_time:249817ms step_avg:173.24ms
step:1453/1530 train_loss:3.2682 train_time:249993ms step_avg:173.25ms
step:1454/1530 train_loss:3.2952 train_time:250170ms step_avg:173.25ms
step:1455/1530 train_loss:3.3386 train_time:250352ms step_avg:173.25ms
step:1456/1530 train_loss:3.2930 train_time:250528ms step_avg:173.26ms
step:1457/1530 train_loss:3.1600 train_time:250706ms step_avg:173.26ms
step:1458/1530 train_loss:3.4321 train_time:250883ms step_avg:173.26ms
step:1459/1530 train_loss:3.2777 train_time:251065ms step_avg:173.27ms
step:1460/1530 train_loss:3.3215 train_time:251243ms step_avg:173.27ms
step:1461/1530 train_loss:3.4335 train_time:251422ms step_avg:173.28ms
step:1462/1530 train_loss:3.2672 train_time:251598ms step_avg:173.28ms
step:1463/1530 train_loss:3.4782 train_time:251782ms step_avg:173.28ms
step:1464/1530 train_loss:3.3661 train_time:251960ms step_avg:173.29ms
step:1465/1530 train_loss:3.3633 train_time:252140ms step_avg:173.29ms
step:1466/1530 train_loss:3.2931 train_time:252317ms step_avg:173.29ms
step:1467/1530 train_loss:3.4011 train_time:252498ms step_avg:173.30ms
step:1468/1530 train_loss:3.2957 train_time:252673ms step_avg:173.30ms
step:1469/1530 train_loss:3.2827 train_time:252851ms step_avg:173.30ms
step:1470/1530 train_loss:3.3362 train_time:253035ms step_avg:173.31ms
step:1471/1530 train_loss:3.2661 train_time:253220ms step_avg:173.32ms
step:1472/1530 train_loss:3.2567 train_time:253404ms step_avg:173.33ms
step:1473/1530 train_loss:3.4487 train_time:253582ms step_avg:173.33ms
step:1474/1530 train_loss:3.3193 train_time:253765ms step_avg:173.34ms
step:1475/1530 train_loss:3.1585 train_time:253950ms step_avg:173.34ms
step:1476/1530 train_loss:3.2730 train_time:254128ms step_avg:173.35ms
step:1477/1530 train_loss:3.2469 train_time:254317ms step_avg:173.36ms
step:1478/1530 train_loss:3.3153 train_time:254502ms step_avg:173.37ms
step:1479/1530 train_loss:3.4021 train_time:254684ms step_avg:173.37ms
step:1480/1530 train_loss:3.2757 train_time:254862ms step_avg:173.38ms
step:1481/1530 train_loss:3.4596 train_time:255042ms step_avg:173.38ms
step:1482/1530 train_loss:3.3741 train_time:255227ms step_avg:173.39ms
step:1483/1530 train_loss:3.2842 train_time:255416ms step_avg:173.40ms
step:1484/1530 train_loss:3.2735 train_time:255603ms step_avg:173.41ms
step:1485/1530 train_loss:3.2858 train_time:255783ms step_avg:173.41ms
step:1486/1530 train_loss:3.2355 train_time:255968ms step_avg:173.42ms
step:1487/1530 train_loss:3.3513 train_time:256150ms step_avg:173.43ms
step:1488/1530 train_loss:3.2552 train_time:256334ms step_avg:173.43ms
step:1489/1530 train_loss:3.3208 train_time:256515ms step_avg:173.44ms
step:1490/1530 train_loss:3.2572 train_time:256696ms step_avg:173.44ms
step:1491/1530 train_loss:3.1662 train_time:256877ms step_avg:173.45ms
step:1492/1530 train_loss:3.2794 train_time:257057ms step_avg:173.45ms
step:1493/1530 train_loss:3.4412 train_time:257237ms step_avg:173.46ms
step:1494/1530 train_loss:3.3054 train_time:257415ms step_avg:173.46ms
step:1495/1530 train_loss:3.0409 train_time:257601ms step_avg:173.47ms
step:1496/1530 train_loss:3.3678 train_time:257783ms step_avg:173.47ms
step:1497/1530 train_loss:3.3166 train_time:257966ms step_avg:173.48ms
step:1498/1530 train_loss:3.3523 train_time:258151ms step_avg:173.49ms
step:1499/1530 train_loss:3.3173 train_time:258339ms step_avg:173.50ms
step:1500/1530 train_loss:3.3035 train_time:258531ms step_avg:173.51ms
step:1500/1530 val_loss:3.2870 train_time:258586ms step_avg:173.55ms
step:1501/1530 train_loss:3.0961 train_time:258721ms step_avg:173.52ms
step:1502/1530 train_loss:3.3678 train_time:258914ms step_avg:173.53ms
step:1503/1530 train_loss:3.2512 train_time:259093ms step_avg:173.54ms
step:1504/1530 train_loss:3.2555 train_time:259274ms step_avg:173.54ms
step:1505/1530 train_loss:3.2202 train_time:259453ms step_avg:173.55ms
step:1506/1530 train_loss:3.2854 train_time:259636ms step_avg:173.55ms
step:1507/1530 train_loss:3.1807 train_time:259831ms step_avg:173.57ms
step:1508/1530 train_loss:3.4863 train_time:260014ms step_avg:173.57ms
step:1509/1530 train_loss:3.2867 train_time:260191ms step_avg:173.58ms
step:1510/1530 train_loss:3.2787 train_time:260372ms step_avg:173.58ms
step:1511/1530 train_loss:3.4186 train_time:260682ms step_avg:173.67ms
step:1512/1530 train_loss:3.4270 train_time:260870ms step_avg:173.68ms
step:1513/1530 train_loss:3.2779 train_time:261054ms step_avg:173.69ms
step:1514/1530 train_loss:3.0924 train_time:261237ms step_avg:173.69ms
step:1515/1530 train_loss:3.2500 train_time:261416ms step_avg:173.70ms
step:1516/1530 train_loss:3.2613 train_time:261601ms step_avg:173.71ms
step:1517/1530 train_loss:3.3065 train_time:261782ms step_avg:173.71ms
step:1518/1530 train_loss:3.2117 train_time:261966ms step_avg:173.72ms
step:1519/1530 train_loss:3.5165 train_time:262292ms step_avg:173.82ms
step:1520/1530 train_loss:3.1396 train_time:262474ms step_avg:173.82ms
step:1521/1530 train_loss:3.2130 train_time:262651ms step_avg:173.83ms
step:1522/1530 train_loss:3.3627 train_time:262836ms step_avg:173.83ms
step:1523/1530 train_loss:3.2349 train_time:263013ms step_avg:173.84ms
step:1524/1530 train_loss:3.3546 train_time:263193ms step_avg:173.84ms
step:1525/1530 train_loss:3.3450 train_time:263380ms step_avg:173.85ms
step:1526/1530 train_loss:3.2862 train_time:263570ms step_avg:173.86ms
step:1527/1530 train_loss:3.2953 train_time:263753ms step_avg:173.86ms
step:1528/1530 train_loss:3.4123 train_time:263932ms step_avg:173.87ms
step:1529/1530 train_loss:3.4156 train_time:264111ms step_avg:173.87ms
step:1530/1530 train_loss:3.2465 train_time:264290ms step_avg:173.88ms
step:1530/1530 val_loss:3.2845 train_time:264344ms step_avg:173.91ms