dca62101-15d4-4c76-842e-99213fa2508b.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
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)
                    if group['nesterov']:
                        g = g.add(buf, alpha=momentum)
                    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

class Rotary(torch.nn.Module):

    def __init__(self, dim, base=10000):
        super().__init__()
        self.dim = dim
        self.base = base
        self.inv_freq = None
        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:
            self.inv_freq = 1.0 / (self.base ** (torch.arange(0, self.dim, 2, device=x.device).float() / self.dim))
            self.seq_len_cached = seq_len
            t = torch.arange(seq_len, device=x.device).type_as(self.inv_freq)
            freqs = torch.outer(t, self.inv_freq)
            self.cos_cached = freqs.cos().bfloat16()
            self.sin_cached = freqs.sin().bfloat16()
        return self.cos_cached[None, :, None, :], self.sin_cached[None, :, None, :]

def apply_rotary_emb(x, cos, sin):
    assert x.ndim == 4 # multihead attention
    d = x.shape[3]//2
    x1 = x[..., :d]
    x2 = x[..., d:]
    y1 = x1 * cos + x2 * sin
    y2 = x1 * (-sin) + x2 * cos
    return torch.cat([y1, y2], 3).type_as(x)

class CastedLinear(nn.Linear):
    def forward(self, x):
        return F.linear(x, self.weight.to(x.dtype))

class CausalSelfAttention(nn.Module):

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

    def forward(self, x, v1, block_mask):
        B, T, C = x.size() # batch size, sequence length, embedding dimensionality (n_embd)
        q = self.c_q(x).view(B, T, self.n_head, self.head_dim)
        k = self.c_k(x).view(B, T, self.n_head, self.head_dim)
        v = self.c_v(x).view(B, T, self.n_head, self.head_dim)
        if v1 is None:
            v1 = v # This happens if we are in the first block. v needs to be accessed by subsequent blocks
        v = (1 - self.lamb) * v + self.lamb * v1.view_as(v) # @Grad62304977
        cos, sin = self.rotary(q)
        q, k = F.rms_norm(q, (q.size(-1),)), F.rms_norm(k, (k.size(-1),)) # QK norm suggested by @Grad62304977
        q, k = apply_rotary_emb(q, cos, sin), apply_rotary_emb(k, cos, sin)
        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, v1

class MLP(nn.Module):

    def __init__(self, config):
        super().__init__()
        self.c_fc    = CastedLinear(config.n_embd, 4 * config.n_embd, bias=False)
        self.c_proj  = CastedLinear(4 * config.n_embd, config.n_embd, bias=False)
        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)
        self.mlp = MLP(config)
        self.lambdas = nn.Parameter(torch.tensor([1., 0.]))

    def forward(self, x, v1, x0, block_mask):
        x = self.lambdas[0] * x + self.lambdas[1] * x0
        x1, v1 = self.attn(F.rms_norm(x, (x.size(-1),)), v1, block_mask)
        x = x + x1
        x = x + self.mlp(F.rms_norm(x, (x.size(-1),)))
        return x, v1

# -----------------------------------------------------------------------------
# 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),
            h = nn.ModuleList([Block(config) for _ in range(config.n_layer)]),
        ))
        self.lm_head = CastedLinear(config.n_embd, config.vocab_size, bias=False)
        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 = F.rms_norm(x, (x.size(-1),)) # @Grad62304977
        x0 = x
        v1 = None

        # 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, v1 = self.transformer.h[i](x, v1, 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, v1 = self.transformer.h[self.num_encoder_layers + i](x, v1, x0, block_mask)

        x = F.rms_norm(x, (x.size(-1),))
        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, B, T, process_rank, num_processes):
        self.process_rank = process_rank
        self.num_processes = num_processes
        self.B = B
        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 * B * 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.B * self.T
        self.tokens = _load_data_shard(self.files[self.current_shard])

    def next_batch(self):
        batch_size = self.B * self.T * self.num_processes
        buf = self.tokens[self.current_position:self.current_position+self.B*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
    device_batch_size : int = 1 # batch size, in sequences, per device
    sequence_length : int = 64*1024 # sequence length, in tokens
    num_iterations : int = 1750 # number of iterations to run
    warmup_iters : int = 0
    cooldown_iters : int = 640 # 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('='*100 + '\n')
        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
B, T = args.device_batch_size, args.sequence_length
# calculate the number of steps to take in the val loop.
assert args.val_tokens % (B * T * ddp_world_size) == 0
val_steps = args.val_tokens // (B * T * ddp_world_size)
# calculate the steps of gradient accumulation required to attain the desired global batch size.
assert args.batch_size % (B * ddp_world_size) == 0
train_accumulation_steps = args.batch_size // (B * ddp_world_size)

# load tokens
train_loader = DistributedDataLoader(args.input_bin, B, T, ddp_rank, ddp_world_size)
val_loader = DistributedDataLoader(args.input_val_bin, B, 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

# CUDNN attention is ~4ms faster than Flash, but doesn't get selected by default in PyTorch 2.5.1
from torch.backends.cuda import enable_cudnn_sdp, enable_flash_sdp, enable_math_sdp, enable_mem_efficient_sdp
enable_cudnn_sdp(True)
enable_flash_sdp(False)
enable_mem_efficient_sdp(False)
enable_math_sdp(False)

# init the optimizer(s)
optimizer1 = torch.optim.Adam([raw_model.transformer.wte.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)
    # Set the attention blocksize for the current step, in chunks of 64
    attn_blocksize = torch.tensor(64*((step/args.num_iterations * (1792 - 64) + 64)//64), dtype=torch.int, device='cuda')
    # 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

    # 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):
        # forward pass
        loss = model(x, y, attn_blocksize=attn_blocksize)
        train_loss = loss.detach()
        # advance the dataset for the next batch
        x, y = train_loader.next_batch()
        # backward pass
        if i < train_accumulation_steps:
            with model.no_sync(): # there's no need to sync gradients every accumulation step
                loss.backward()
        else:
            loss.backward() # just sync on the last step
    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.dev20241124+cu124 compiled for CUDA 12.4
nvidia-smi:
Mon Nov 25 00:11:20 2024       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 555.42.06              Driver Version: 555.42.06      CUDA Version: 12.5     |
|-----------------------------------------+------------------------+----------------------+
| 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          Off |   00000000:18:00.0 Off |                    0 |
| N/A   32C    P0             69W /  700W |       4MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA H100 80GB HBM3          Off |   00000000:2A:00.0 Off |                    0 |
| N/A   38C    P0            102W /  700W |      22MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   2  NVIDIA H100 80GB HBM3          Off |   00000000:3A:00.0 Off |                    0 |
| N/A   38C    P0             69W /  700W |       4MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   3  NVIDIA H100 80GB HBM3          Off |   00000000:5D:00.0 Off |                    0 |
| N/A   31C    P0             70W /  700W |       4MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   4  NVIDIA H100 80GB HBM3          Off |   00000000:84:00.0 Off |                    0 |
| N/A   31C    P0             71W /  700W |       4MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   5  NVIDIA H100 80GB HBM3          Off |   00000000:8B:00.0 Off |                    0 |
| N/A   37C    P0             71W /  700W |       4MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   6  NVIDIA H100 80GB HBM3          Off |   00000000:91:00.0 Off |                    0 |
| N/A   35C    P0            114W /  700W |      23MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
|   7  NVIDIA H100 80GB HBM3          Off |   00000000:E4:00.0 Off |                    0 |
| N/A   31C    P0             72W /  700W |       4MiB /  81559MiB |      0%      Default |
|                                         |                        |             Disabled |
+-----------------------------------------+------------------------+----------------------+
                                                                                         
+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    1   N/A  N/A     34414      C   /usr/bin/python3                                0MiB |
|    6   N/A  N/A     34419      C   /usr/bin/python3                                0MiB |
+-----------------------------------------------------------------------------------------+

====================================================================================================
Training DataLoader: total number of tokens: 1800000000 across 18 files
Validation DataLoader: total number of tokens: 100000000 across 1 files
====================================================================================================
step:0/1750 val_loss:10.8258 train_time:0ms step_avg:nanms
step:1/1750 train_loss:10.8258 train_time:29542ms step_avg:nanms
step:2/1750 train_loss:10.0684 train_time:29653ms step_avg:nanms
step:3/1750 train_loss:8.3706 train_time:29797ms step_avg:nanms
step:4/1750 train_loss:7.6195 train_time:29945ms step_avg:nanms
step:5/1750 train_loss:7.4804 train_time:30093ms step_avg:nanms
step:6/1750 train_loss:6.9973 train_time:30240ms step_avg:nanms
step:7/1750 train_loss:7.2450 train_time:30387ms step_avg:nanms
step:8/1750 train_loss:6.7453 train_time:30535ms step_avg:nanms
step:9/1750 train_loss:6.6253 train_time:30681ms step_avg:nanms
step:10/1750 train_loss:6.5351 train_time:30830ms step_avg:nanms
step:11/1750 train_loss:6.5154 train_time:110ms step_avg:nanms
step:12/1750 train_loss:6.3730 train_time:257ms step_avg:nanms
step:13/1750 train_loss:6.2362 train_time:406ms step_avg:135.27ms
step:14/1750 train_loss:6.2109 train_time:552ms step_avg:138.06ms
step:15/1750 train_loss:6.1753 train_time:700ms step_avg:140.10ms
step:16/1750 train_loss:6.1661 train_time:848ms step_avg:141.31ms
step:17/1750 train_loss:6.2227 train_time:996ms step_avg:142.31ms
step:18/1750 train_loss:6.0256 train_time:1144ms step_avg:142.99ms
step:19/1750 train_loss:6.0295 train_time:1291ms step_avg:143.44ms
step:20/1750 train_loss:5.7402 train_time:1438ms step_avg:143.83ms
step:21/1750 train_loss:6.0167 train_time:1586ms step_avg:144.20ms
step:22/1750 train_loss:6.2523 train_time:1733ms step_avg:144.38ms
step:23/1750 train_loss:5.9193 train_time:1881ms step_avg:144.70ms
step:24/1750 train_loss:6.1039 train_time:2029ms step_avg:144.92ms
step:25/1750 train_loss:5.7746 train_time:2176ms step_avg:145.09ms
step:26/1750 train_loss:5.6894 train_time:2327ms step_avg:145.42ms
step:27/1750 train_loss:5.8944 train_time:2474ms step_avg:145.54ms
step:28/1750 train_loss:5.4991 train_time:2622ms step_avg:145.65ms
step:29/1750 train_loss:5.7636 train_time:2769ms step_avg:145.72ms
step:30/1750 train_loss:5.5876 train_time:2916ms step_avg:145.80ms
step:31/1750 train_loss:5.5462 train_time:3064ms step_avg:145.89ms
step:32/1750 train_loss:5.4139 train_time:3211ms step_avg:145.94ms
step:33/1750 train_loss:5.7045 train_time:3358ms step_avg:146.01ms
step:34/1750 train_loss:5.6040 train_time:3506ms step_avg:146.08ms
step:35/1750 train_loss:5.6913 train_time:3652ms step_avg:146.08ms
step:36/1750 train_loss:5.6551 train_time:3801ms step_avg:146.17ms
step:37/1750 train_loss:5.5731 train_time:3948ms step_avg:146.21ms
step:38/1750 train_loss:5.4090 train_time:4096ms step_avg:146.28ms
step:39/1750 train_loss:5.4200 train_time:4243ms step_avg:146.33ms
step:40/1750 train_loss:5.3542 train_time:4391ms step_avg:146.35ms
step:41/1750 train_loss:5.3240 train_time:4538ms step_avg:146.38ms
step:42/1750 train_loss:5.2552 train_time:4686ms step_avg:146.43ms
step:43/1750 train_loss:5.3456 train_time:4832ms step_avg:146.44ms
step:44/1750 train_loss:5.3128 train_time:4981ms step_avg:146.49ms
step:45/1750 train_loss:5.4665 train_time:5129ms step_avg:146.54ms
step:46/1750 train_loss:5.2496 train_time:5276ms step_avg:146.55ms
step:47/1750 train_loss:5.1535 train_time:5424ms step_avg:146.58ms
step:48/1750 train_loss:5.3197 train_time:5570ms step_avg:146.58ms
step:49/1750 train_loss:5.2384 train_time:5718ms step_avg:146.61ms
step:50/1750 train_loss:5.3184 train_time:5866ms step_avg:146.65ms
step:51/1750 train_loss:5.2112 train_time:6013ms step_avg:146.66ms
step:52/1750 train_loss:5.1192 train_time:6162ms step_avg:146.70ms
step:53/1750 train_loss:5.2431 train_time:6309ms step_avg:146.71ms
step:54/1750 train_loss:5.1014 train_time:6456ms step_avg:146.72ms
step:55/1750 train_loss:5.4870 train_time:6605ms step_avg:146.77ms
step:56/1750 train_loss:5.1101 train_time:6750ms step_avg:146.75ms
step:57/1750 train_loss:4.9616 train_time:6899ms step_avg:146.79ms
step:58/1750 train_loss:5.0903 train_time:7046ms step_avg:146.80ms
step:59/1750 train_loss:5.0937 train_time:7193ms step_avg:146.81ms
step:60/1750 train_loss:5.1986 train_time:7342ms step_avg:146.83ms
step:61/1750 train_loss:4.9161 train_time:7488ms step_avg:146.83ms
step:62/1750 train_loss:5.0517 train_time:7636ms step_avg:146.84ms
step:63/1750 train_loss:5.0449 train_time:7784ms step_avg:146.88ms
step:64/1750 train_loss:4.9396 train_time:7931ms step_avg:146.88ms
step:65/1750 train_loss:4.9000 train_time:8079ms step_avg:146.89ms
step:66/1750 train_loss:5.0698 train_time:8227ms step_avg:146.91ms
step:67/1750 train_loss:4.9068 train_time:8374ms step_avg:146.92ms
step:68/1750 train_loss:5.2049 train_time:8522ms step_avg:146.93ms
step:69/1750 train_loss:4.8059 train_time:8669ms step_avg:146.93ms
step:70/1750 train_loss:4.9253 train_time:8817ms step_avg:146.94ms
step:71/1750 train_loss:5.0507 train_time:8965ms step_avg:146.96ms
step:72/1750 train_loss:4.9639 train_time:9111ms step_avg:146.95ms
step:73/1750 train_loss:4.8449 train_time:9259ms step_avg:146.97ms
step:74/1750 train_loss:4.9830 train_time:9407ms step_avg:146.98ms
step:75/1750 train_loss:4.9363 train_time:9554ms step_avg:146.99ms
step:76/1750 train_loss:4.8639 train_time:9703ms step_avg:147.02ms
step:77/1750 train_loss:4.9869 train_time:9849ms step_avg:147.00ms
step:78/1750 train_loss:5.1764 train_time:9997ms step_avg:147.02ms
step:79/1750 train_loss:4.8735 train_time:10145ms step_avg:147.02ms
step:80/1750 train_loss:4.9202 train_time:10292ms step_avg:147.03ms
step:81/1750 train_loss:4.7066 train_time:10439ms step_avg:147.03ms
step:82/1750 train_loss:4.8751 train_time:10587ms step_avg:147.04ms
step:83/1750 train_loss:4.8395 train_time:10734ms step_avg:147.04ms
step:84/1750 train_loss:4.8301 train_time:10882ms step_avg:147.05ms
step:85/1750 train_loss:4.6826 train_time:11029ms step_avg:147.05ms
step:86/1750 train_loss:4.9128 train_time:11176ms step_avg:147.06ms
step:87/1750 train_loss:4.7982 train_time:11324ms step_avg:147.07ms
step:88/1750 train_loss:4.8143 train_time:11471ms step_avg:147.07ms
step:89/1750 train_loss:4.7354 train_time:11618ms step_avg:147.07ms
step:90/1750 train_loss:4.6997 train_time:11767ms step_avg:147.09ms
step:91/1750 train_loss:4.6700 train_time:11913ms step_avg:147.07ms
step:92/1750 train_loss:4.8268 train_time:12061ms step_avg:147.08ms
step:93/1750 train_loss:4.6505 train_time:12208ms step_avg:147.09ms
step:94/1750 train_loss:4.6797 train_time:12356ms step_avg:147.09ms
step:95/1750 train_loss:4.7045 train_time:12504ms step_avg:147.10ms
step:96/1750 train_loss:4.6314 train_time:12650ms step_avg:147.09ms
step:97/1750 train_loss:4.6847 train_time:12798ms step_avg:147.10ms
step:98/1750 train_loss:4.6169 train_time:12945ms step_avg:147.10ms
step:99/1750 train_loss:4.6971 train_time:13092ms step_avg:147.10ms
step:100/1750 train_loss:4.7039 train_time:13238ms step_avg:147.09ms
step:101/1750 train_loss:4.5537 train_time:13387ms step_avg:147.11ms
step:102/1750 train_loss:4.7359 train_time:13534ms step_avg:147.11ms
step:103/1750 train_loss:4.6011 train_time:13683ms step_avg:147.13ms
step:104/1750 train_loss:4.5406 train_time:13830ms step_avg:147.12ms
step:105/1750 train_loss:4.5885 train_time:13978ms step_avg:147.13ms
step:106/1750 train_loss:4.6450 train_time:14125ms step_avg:147.13ms
step:107/1750 train_loss:4.5390 train_time:14271ms step_avg:147.13ms
step:108/1750 train_loss:4.3957 train_time:14418ms step_avg:147.13ms
step:109/1750 train_loss:4.5137 train_time:14567ms step_avg:147.14ms
step:110/1750 train_loss:4.5070 train_time:14713ms step_avg:147.13ms
step:111/1750 train_loss:4.4628 train_time:14861ms step_avg:147.14ms
step:112/1750 train_loss:4.5999 train_time:15008ms step_avg:147.14ms
step:113/1750 train_loss:4.5055 train_time:15156ms step_avg:147.14ms
step:114/1750 train_loss:4.3878 train_time:15304ms step_avg:147.15ms
step:115/1750 train_loss:4.5353 train_time:15451ms step_avg:147.15ms
step:116/1750 train_loss:4.5074 train_time:15599ms step_avg:147.16ms
step:117/1750 train_loss:4.4020 train_time:15745ms step_avg:147.15ms
step:118/1750 train_loss:4.6398 train_time:15893ms step_avg:147.16ms
step:119/1750 train_loss:4.4919 train_time:16040ms step_avg:147.16ms
step:120/1750 train_loss:4.3905 train_time:16187ms step_avg:147.16ms
step:121/1750 train_loss:4.3340 train_time:16334ms step_avg:147.15ms
step:122/1750 train_loss:4.4865 train_time:16482ms step_avg:147.16ms
step:123/1750 train_loss:4.3264 train_time:16629ms step_avg:147.16ms
step:124/1750 train_loss:4.6324 train_time:16776ms step_avg:147.16ms
step:125/1750 train_loss:4.4947 train_time:16923ms step_avg:147.16ms
step:125/1750 val_loss:4.4390 train_time:16961ms step_avg:147.49ms
step:126/1750 train_loss:4.4504 train_time:17072ms step_avg:147.17ms
step:127/1750 train_loss:4.4893 train_time:17220ms step_avg:147.18ms
step:128/1750 train_loss:4.4121 train_time:17367ms step_avg:147.18ms
step:129/1750 train_loss:4.7284 train_time:17514ms step_avg:147.18ms
step:130/1750 train_loss:4.4119 train_time:17662ms step_avg:147.18ms
step:131/1750 train_loss:4.4322 train_time:17811ms step_avg:147.20ms
step:132/1750 train_loss:4.3737 train_time:17963ms step_avg:147.24ms
step:133/1750 train_loss:4.4727 train_time:18114ms step_avg:147.26ms
step:134/1750 train_loss:4.2776 train_time:18266ms step_avg:147.30ms
step:135/1750 train_loss:4.4612 train_time:18418ms step_avg:147.34ms
step:136/1750 train_loss:4.2324 train_time:18569ms step_avg:147.37ms
step:137/1750 train_loss:4.3928 train_time:18720ms step_avg:147.40ms
step:138/1750 train_loss:4.2974 train_time:18870ms step_avg:147.42ms
step:139/1750 train_loss:4.3957 train_time:19021ms step_avg:147.45ms
step:140/1750 train_loss:4.4818 train_time:19171ms step_avg:147.47ms
step:141/1750 train_loss:4.3193 train_time:19321ms step_avg:147.49ms
step:142/1750 train_loss:4.3226 train_time:19471ms step_avg:147.51ms
step:143/1750 train_loss:4.2781 train_time:19621ms step_avg:147.53ms
step:144/1750 train_loss:4.3665 train_time:19771ms step_avg:147.54ms
step:145/1750 train_loss:4.3223 train_time:19921ms step_avg:147.56ms
step:146/1750 train_loss:4.1790 train_time:20071ms step_avg:147.58ms
step:147/1750 train_loss:4.3394 train_time:20222ms step_avg:147.61ms
step:148/1750 train_loss:4.3627 train_time:20372ms step_avg:147.62ms
step:149/1750 train_loss:4.3105 train_time:20522ms step_avg:147.64ms
step:150/1750 train_loss:4.4559 train_time:20672ms step_avg:147.66ms
step:151/1750 train_loss:4.2800 train_time:20822ms step_avg:147.68ms
step:152/1750 train_loss:4.2858 train_time:20972ms step_avg:147.69ms
step:153/1750 train_loss:4.3804 train_time:21123ms step_avg:147.71ms
step:154/1750 train_loss:4.3760 train_time:21273ms step_avg:147.73ms
step:155/1750 train_loss:4.2905 train_time:21424ms step_avg:147.75ms
step:156/1750 train_loss:4.3549 train_time:21574ms step_avg:147.77ms
step:157/1750 train_loss:4.4245 train_time:21725ms step_avg:147.79ms
step:158/1750 train_loss:4.2708 train_time:21876ms step_avg:147.81ms
step:159/1750 train_loss:4.3258 train_time:22027ms step_avg:147.83ms
step:160/1750 train_loss:4.1274 train_time:22177ms step_avg:147.84ms
step:161/1750 train_loss:4.3630 train_time:22327ms step_avg:147.86ms
step:162/1750 train_loss:4.3634 train_time:22478ms step_avg:147.88ms
step:163/1750 train_loss:4.3536 train_time:22627ms step_avg:147.89ms
step:164/1750 train_loss:4.2012 train_time:22778ms step_avg:147.91ms
step:165/1750 train_loss:4.3057 train_time:22928ms step_avg:147.92ms
step:166/1750 train_loss:4.3738 train_time:23078ms step_avg:147.94ms
step:167/1750 train_loss:4.2172 train_time:23228ms step_avg:147.95ms
step:168/1750 train_loss:4.2976 train_time:23378ms step_avg:147.96ms
step:169/1750 train_loss:4.1670 train_time:23529ms step_avg:147.98ms
step:170/1750 train_loss:4.0416 train_time:23680ms step_avg:148.00ms
step:171/1750 train_loss:4.2132 train_time:23829ms step_avg:148.01ms
step:172/1750 train_loss:4.2278 train_time:23980ms step_avg:148.02ms
step:173/1750 train_loss:4.2850 train_time:24130ms step_avg:148.04ms
step:174/1750 train_loss:4.4364 train_time:24281ms step_avg:148.05ms
step:175/1750 train_loss:4.2676 train_time:24431ms step_avg:148.06ms
step:176/1750 train_loss:4.1136 train_time:24580ms step_avg:148.07ms
step:177/1750 train_loss:4.0858 train_time:24731ms step_avg:148.09ms
step:178/1750 train_loss:4.2087 train_time:24881ms step_avg:148.10ms
step:179/1750 train_loss:4.1514 train_time:25031ms step_avg:148.11ms
step:180/1750 train_loss:4.1321 train_time:25181ms step_avg:148.12ms
step:181/1750 train_loss:4.3230 train_time:25331ms step_avg:148.13ms
step:182/1750 train_loss:4.1881 train_time:25481ms step_avg:148.15ms
step:183/1750 train_loss:4.1494 train_time:25631ms step_avg:148.16ms
step:184/1750 train_loss:4.1544 train_time:25781ms step_avg:148.17ms
step:185/1750 train_loss:4.2340 train_time:25931ms step_avg:148.18ms
step:186/1750 train_loss:4.1997 train_time:26081ms step_avg:148.19ms
step:187/1750 train_loss:4.2509 train_time:26231ms step_avg:148.20ms
step:188/1750 train_loss:4.1955 train_time:26504ms step_avg:148.90ms
step:189/1750 train_loss:4.1414 train_time:26805ms step_avg:149.75ms
step:190/1750 train_loss:4.2368 train_time:26956ms step_avg:149.76ms
step:191/1750 train_loss:4.1065 train_time:27106ms step_avg:149.76ms
step:192/1750 train_loss:4.0548 train_time:27258ms step_avg:149.77ms
step:193/1750 train_loss:4.2763 train_time:27408ms step_avg:149.77ms
step:194/1750 train_loss:4.1885 train_time:27559ms step_avg:149.78ms
step:195/1750 train_loss:4.3725 train_time:27709ms step_avg:149.78ms
step:196/1750 train_loss:4.2025 train_time:27859ms step_avg:149.78ms
step:197/1750 train_loss:4.0726 train_time:28009ms step_avg:149.78ms
step:198/1750 train_loss:4.1861 train_time:28158ms step_avg:149.78ms
step:199/1750 train_loss:4.0388 train_time:28307ms step_avg:149.77ms
step:200/1750 train_loss:4.1322 train_time:28457ms step_avg:149.77ms
step:201/1750 train_loss:4.0231 train_time:28605ms step_avg:149.77ms
step:202/1750 train_loss:4.2611 train_time:28755ms step_avg:149.77ms
step:203/1750 train_loss:4.0731 train_time:28905ms step_avg:149.77ms
step:204/1750 train_loss:4.2092 train_time:29054ms step_avg:149.76ms
step:205/1750 train_loss:4.2609 train_time:29204ms step_avg:149.76ms
step:206/1750 train_loss:3.9490 train_time:29353ms step_avg:149.76ms
step:207/1750 train_loss:4.0959 train_time:29505ms step_avg:149.77ms
step:208/1750 train_loss:4.1077 train_time:29653ms step_avg:149.76ms
step:209/1750 train_loss:4.2502 train_time:29804ms step_avg:149.77ms
step:210/1750 train_loss:4.2013 train_time:29953ms step_avg:149.76ms
step:211/1750 train_loss:4.0603 train_time:30102ms step_avg:149.76ms
step:212/1750 train_loss:4.1282 train_time:30251ms step_avg:149.76ms
step:213/1750 train_loss:4.0549 train_time:30401ms step_avg:149.76ms
step:214/1750 train_loss:4.1243 train_time:30549ms step_avg:149.75ms
step:215/1750 train_loss:3.9645 train_time:30700ms step_avg:149.76ms
step:216/1750 train_loss:4.0108 train_time:30849ms step_avg:149.75ms
step:217/1750 train_loss:4.0207 train_time:31000ms step_avg:149.76ms
step:218/1750 train_loss:4.0934 train_time:31149ms step_avg:149.75ms
step:219/1750 train_loss:4.0818 train_time:31299ms step_avg:149.76ms
step:220/1750 train_loss:4.1000 train_time:31447ms step_avg:149.75ms
step:221/1750 train_loss:4.1067 train_time:31597ms step_avg:149.75ms
step:222/1750 train_loss:4.0046 train_time:31746ms step_avg:149.74ms
step:223/1750 train_loss:4.0005 train_time:31894ms step_avg:149.74ms
step:224/1750 train_loss:4.3115 train_time:32044ms step_avg:149.74ms
step:225/1750 train_loss:3.9130 train_time:32194ms step_avg:149.74ms
step:226/1750 train_loss:3.9981 train_time:32351ms step_avg:149.77ms
step:227/1750 train_loss:3.9931 train_time:32493ms step_avg:149.74ms
step:228/1750 train_loss:4.1522 train_time:32643ms step_avg:149.74ms
step:229/1750 train_loss:3.9398 train_time:32793ms step_avg:149.74ms
step:230/1750 train_loss:4.0671 train_time:32942ms step_avg:149.74ms
step:231/1750 train_loss:3.9226 train_time:33091ms step_avg:149.73ms
step:232/1750 train_loss:3.9843 train_time:33242ms step_avg:149.74ms
step:233/1750 train_loss:4.1037 train_time:33389ms step_avg:149.73ms
step:234/1750 train_loss:4.0538 train_time:33540ms step_avg:149.73ms
step:235/1750 train_loss:3.9287 train_time:33689ms step_avg:149.73ms
step:236/1750 train_loss:4.1042 train_time:33839ms step_avg:149.73ms
step:237/1750 train_loss:4.0910 train_time:33987ms step_avg:149.72ms
step:238/1750 train_loss:3.9585 train_time:34137ms step_avg:149.72ms
step:239/1750 train_loss:4.1032 train_time:34286ms step_avg:149.72ms
step:240/1750 train_loss:4.1335 train_time:34437ms step_avg:149.72ms
step:241/1750 train_loss:3.9906 train_time:34585ms step_avg:149.72ms
step:242/1750 train_loss:4.1584 train_time:34735ms step_avg:149.72ms
step:243/1750 train_loss:4.0334 train_time:34883ms step_avg:149.71ms
step:244/1750 train_loss:4.0982 train_time:35033ms step_avg:149.72ms
step:245/1750 train_loss:4.1620 train_time:35182ms step_avg:149.71ms
step:246/1750 train_loss:4.0828 train_time:35332ms step_avg:149.71ms
step:247/1750 train_loss:4.0238 train_time:35481ms step_avg:149.71ms
step:248/1750 train_loss:4.1345 train_time:35630ms step_avg:149.71ms
step:249/1750 train_loss:3.9336 train_time:35780ms step_avg:149.71ms
step:250/1750 train_loss:3.9939 train_time:35929ms step_avg:149.70ms
step:250/1750 val_loss:4.0286 train_time:35967ms step_avg:149.86ms
step:251/1750 train_loss:4.0993 train_time:36080ms step_avg:149.71ms
step:252/1750 train_loss:4.1796 train_time:36233ms step_avg:149.72ms
step:253/1750 train_loss:3.9579 train_time:36382ms step_avg:149.72ms
step:254/1750 train_loss:3.9013 train_time:36530ms step_avg:149.71ms
step:255/1750 train_loss:4.0930 train_time:36679ms step_avg:149.71ms
step:256/1750 train_loss:4.0144 train_time:36828ms step_avg:149.71ms
step:257/1750 train_loss:4.0145 train_time:36978ms step_avg:149.71ms
step:258/1750 train_loss:4.0026 train_time:37127ms step_avg:149.71ms
step:259/1750 train_loss:4.0450 train_time:37278ms step_avg:149.71ms
step:260/1750 train_loss:4.0767 train_time:37427ms step_avg:149.71ms
step:261/1750 train_loss:4.0402 train_time:37579ms step_avg:149.72ms
step:262/1750 train_loss:4.0027 train_time:37731ms step_avg:149.73ms
step:263/1750 train_loss:3.9046 train_time:37885ms step_avg:149.74ms
step:264/1750 train_loss:4.0031 train_time:38038ms step_avg:149.76ms
step:265/1750 train_loss:3.8831 train_time:38191ms step_avg:149.77ms
step:266/1750 train_loss:3.9317 train_time:38344ms step_avg:149.78ms
step:267/1750 train_loss:3.9416 train_time:38496ms step_avg:149.79ms
step:268/1750 train_loss:3.9794 train_time:38648ms step_avg:149.80ms
step:269/1750 train_loss:3.8714 train_time:38801ms step_avg:149.81ms
step:270/1750 train_loss:4.1141 train_time:38953ms step_avg:149.82ms
step:271/1750 train_loss:3.9862 train_time:39105ms step_avg:149.83ms
step:272/1750 train_loss:3.9308 train_time:39258ms step_avg:149.84ms
step:273/1750 train_loss:3.9533 train_time:39410ms step_avg:149.85ms
step:274/1750 train_loss:4.0472 train_time:39565ms step_avg:149.87ms
step:275/1750 train_loss:4.0701 train_time:39717ms step_avg:149.88ms
step:276/1750 train_loss:4.2323 train_time:39870ms step_avg:149.89ms
step:277/1750 train_loss:4.0509 train_time:40023ms step_avg:149.90ms
step:278/1750 train_loss:4.1100 train_time:40175ms step_avg:149.91ms
step:279/1750 train_loss:4.0144 train_time:40327ms step_avg:149.92ms
step:280/1750 train_loss:4.2045 train_time:40482ms step_avg:149.93ms
step:281/1750 train_loss:3.9800 train_time:40634ms step_avg:149.94ms
step:282/1750 train_loss:3.9628 train_time:40787ms step_avg:149.95ms
step:283/1750 train_loss:3.9303 train_time:40938ms step_avg:149.96ms
step:284/1750 train_loss:4.0713 train_time:41093ms step_avg:149.97ms
step:285/1750 train_loss:4.0814 train_time:41246ms step_avg:149.98ms
step:286/1750 train_loss:4.1092 train_time:41398ms step_avg:149.99ms
step:287/1750 train_loss:3.9241 train_time:41552ms step_avg:150.01ms
step:288/1750 train_loss:4.0256 train_time:41704ms step_avg:150.01ms
step:289/1750 train_loss:3.9011 train_time:41858ms step_avg:150.03ms
step:290/1750 train_loss:3.8727 train_time:42009ms step_avg:150.03ms
step:291/1750 train_loss:3.9233 train_time:42162ms step_avg:150.04ms
step:292/1750 train_loss:3.8782 train_time:42314ms step_avg:150.05ms
step:293/1750 train_loss:3.9263 train_time:42467ms step_avg:150.06ms
step:294/1750 train_loss:3.9564 train_time:42620ms step_avg:150.07ms
step:295/1750 train_loss:3.8526 train_time:42771ms step_avg:150.07ms
step:296/1750 train_loss:3.8774 train_time:42925ms step_avg:150.09ms
step:297/1750 train_loss:3.8909 train_time:43078ms step_avg:150.10ms
step:298/1750 train_loss:3.9968 train_time:43230ms step_avg:150.10ms
step:299/1750 train_loss:3.8436 train_time:43383ms step_avg:150.11ms
step:300/1750 train_loss:3.9825 train_time:43535ms step_avg:150.12ms
step:301/1750 train_loss:3.9892 train_time:43688ms step_avg:150.13ms
step:302/1750 train_loss:3.9452 train_time:43842ms step_avg:150.14ms
step:303/1750 train_loss:3.9983 train_time:43994ms step_avg:150.15ms
step:304/1750 train_loss:3.9806 train_time:44146ms step_avg:150.16ms
step:305/1750 train_loss:4.4694 train_time:44298ms step_avg:150.16ms
step:306/1750 train_loss:3.9538 train_time:44450ms step_avg:150.17ms
step:307/1750 train_loss:3.8509 train_time:44603ms step_avg:150.18ms
step:308/1750 train_loss:4.0087 train_time:44755ms step_avg:150.18ms
step:309/1750 train_loss:3.8790 train_time:44907ms step_avg:150.19ms
step:310/1750 train_loss:4.0981 train_time:45060ms step_avg:150.20ms
step:311/1750 train_loss:3.9442 train_time:45212ms step_avg:150.21ms
step:312/1750 train_loss:3.8865 train_time:45365ms step_avg:150.21ms
step:313/1750 train_loss:3.9652 train_time:45518ms step_avg:150.22ms
step:314/1750 train_loss:4.0852 train_time:45670ms step_avg:150.23ms
step:315/1750 train_loss:3.9690 train_time:45822ms step_avg:150.24ms
step:316/1750 train_loss:3.8092 train_time:45973ms step_avg:150.24ms
step:317/1750 train_loss:3.8903 train_time:46127ms step_avg:150.25ms
step:318/1750 train_loss:3.9397 train_time:46279ms step_avg:150.26ms
step:319/1750 train_loss:3.9115 train_time:46431ms step_avg:150.26ms
step:320/1750 train_loss:4.0325 train_time:46584ms step_avg:150.27ms
step:321/1750 train_loss:3.9803 train_time:46737ms step_avg:150.28ms
step:322/1750 train_loss:3.9435 train_time:46890ms step_avg:150.29ms
step:323/1750 train_loss:4.0233 train_time:47044ms step_avg:150.30ms
step:324/1750 train_loss:3.9738 train_time:47196ms step_avg:150.31ms
step:325/1750 train_loss:4.0326 train_time:47349ms step_avg:150.31ms
step:326/1750 train_loss:3.9090 train_time:47501ms step_avg:150.32ms
step:327/1750 train_loss:4.4132 train_time:47651ms step_avg:150.32ms
step:328/1750 train_loss:4.0915 train_time:47803ms step_avg:150.32ms
step:329/1750 train_loss:3.8215 train_time:47955ms step_avg:150.33ms
step:330/1750 train_loss:3.7594 train_time:48107ms step_avg:150.33ms
step:331/1750 train_loss:3.9935 train_time:48261ms step_avg:150.35ms
step:332/1750 train_loss:3.9249 train_time:48411ms step_avg:150.35ms
step:333/1750 train_loss:3.8980 train_time:48564ms step_avg:150.35ms
step:334/1750 train_loss:3.8645 train_time:48715ms step_avg:150.36ms
step:335/1750 train_loss:4.0315 train_time:48868ms step_avg:150.36ms
step:336/1750 train_loss:3.9760 train_time:49021ms step_avg:150.37ms
step:337/1750 train_loss:4.4388 train_time:49174ms step_avg:150.38ms
step:338/1750 train_loss:3.9523 train_time:49324ms step_avg:150.38ms
step:339/1750 train_loss:3.8808 train_time:49475ms step_avg:150.38ms
step:340/1750 train_loss:3.9487 train_time:49627ms step_avg:150.39ms
step:341/1750 train_loss:3.8735 train_time:49780ms step_avg:150.39ms
step:342/1750 train_loss:3.8370 train_time:49931ms step_avg:150.39ms
step:343/1750 train_loss:3.8545 train_time:50084ms step_avg:150.40ms
step:344/1750 train_loss:4.0075 train_time:50235ms step_avg:150.40ms
step:345/1750 train_loss:3.8317 train_time:50387ms step_avg:150.41ms
step:346/1750 train_loss:3.7825 train_time:50539ms step_avg:150.42ms
step:347/1750 train_loss:3.8196 train_time:50692ms step_avg:150.42ms
step:348/1750 train_loss:3.8767 train_time:50844ms step_avg:150.43ms
step:349/1750 train_loss:3.8508 train_time:50995ms step_avg:150.43ms
step:350/1750 train_loss:3.5820 train_time:51147ms step_avg:150.43ms
step:351/1750 train_loss:3.8431 train_time:51299ms step_avg:150.44ms
step:352/1750 train_loss:4.2056 train_time:51451ms step_avg:150.44ms
step:353/1750 train_loss:3.6813 train_time:51602ms step_avg:150.44ms
step:354/1750 train_loss:3.9403 train_time:51753ms step_avg:150.45ms
step:355/1750 train_loss:3.8021 train_time:51904ms step_avg:150.45ms
step:356/1750 train_loss:3.8965 train_time:52057ms step_avg:150.45ms
step:357/1750 train_loss:3.7803 train_time:52209ms step_avg:150.46ms
step:358/1750 train_loss:3.8715 train_time:52361ms step_avg:150.46ms
step:359/1750 train_loss:3.8113 train_time:52513ms step_avg:150.47ms
step:360/1750 train_loss:3.4454 train_time:52665ms step_avg:150.47ms
step:361/1750 train_loss:4.0502 train_time:52816ms step_avg:150.47ms
step:362/1750 train_loss:3.9392 train_time:52968ms step_avg:150.48ms
step:363/1750 train_loss:3.8621 train_time:53120ms step_avg:150.48ms
step:364/1750 train_loss:3.7617 train_time:53271ms step_avg:150.48ms
step:365/1750 train_loss:3.9274 train_time:53423ms step_avg:150.49ms
step:366/1750 train_loss:3.8850 train_time:53574ms step_avg:150.49ms
step:367/1750 train_loss:3.8765 train_time:53726ms step_avg:150.49ms
step:368/1750 train_loss:3.8600 train_time:53878ms step_avg:150.50ms
step:369/1750 train_loss:3.7610 train_time:54030ms step_avg:150.50ms
step:370/1750 train_loss:3.9102 train_time:54182ms step_avg:150.51ms
step:371/1750 train_loss:3.7579 train_time:54334ms step_avg:150.51ms
step:372/1750 train_loss:3.7095 train_time:54485ms step_avg:150.51ms
step:373/1750 train_loss:3.9375 train_time:54637ms step_avg:150.52ms
step:374/1750 train_loss:3.8442 train_time:54790ms step_avg:150.52ms
step:375/1750 train_loss:3.8164 train_time:54942ms step_avg:150.53ms
step:375/1750 val_loss:3.8441 train_time:54981ms step_avg:150.63ms
step:376/1750 train_loss:3.8838 train_time:55095ms step_avg:150.53ms
step:377/1750 train_loss:3.8150 train_time:55375ms step_avg:150.89ms
step:378/1750 train_loss:3.8646 train_time:55536ms step_avg:150.91ms
step:379/1750 train_loss:3.9030 train_time:55835ms step_avg:151.31ms
step:380/1750 train_loss:3.9681 train_time:55986ms step_avg:151.31ms
step:381/1750 train_loss:3.8639 train_time:56139ms step_avg:151.32ms
step:382/1750 train_loss:3.8296 train_time:56291ms step_avg:151.32ms
step:383/1750 train_loss:3.8181 train_time:56444ms step_avg:151.32ms
step:384/1750 train_loss:3.8853 train_time:56595ms step_avg:151.32ms
step:385/1750 train_loss:3.8079 train_time:56747ms step_avg:151.32ms
step:386/1750 train_loss:3.9114 train_time:56898ms step_avg:151.32ms
step:387/1750 train_loss:4.0893 train_time:57050ms step_avg:151.33ms
step:388/1750 train_loss:3.8171 train_time:57201ms step_avg:151.32ms
step:389/1750 train_loss:3.8173 train_time:57352ms step_avg:151.33ms
step:390/1750 train_loss:3.9083 train_time:57506ms step_avg:151.33ms
step:391/1750 train_loss:3.8283 train_time:57662ms step_avg:151.34ms
step:392/1750 train_loss:3.9381 train_time:57813ms step_avg:151.34ms
step:393/1750 train_loss:3.7764 train_time:57969ms step_avg:151.36ms
step:394/1750 train_loss:3.9032 train_time:58124ms step_avg:151.36ms
step:395/1750 train_loss:3.6474 train_time:58278ms step_avg:151.37ms
step:396/1750 train_loss:3.8532 train_time:58433ms step_avg:151.38ms
step:397/1750 train_loss:3.8921 train_time:58588ms step_avg:151.39ms
step:398/1750 train_loss:3.8920 train_time:58743ms step_avg:151.40ms
step:399/1750 train_loss:3.7910 train_time:58896ms step_avg:151.40ms
step:400/1750 train_loss:3.8472 train_time:59050ms step_avg:151.41ms
step:401/1750 train_loss:3.9319 train_time:59204ms step_avg:151.42ms
step:402/1750 train_loss:3.8622 train_time:59358ms step_avg:151.42ms
step:403/1750 train_loss:3.9761 train_time:59513ms step_avg:151.43ms
step:404/1750 train_loss:3.6996 train_time:59667ms step_avg:151.44ms
step:405/1750 train_loss:3.8070 train_time:59822ms step_avg:151.45ms
step:406/1750 train_loss:4.1120 train_time:59975ms step_avg:151.45ms
step:407/1750 train_loss:3.8007 train_time:60130ms step_avg:151.46ms
step:408/1750 train_loss:3.8343 train_time:60283ms step_avg:151.46ms
step:409/1750 train_loss:3.8794 train_time:60438ms step_avg:151.47ms
step:410/1750 train_loss:3.7795 train_time:60591ms step_avg:151.48ms
step:411/1750 train_loss:3.7781 train_time:60746ms step_avg:151.49ms
step:412/1750 train_loss:4.2013 train_time:60901ms step_avg:151.49ms
step:413/1750 train_loss:3.6627 train_time:61054ms step_avg:151.50ms
step:414/1750 train_loss:4.0295 train_time:61209ms step_avg:151.51ms
step:415/1750 train_loss:3.7768 train_time:61363ms step_avg:151.51ms
step:416/1750 train_loss:3.7782 train_time:61517ms step_avg:151.52ms
step:417/1750 train_loss:3.9748 train_time:61671ms step_avg:151.52ms
step:418/1750 train_loss:3.7082 train_time:61824ms step_avg:151.53ms
step:419/1750 train_loss:3.8269 train_time:61979ms step_avg:151.54ms
step:420/1750 train_loss:3.7233 train_time:62133ms step_avg:151.54ms
step:421/1750 train_loss:3.6700 train_time:62287ms step_avg:151.55ms
step:422/1750 train_loss:3.8082 train_time:62442ms step_avg:151.56ms
step:423/1750 train_loss:3.8901 train_time:62597ms step_avg:151.57ms
step:424/1750 train_loss:3.6383 train_time:62751ms step_avg:151.57ms
step:425/1750 train_loss:3.8166 train_time:62907ms step_avg:151.58ms
step:426/1750 train_loss:3.6834 train_time:63061ms step_avg:151.59ms
step:427/1750 train_loss:3.9126 train_time:63215ms step_avg:151.59ms
step:428/1750 train_loss:3.8358 train_time:63369ms step_avg:151.60ms
step:429/1750 train_loss:3.7749 train_time:63523ms step_avg:151.61ms
step:430/1750 train_loss:3.7411 train_time:63677ms step_avg:151.61ms
step:431/1750 train_loss:3.6424 train_time:63832ms step_avg:151.62ms
step:432/1750 train_loss:3.7862 train_time:63987ms step_avg:151.63ms
step:433/1750 train_loss:3.8426 train_time:64143ms step_avg:151.64ms
step:434/1750 train_loss:3.7947 train_time:64297ms step_avg:151.64ms
step:435/1750 train_loss:3.8318 train_time:64450ms step_avg:151.65ms
step:436/1750 train_loss:3.8575 train_time:64604ms step_avg:151.65ms
step:437/1750 train_loss:3.7365 train_time:64761ms step_avg:151.66ms
step:438/1750 train_loss:3.7267 train_time:64914ms step_avg:151.67ms
step:439/1750 train_loss:3.7265 train_time:65068ms step_avg:151.67ms
step:440/1750 train_loss:3.9141 train_time:65223ms step_avg:151.68ms
step:441/1750 train_loss:3.7805 train_time:65376ms step_avg:151.69ms
step:442/1750 train_loss:3.7593 train_time:65531ms step_avg:151.69ms
step:443/1750 train_loss:3.6407 train_time:65686ms step_avg:151.70ms
step:444/1750 train_loss:3.9435 train_time:65842ms step_avg:151.71ms
step:445/1750 train_loss:3.8679 train_time:65995ms step_avg:151.71ms
step:446/1750 train_loss:3.8550 train_time:66151ms step_avg:151.72ms
step:447/1750 train_loss:3.7688 train_time:66304ms step_avg:151.73ms
step:448/1750 train_loss:3.8712 train_time:66461ms step_avg:151.74ms
step:449/1750 train_loss:3.7104 train_time:66615ms step_avg:151.74ms
step:450/1750 train_loss:3.7440 train_time:66770ms step_avg:151.75ms
step:451/1750 train_loss:3.6094 train_time:66925ms step_avg:151.76ms
step:452/1750 train_loss:3.7275 train_time:67079ms step_avg:151.76ms
step:453/1750 train_loss:3.7029 train_time:67234ms step_avg:151.77ms
step:454/1750 train_loss:3.6592 train_time:67388ms step_avg:151.77ms
step:455/1750 train_loss:3.8591 train_time:67544ms step_avg:151.78ms
step:456/1750 train_loss:3.7475 train_time:67697ms step_avg:151.79ms
step:457/1750 train_loss:3.8038 train_time:67851ms step_avg:151.79ms
step:458/1750 train_loss:3.8523 train_time:68004ms step_avg:151.79ms
step:459/1750 train_loss:3.6535 train_time:68159ms step_avg:151.80ms
step:460/1750 train_loss:3.8092 train_time:68311ms step_avg:151.80ms
step:461/1750 train_loss:3.7069 train_time:68467ms step_avg:151.81ms
step:462/1750 train_loss:3.7567 train_time:68622ms step_avg:151.82ms
step:463/1750 train_loss:3.7997 train_time:68775ms step_avg:151.82ms
step:464/1750 train_loss:3.7349 train_time:68929ms step_avg:151.83ms
step:465/1750 train_loss:3.7412 train_time:69082ms step_avg:151.83ms
step:466/1750 train_loss:3.8268 train_time:69234ms step_avg:151.83ms
step:467/1750 train_loss:3.8421 train_time:69388ms step_avg:151.83ms
step:468/1750 train_loss:3.8119 train_time:69542ms step_avg:151.84ms
step:469/1750 train_loss:3.7032 train_time:69696ms step_avg:151.84ms
step:470/1750 train_loss:3.7894 train_time:69850ms step_avg:151.85ms
step:471/1750 train_loss:3.8339 train_time:70004ms step_avg:151.85ms
step:472/1750 train_loss:3.8026 train_time:70160ms step_avg:151.86ms
step:473/1750 train_loss:3.7345 train_time:70312ms step_avg:151.86ms
step:474/1750 train_loss:3.6034 train_time:70466ms step_avg:151.87ms
step:475/1750 train_loss:4.0385 train_time:70621ms step_avg:151.87ms
step:476/1750 train_loss:3.7840 train_time:70773ms step_avg:151.87ms
step:477/1750 train_loss:3.6103 train_time:70928ms step_avg:151.88ms
step:478/1750 train_loss:3.8447 train_time:71082ms step_avg:151.88ms
step:479/1750 train_loss:3.7899 train_time:71235ms step_avg:151.89ms
step:480/1750 train_loss:3.9397 train_time:71388ms step_avg:151.89ms
step:481/1750 train_loss:3.7422 train_time:71544ms step_avg:151.90ms
step:482/1750 train_loss:3.5509 train_time:71697ms step_avg:151.90ms
step:483/1750 train_loss:3.8344 train_time:71852ms step_avg:151.91ms
step:484/1750 train_loss:3.6824 train_time:72005ms step_avg:151.91ms
step:485/1750 train_loss:3.6797 train_time:72161ms step_avg:151.92ms
step:486/1750 train_loss:3.5946 train_time:72313ms step_avg:151.92ms
step:487/1750 train_loss:3.6988 train_time:72466ms step_avg:151.92ms
step:488/1750 train_loss:3.8981 train_time:72620ms step_avg:151.93ms
step:489/1750 train_loss:3.7316 train_time:72772ms step_avg:151.93ms
step:490/1750 train_loss:3.6128 train_time:72927ms step_avg:151.93ms
step:491/1750 train_loss:3.6328 train_time:73082ms step_avg:151.94ms
step:492/1750 train_loss:3.7526 train_time:73235ms step_avg:151.94ms
step:493/1750 train_loss:3.6022 train_time:73388ms step_avg:151.94ms
step:494/1750 train_loss:3.7210 train_time:73543ms step_avg:151.95ms
step:495/1750 train_loss:3.6804 train_time:73697ms step_avg:151.95ms
step:496/1750 train_loss:3.5389 train_time:73851ms step_avg:151.96ms
step:497/1750 train_loss:3.7528 train_time:74004ms step_avg:151.96ms
step:498/1750 train_loss:3.8051 train_time:74158ms step_avg:151.96ms
step:499/1750 train_loss:3.8377 train_time:74311ms step_avg:151.96ms
step:500/1750 train_loss:3.7551 train_time:74464ms step_avg:151.97ms
step:500/1750 val_loss:3.7255 train_time:74504ms step_avg:152.05ms
step:501/1750 train_loss:3.8210 train_time:74621ms step_avg:151.98ms
step:502/1750 train_loss:3.7680 train_time:74775ms step_avg:151.98ms
step:503/1750 train_loss:3.7945 train_time:74927ms step_avg:151.98ms
step:504/1750 train_loss:3.7527 train_time:75080ms step_avg:151.98ms
step:505/1750 train_loss:3.8235 train_time:75234ms step_avg:151.99ms
step:506/1750 train_loss:3.6798 train_time:75388ms step_avg:151.99ms
step:507/1750 train_loss:3.7836 train_time:75541ms step_avg:151.99ms
step:508/1750 train_loss:3.8465 train_time:75698ms step_avg:152.00ms
step:509/1750 train_loss:3.7984 train_time:75851ms step_avg:152.01ms
step:510/1750 train_loss:3.5987 train_time:76005ms step_avg:152.01ms
step:511/1750 train_loss:3.8030 train_time:76158ms step_avg:152.01ms
step:512/1750 train_loss:3.7536 train_time:76312ms step_avg:152.02ms
step:513/1750 train_loss:3.6848 train_time:76465ms step_avg:152.02ms
step:514/1750 train_loss:3.8531 train_time:76620ms step_avg:152.02ms
step:515/1750 train_loss:3.7572 train_time:76774ms step_avg:152.03ms
step:516/1750 train_loss:4.1038 train_time:76931ms step_avg:152.04ms
step:517/1750 train_loss:3.7041 train_time:77084ms step_avg:152.04ms
step:518/1750 train_loss:3.7908 train_time:77239ms step_avg:152.04ms
step:519/1750 train_loss:3.6829 train_time:77392ms step_avg:152.05ms
step:520/1750 train_loss:3.7037 train_time:77548ms step_avg:152.05ms
step:521/1750 train_loss:3.6737 train_time:77705ms step_avg:152.06ms
step:522/1750 train_loss:3.6724 train_time:77860ms step_avg:152.07ms
step:523/1750 train_loss:4.3148 train_time:78019ms step_avg:152.08ms
step:524/1750 train_loss:3.7586 train_time:78176ms step_avg:152.09ms
step:525/1750 train_loss:3.7041 train_time:78331ms step_avg:152.10ms
step:526/1750 train_loss:3.7189 train_time:78488ms step_avg:152.11ms
step:527/1750 train_loss:3.6784 train_time:78644ms step_avg:152.12ms
step:528/1750 train_loss:3.6506 train_time:78801ms step_avg:152.13ms
step:529/1750 train_loss:3.8666 train_time:78958ms step_avg:152.13ms
step:530/1750 train_loss:3.6610 train_time:79114ms step_avg:152.14ms
step:531/1750 train_loss:3.9413 train_time:79269ms step_avg:152.15ms
step:532/1750 train_loss:3.7521 train_time:79426ms step_avg:152.16ms
step:533/1750 train_loss:3.6763 train_time:79583ms step_avg:152.17ms
step:534/1750 train_loss:3.6887 train_time:79738ms step_avg:152.17ms
step:535/1750 train_loss:3.6246 train_time:79894ms step_avg:152.18ms
step:536/1750 train_loss:3.7705 train_time:80052ms step_avg:152.19ms
step:537/1750 train_loss:3.7458 train_time:80207ms step_avg:152.20ms
step:538/1750 train_loss:3.6524 train_time:80362ms step_avg:152.20ms
step:539/1750 train_loss:4.1415 train_time:80522ms step_avg:152.22ms
step:540/1750 train_loss:3.6938 train_time:80679ms step_avg:152.22ms
step:541/1750 train_loss:3.8114 train_time:80834ms step_avg:152.23ms
step:542/1750 train_loss:3.6152 train_time:80991ms step_avg:152.24ms
step:543/1750 train_loss:3.6133 train_time:81147ms step_avg:152.25ms
step:544/1750 train_loss:3.6695 train_time:81304ms step_avg:152.25ms
step:545/1750 train_loss:3.6066 train_time:81460ms step_avg:152.26ms
step:546/1750 train_loss:3.6450 train_time:81615ms step_avg:152.27ms
step:547/1750 train_loss:3.6692 train_time:81771ms step_avg:152.27ms
step:548/1750 train_loss:3.6381 train_time:81928ms step_avg:152.28ms
step:549/1750 train_loss:3.7480 train_time:82084ms step_avg:152.29ms
step:550/1750 train_loss:3.6367 train_time:82242ms step_avg:152.30ms
step:551/1750 train_loss:3.6477 train_time:82398ms step_avg:152.31ms
step:552/1750 train_loss:3.9564 train_time:82554ms step_avg:152.31ms
step:553/1750 train_loss:3.7831 train_time:82711ms step_avg:152.32ms
step:554/1750 train_loss:3.7361 train_time:82866ms step_avg:152.33ms
step:555/1750 train_loss:3.6545 train_time:83023ms step_avg:152.34ms
step:556/1750 train_loss:3.7179 train_time:83179ms step_avg:152.34ms
step:557/1750 train_loss:3.3304 train_time:83334ms step_avg:152.35ms
step:558/1750 train_loss:3.6334 train_time:83490ms step_avg:152.35ms
step:559/1750 train_loss:3.6684 train_time:83645ms step_avg:152.36ms
step:560/1750 train_loss:3.7141 train_time:83803ms step_avg:152.37ms
step:561/1750 train_loss:3.6332 train_time:83958ms step_avg:152.37ms
step:562/1750 train_loss:3.5758 train_time:84114ms step_avg:152.38ms
step:563/1750 train_loss:3.7816 train_time:84268ms step_avg:152.38ms
step:564/1750 train_loss:3.5931 train_time:84426ms step_avg:152.39ms
step:565/1750 train_loss:3.7048 train_time:84582ms step_avg:152.40ms
step:566/1750 train_loss:3.6477 train_time:84865ms step_avg:152.64ms
step:567/1750 train_loss:3.6201 train_time:85029ms step_avg:152.65ms
step:568/1750 train_loss:3.7130 train_time:85184ms step_avg:152.66ms
step:569/1750 train_loss:3.6725 train_time:85531ms step_avg:153.01ms
step:570/1750 train_loss:3.7091 train_time:85685ms step_avg:153.01ms
step:571/1750 train_loss:3.7849 train_time:85840ms step_avg:153.01ms
step:572/1750 train_loss:3.7465 train_time:85997ms step_avg:153.02ms
step:573/1750 train_loss:3.7628 train_time:86152ms step_avg:153.02ms
step:574/1750 train_loss:3.8042 train_time:86311ms step_avg:153.03ms
step:575/1750 train_loss:3.7462 train_time:86466ms step_avg:153.04ms
step:576/1750 train_loss:3.7818 train_time:86623ms step_avg:153.04ms
step:577/1750 train_loss:3.6946 train_time:86779ms step_avg:153.05ms
step:578/1750 train_loss:3.6959 train_time:86935ms step_avg:153.06ms
step:579/1750 train_loss:3.6922 train_time:87091ms step_avg:153.06ms
step:580/1750 train_loss:3.6215 train_time:87247ms step_avg:153.07ms
step:581/1750 train_loss:3.6593 train_time:87405ms step_avg:153.07ms
step:582/1750 train_loss:3.8768 train_time:87561ms step_avg:153.08ms
step:583/1750 train_loss:3.6479 train_time:87717ms step_avg:153.08ms
step:584/1750 train_loss:3.6166 train_time:87873ms step_avg:153.09ms
step:585/1750 train_loss:3.8099 train_time:88027ms step_avg:153.09ms
step:586/1750 train_loss:3.5314 train_time:88183ms step_avg:153.10ms
step:587/1750 train_loss:3.6917 train_time:88338ms step_avg:153.10ms
step:588/1750 train_loss:3.6748 train_time:88493ms step_avg:153.10ms
step:589/1750 train_loss:4.0170 train_time:88650ms step_avg:153.11ms
step:590/1750 train_loss:3.8086 train_time:88806ms step_avg:153.11ms
step:591/1750 train_loss:3.5359 train_time:88961ms step_avg:153.12ms
step:592/1750 train_loss:3.5638 train_time:89118ms step_avg:153.12ms
step:593/1750 train_loss:3.5284 train_time:89273ms step_avg:153.13ms
step:594/1750 train_loss:3.5778 train_time:89430ms step_avg:153.13ms
step:595/1750 train_loss:3.9476 train_time:89587ms step_avg:153.14ms
step:596/1750 train_loss:3.6638 train_time:89743ms step_avg:153.15ms
step:597/1750 train_loss:3.6086 train_time:89898ms step_avg:153.15ms
step:598/1750 train_loss:3.6852 train_time:90052ms step_avg:153.15ms
step:599/1750 train_loss:3.4921 train_time:90209ms step_avg:153.16ms
step:600/1750 train_loss:3.6204 train_time:90365ms step_avg:153.16ms
step:601/1750 train_loss:3.6601 train_time:90521ms step_avg:153.17ms
step:602/1750 train_loss:3.6845 train_time:90677ms step_avg:153.17ms
step:603/1750 train_loss:3.8057 train_time:90833ms step_avg:153.18ms
step:604/1750 train_loss:3.6377 train_time:90989ms step_avg:153.18ms
step:605/1750 train_loss:3.6370 train_time:91145ms step_avg:153.19ms
step:606/1750 train_loss:3.5943 train_time:91302ms step_avg:153.19ms
step:607/1750 train_loss:3.8601 train_time:91458ms step_avg:153.20ms
step:608/1750 train_loss:3.6643 train_time:91614ms step_avg:153.20ms
step:609/1750 train_loss:3.6307 train_time:91768ms step_avg:153.20ms
step:610/1750 train_loss:3.7294 train_time:91923ms step_avg:153.21ms
step:611/1750 train_loss:3.6233 train_time:92080ms step_avg:153.21ms
step:612/1750 train_loss:3.5943 train_time:92235ms step_avg:153.21ms
step:613/1750 train_loss:3.7865 train_time:92390ms step_avg:153.22ms
step:614/1750 train_loss:3.7243 train_time:92546ms step_avg:153.22ms
step:615/1750 train_loss:3.7071 train_time:92702ms step_avg:153.23ms
step:616/1750 train_loss:3.6498 train_time:92857ms step_avg:153.23ms
step:617/1750 train_loss:3.5797 train_time:93012ms step_avg:153.23ms
step:618/1750 train_loss:3.7105 train_time:93166ms step_avg:153.23ms
step:619/1750 train_loss:3.5854 train_time:93323ms step_avg:153.24ms
step:620/1750 train_loss:3.6126 train_time:93479ms step_avg:153.24ms
step:621/1750 train_loss:3.9479 train_time:93634ms step_avg:153.25ms
step:622/1750 train_loss:3.5934 train_time:93789ms step_avg:153.25ms
step:623/1750 train_loss:3.6256 train_time:93946ms step_avg:153.26ms
step:624/1750 train_loss:3.7191 train_time:94104ms step_avg:153.26ms
step:625/1750 train_loss:3.7304 train_time:94259ms step_avg:153.27ms
step:625/1750 val_loss:3.6455 train_time:94299ms step_avg:153.33ms
step:626/1750 train_loss:3.7650 train_time:94417ms step_avg:153.27ms
step:627/1750 train_loss:3.7331 train_time:94572ms step_avg:153.28ms
step:628/1750 train_loss:3.7899 train_time:94727ms step_avg:153.28ms
step:629/1750 train_loss:3.6126 train_time:94883ms step_avg:153.28ms
step:630/1750 train_loss:3.7480 train_time:95037ms step_avg:153.29ms
step:631/1750 train_loss:3.7654 train_time:95191ms step_avg:153.29ms
step:632/1750 train_loss:3.6741 train_time:95348ms step_avg:153.29ms
step:633/1750 train_loss:3.6208 train_time:95504ms step_avg:153.30ms
step:634/1750 train_loss:3.7267 train_time:95660ms step_avg:153.30ms
step:635/1750 train_loss:3.9796 train_time:95815ms step_avg:153.30ms
step:636/1750 train_loss:3.5647 train_time:95971ms step_avg:153.31ms
step:637/1750 train_loss:3.3699 train_time:96127ms step_avg:153.31ms
step:638/1750 train_loss:3.6113 train_time:96282ms step_avg:153.32ms
step:639/1750 train_loss:3.6531 train_time:96438ms step_avg:153.32ms
step:640/1750 train_loss:3.5879 train_time:96593ms step_avg:153.32ms
step:641/1750 train_loss:3.6120 train_time:96748ms step_avg:153.32ms
step:642/1750 train_loss:3.6530 train_time:96903ms step_avg:153.33ms
step:643/1750 train_loss:3.6308 train_time:97059ms step_avg:153.33ms
step:644/1750 train_loss:3.5791 train_time:97214ms step_avg:153.33ms
step:645/1750 train_loss:3.8021 train_time:97370ms step_avg:153.34ms
step:646/1750 train_loss:3.6963 train_time:97526ms step_avg:153.34ms
step:647/1750 train_loss:3.6893 train_time:97681ms step_avg:153.35ms
step:648/1750 train_loss:3.7297 train_time:97837ms step_avg:153.35ms
step:649/1750 train_loss:3.7875 train_time:97992ms step_avg:153.35ms
step:650/1750 train_loss:3.6409 train_time:98150ms step_avg:153.36ms
step:651/1750 train_loss:3.7936 train_time:98309ms step_avg:153.37ms
step:652/1750 train_loss:3.6054 train_time:98468ms step_avg:153.38ms
step:653/1750 train_loss:3.6884 train_time:98625ms step_avg:153.38ms
step:654/1750 train_loss:3.4483 train_time:98784ms step_avg:153.39ms
step:655/1750 train_loss:3.6000 train_time:98940ms step_avg:153.40ms
step:656/1750 train_loss:3.5964 train_time:99099ms step_avg:153.40ms
step:657/1750 train_loss:3.5285 train_time:99259ms step_avg:153.41ms
step:658/1750 train_loss:3.7134 train_time:99417ms step_avg:153.42ms
step:659/1750 train_loss:3.6141 train_time:99576ms step_avg:153.43ms
step:660/1750 train_loss:3.7102 train_time:99736ms step_avg:153.44ms
step:661/1750 train_loss:3.7725 train_time:99894ms step_avg:153.45ms
step:662/1750 train_loss:3.6889 train_time:100050ms step_avg:153.45ms
step:663/1750 train_loss:3.5740 train_time:100207ms step_avg:153.46ms
step:664/1750 train_loss:3.6392 train_time:100366ms step_avg:153.47ms
step:665/1750 train_loss:3.5164 train_time:100524ms step_avg:153.47ms
step:666/1750 train_loss:3.8074 train_time:100680ms step_avg:153.48ms
step:667/1750 train_loss:3.6345 train_time:100840ms step_avg:153.49ms
step:668/1750 train_loss:3.6676 train_time:101000ms step_avg:153.50ms
step:669/1750 train_loss:3.5049 train_time:101162ms step_avg:153.51ms
step:670/1750 train_loss:3.6295 train_time:101319ms step_avg:153.51ms
step:671/1750 train_loss:3.5854 train_time:101475ms step_avg:153.52ms
step:672/1750 train_loss:3.5928 train_time:101633ms step_avg:153.52ms
step:673/1750 train_loss:3.8751 train_time:101792ms step_avg:153.53ms
step:674/1750 train_loss:3.6558 train_time:101952ms step_avg:153.54ms
step:675/1750 train_loss:3.7344 train_time:102109ms step_avg:153.55ms
step:676/1750 train_loss:3.5076 train_time:102268ms step_avg:153.56ms
step:677/1750 train_loss:3.6181 train_time:102427ms step_avg:153.56ms
step:678/1750 train_loss:3.5759 train_time:102584ms step_avg:153.57ms
step:679/1750 train_loss:3.6985 train_time:102744ms step_avg:153.58ms
step:680/1750 train_loss:3.6058 train_time:102903ms step_avg:153.59ms
step:681/1750 train_loss:3.6333 train_time:103060ms step_avg:153.59ms
step:682/1750 train_loss:3.6893 train_time:103220ms step_avg:153.60ms
step:683/1750 train_loss:3.7554 train_time:103379ms step_avg:153.61ms
step:684/1750 train_loss:3.6692 train_time:103540ms step_avg:153.62ms
step:685/1750 train_loss:3.7095 train_time:103699ms step_avg:153.63ms
step:686/1750 train_loss:3.6592 train_time:103859ms step_avg:153.64ms
step:687/1750 train_loss:3.6911 train_time:104014ms step_avg:153.64ms
step:688/1750 train_loss:3.2490 train_time:104175ms step_avg:153.65ms
step:689/1750 train_loss:3.4325 train_time:104334ms step_avg:153.66ms
step:690/1750 train_loss:3.5669 train_time:104494ms step_avg:153.67ms
step:691/1750 train_loss:3.4415 train_time:104651ms step_avg:153.67ms
step:692/1750 train_loss:3.6513 train_time:104807ms step_avg:153.68ms
step:693/1750 train_loss:3.6703 train_time:104966ms step_avg:153.68ms
step:694/1750 train_loss:3.5808 train_time:105124ms step_avg:153.69ms
step:695/1750 train_loss:3.5584 train_time:105281ms step_avg:153.69ms
step:696/1750 train_loss:3.8811 train_time:105440ms step_avg:153.70ms
step:697/1750 train_loss:3.6147 train_time:105601ms step_avg:153.71ms
step:698/1750 train_loss:3.6710 train_time:105761ms step_avg:153.72ms
step:699/1750 train_loss:3.7926 train_time:105920ms step_avg:153.73ms
step:700/1750 train_loss:3.5912 train_time:106078ms step_avg:153.74ms
step:701/1750 train_loss:3.5688 train_time:106237ms step_avg:153.74ms
step:702/1750 train_loss:3.5331 train_time:106395ms step_avg:153.75ms
step:703/1750 train_loss:3.5184 train_time:106555ms step_avg:153.76ms
step:704/1750 train_loss:3.5893 train_time:106711ms step_avg:153.76ms
step:705/1750 train_loss:3.5761 train_time:106872ms step_avg:153.77ms
step:706/1750 train_loss:3.6070 train_time:107032ms step_avg:153.78ms
step:707/1750 train_loss:3.6757 train_time:107190ms step_avg:153.79ms
step:708/1750 train_loss:3.6223 train_time:107349ms step_avg:153.79ms
step:709/1750 train_loss:3.6028 train_time:107508ms step_avg:153.80ms
step:710/1750 train_loss:3.5637 train_time:107665ms step_avg:153.81ms
step:711/1750 train_loss:3.6162 train_time:107824ms step_avg:153.81ms
step:712/1750 train_loss:3.6744 train_time:107985ms step_avg:153.82ms
step:713/1750 train_loss:3.6734 train_time:108145ms step_avg:153.83ms
step:714/1750 train_loss:3.5815 train_time:108302ms step_avg:153.84ms
step:715/1750 train_loss:3.5904 train_time:108460ms step_avg:153.84ms
step:716/1750 train_loss:3.6091 train_time:108616ms step_avg:153.85ms
step:717/1750 train_loss:3.7390 train_time:108774ms step_avg:153.85ms
step:718/1750 train_loss:3.6209 train_time:108931ms step_avg:153.86ms
step:719/1750 train_loss:3.6978 train_time:109087ms step_avg:153.86ms
step:720/1750 train_loss:3.8682 train_time:109247ms step_avg:153.87ms
step:721/1750 train_loss:3.4855 train_time:109404ms step_avg:153.87ms
step:722/1750 train_loss:3.7562 train_time:109562ms step_avg:153.88ms
step:723/1750 train_loss:3.7913 train_time:109718ms step_avg:153.88ms
step:724/1750 train_loss:3.5867 train_time:109874ms step_avg:153.89ms
step:725/1750 train_loss:3.6725 train_time:110033ms step_avg:153.89ms
step:726/1750 train_loss:3.5627 train_time:110192ms step_avg:153.90ms
step:727/1750 train_loss:3.5974 train_time:110352ms step_avg:153.91ms
step:728/1750 train_loss:3.7562 train_time:110508ms step_avg:153.91ms
step:729/1750 train_loss:3.6947 train_time:110666ms step_avg:153.92ms
step:730/1750 train_loss:3.6923 train_time:110824ms step_avg:153.92ms
step:731/1750 train_loss:3.5841 train_time:110980ms step_avg:153.92ms
step:732/1750 train_loss:3.6187 train_time:111136ms step_avg:153.93ms
step:733/1750 train_loss:3.8632 train_time:111294ms step_avg:153.93ms
step:734/1750 train_loss:3.5861 train_time:111451ms step_avg:153.94ms
step:735/1750 train_loss:3.6359 train_time:111607ms step_avg:153.94ms
step:736/1750 train_loss:3.7587 train_time:111765ms step_avg:153.95ms
step:737/1750 train_loss:3.6977 train_time:111922ms step_avg:153.95ms
step:738/1750 train_loss:3.6218 train_time:112078ms step_avg:153.95ms
step:739/1750 train_loss:3.5251 train_time:112235ms step_avg:153.96ms
step:740/1750 train_loss:4.1344 train_time:112396ms step_avg:153.97ms
step:741/1750 train_loss:3.5145 train_time:112553ms step_avg:153.97ms
step:742/1750 train_loss:3.5826 train_time:112709ms step_avg:153.97ms
step:743/1750 train_loss:3.6017 train_time:112868ms step_avg:153.98ms
step:744/1750 train_loss:3.6745 train_time:113026ms step_avg:153.99ms
step:745/1750 train_loss:3.6130 train_time:113184ms step_avg:153.99ms
step:746/1750 train_loss:3.6134 train_time:113341ms step_avg:154.00ms
step:747/1750 train_loss:3.6699 train_time:113498ms step_avg:154.00ms
step:748/1750 train_loss:3.5870 train_time:113658ms step_avg:154.01ms
step:749/1750 train_loss:3.5870 train_time:113815ms step_avg:154.01ms
step:750/1750 train_loss:3.6310 train_time:113973ms step_avg:154.02ms
step:750/1750 val_loss:3.5923 train_time:114015ms step_avg:154.07ms
step:751/1750 train_loss:3.5933 train_time:114132ms step_avg:154.02ms
step:752/1750 train_loss:3.6381 train_time:114292ms step_avg:154.03ms
step:753/1750 train_loss:3.6411 train_time:114450ms step_avg:154.04ms
step:754/1750 train_loss:3.6113 train_time:114606ms step_avg:154.04ms
step:755/1750 train_loss:3.6992 train_time:114892ms step_avg:154.22ms
step:756/1750 train_loss:3.4844 train_time:115056ms step_avg:154.23ms
step:757/1750 train_loss:3.7509 train_time:115218ms step_avg:154.24ms
step:758/1750 train_loss:3.6791 train_time:115372ms step_avg:154.24ms
step:759/1750 train_loss:3.6191 train_time:115672ms step_avg:154.44ms
step:760/1750 train_loss:3.7282 train_time:115828ms step_avg:154.44ms
step:761/1750 train_loss:3.4213 train_time:115984ms step_avg:154.44ms
step:762/1750 train_loss:3.5710 train_time:116141ms step_avg:154.44ms
step:763/1750 train_loss:3.6922 train_time:116297ms step_avg:154.45ms
step:764/1750 train_loss:3.3377 train_time:116454ms step_avg:154.45ms
step:765/1750 train_loss:3.7597 train_time:116611ms step_avg:154.45ms
step:766/1750 train_loss:3.6072 train_time:116769ms step_avg:154.46ms
step:767/1750 train_loss:3.5881 train_time:116926ms step_avg:154.46ms
step:768/1750 train_loss:3.5954 train_time:117084ms step_avg:154.46ms
step:769/1750 train_loss:3.6106 train_time:117243ms step_avg:154.47ms
step:770/1750 train_loss:3.6671 train_time:117402ms step_avg:154.48ms
step:771/1750 train_loss:3.9054 train_time:117561ms step_avg:154.48ms
step:772/1750 train_loss:3.4761 train_time:117718ms step_avg:154.49ms
step:773/1750 train_loss:3.6558 train_time:117874ms step_avg:154.49ms
step:774/1750 train_loss:3.6659 train_time:118031ms step_avg:154.49ms
step:775/1750 train_loss:3.6346 train_time:118186ms step_avg:154.49ms
step:776/1750 train_loss:3.4179 train_time:118344ms step_avg:154.50ms
step:777/1750 train_loss:3.4158 train_time:118501ms step_avg:154.50ms
step:778/1750 train_loss:3.5127 train_time:118658ms step_avg:154.50ms
step:779/1750 train_loss:3.6052 train_time:118816ms step_avg:154.51ms
step:780/1750 train_loss:3.6138 train_time:118976ms step_avg:154.51ms
step:781/1750 train_loss:3.6946 train_time:119135ms step_avg:154.52ms
step:782/1750 train_loss:3.6074 train_time:119295ms step_avg:154.53ms
step:783/1750 train_loss:3.5967 train_time:119452ms step_avg:154.53ms
step:784/1750 train_loss:3.6206 train_time:119612ms step_avg:154.54ms
step:785/1750 train_loss:3.5882 train_time:119769ms step_avg:154.54ms
step:786/1750 train_loss:3.4666 train_time:119931ms step_avg:154.55ms
step:787/1750 train_loss:3.7593 train_time:120089ms step_avg:154.55ms
step:788/1750 train_loss:3.5205 train_time:120247ms step_avg:154.56ms
step:789/1750 train_loss:3.5757 train_time:120405ms step_avg:154.56ms
step:790/1750 train_loss:3.6454 train_time:120566ms step_avg:154.57ms
step:791/1750 train_loss:3.7962 train_time:120727ms step_avg:154.58ms
step:792/1750 train_loss:3.7871 train_time:120886ms step_avg:154.59ms
step:793/1750 train_loss:3.4948 train_time:121044ms step_avg:154.59ms
step:794/1750 train_loss:3.6195 train_time:121206ms step_avg:154.60ms
step:795/1750 train_loss:3.6940 train_time:121366ms step_avg:154.61ms
step:796/1750 train_loss:3.7524 train_time:121527ms step_avg:154.61ms
step:797/1750 train_loss:3.5470 train_time:121685ms step_avg:154.62ms
step:798/1750 train_loss:3.6727 train_time:121846ms step_avg:154.63ms
step:799/1750 train_loss:3.5693 train_time:122008ms step_avg:154.64ms
step:800/1750 train_loss:3.5570 train_time:122166ms step_avg:154.64ms
step:801/1750 train_loss:3.6611 train_time:122326ms step_avg:154.65ms
step:802/1750 train_loss:3.5155 train_time:122487ms step_avg:154.66ms
step:803/1750 train_loss:3.5380 train_time:122645ms step_avg:154.66ms
step:804/1750 train_loss:3.6454 train_time:122806ms step_avg:154.67ms
step:805/1750 train_loss:3.5479 train_time:122968ms step_avg:154.68ms
step:806/1750 train_loss:3.5856 train_time:123126ms step_avg:154.68ms
step:807/1750 train_loss:3.6691 train_time:123286ms step_avg:154.69ms
step:808/1750 train_loss:3.5814 train_time:123445ms step_avg:154.69ms
step:809/1750 train_loss:3.5131 train_time:123603ms step_avg:154.70ms
step:810/1750 train_loss:3.5924 train_time:123762ms step_avg:154.70ms
step:811/1750 train_loss:3.6132 train_time:123923ms step_avg:154.71ms
step:812/1750 train_loss:3.6180 train_time:124081ms step_avg:154.71ms
step:813/1750 train_loss:3.6489 train_time:124240ms step_avg:154.72ms
step:814/1750 train_loss:3.5966 train_time:124400ms step_avg:154.73ms
step:815/1750 train_loss:3.5885 train_time:124559ms step_avg:154.73ms
step:816/1750 train_loss:3.7121 train_time:124721ms step_avg:154.74ms
step:817/1750 train_loss:3.7999 train_time:124879ms step_avg:154.74ms
step:818/1750 train_loss:3.5514 train_time:125037ms step_avg:154.75ms
step:819/1750 train_loss:3.7477 train_time:125198ms step_avg:154.76ms
step:820/1750 train_loss:3.5228 train_time:125357ms step_avg:154.76ms
step:821/1750 train_loss:3.5848 train_time:125514ms step_avg:154.76ms
step:822/1750 train_loss:3.7179 train_time:125675ms step_avg:154.77ms
step:823/1750 train_loss:3.6057 train_time:125835ms step_avg:154.78ms
step:824/1750 train_loss:3.5361 train_time:125992ms step_avg:154.78ms
step:825/1750 train_loss:3.6428 train_time:126154ms step_avg:154.79ms
step:826/1750 train_loss:3.5026 train_time:126319ms step_avg:154.80ms
step:827/1750 train_loss:3.7560 train_time:126478ms step_avg:154.81ms
step:828/1750 train_loss:3.6422 train_time:126638ms step_avg:154.81ms
step:829/1750 train_loss:3.6645 train_time:126798ms step_avg:154.82ms
step:830/1750 train_loss:3.5592 train_time:126957ms step_avg:154.83ms
step:831/1750 train_loss:3.6253 train_time:127118ms step_avg:154.83ms
step:832/1750 train_loss:3.5434 train_time:127278ms step_avg:154.84ms
step:833/1750 train_loss:3.6797 train_time:127439ms step_avg:154.85ms
step:834/1750 train_loss:3.5033 train_time:127599ms step_avg:154.85ms
step:835/1750 train_loss:3.4817 train_time:127759ms step_avg:154.86ms
step:836/1750 train_loss:3.7497 train_time:127920ms step_avg:154.87ms
step:837/1750 train_loss:3.4288 train_time:128079ms step_avg:154.87ms
step:838/1750 train_loss:3.6167 train_time:128238ms step_avg:154.88ms
step:839/1750 train_loss:3.4388 train_time:128397ms step_avg:154.88ms
step:840/1750 train_loss:3.4962 train_time:128554ms step_avg:154.88ms
step:841/1750 train_loss:3.5839 train_time:128712ms step_avg:154.89ms
step:842/1750 train_loss:3.6061 train_time:128875ms step_avg:154.90ms
step:843/1750 train_loss:3.5942 train_time:129034ms step_avg:154.90ms
step:844/1750 train_loss:3.4517 train_time:129191ms step_avg:154.91ms
step:845/1750 train_loss:3.6935 train_time:129352ms step_avg:154.91ms
step:846/1750 train_loss:3.5435 train_time:129514ms step_avg:154.92ms
step:847/1750 train_loss:3.5202 train_time:129673ms step_avg:154.93ms
step:848/1750 train_loss:3.6619 train_time:129829ms step_avg:154.93ms
step:849/1750 train_loss:3.5245 train_time:129988ms step_avg:154.93ms
step:850/1750 train_loss:3.4673 train_time:130146ms step_avg:154.94ms
step:851/1750 train_loss:3.7668 train_time:130307ms step_avg:154.94ms
step:852/1750 train_loss:3.4671 train_time:130464ms step_avg:154.95ms
step:853/1750 train_loss:3.5894 train_time:130623ms step_avg:154.95ms
step:854/1750 train_loss:3.6830 train_time:130780ms step_avg:154.95ms
step:855/1750 train_loss:3.5481 train_time:130940ms step_avg:154.96ms
step:856/1750 train_loss:3.5679 train_time:131100ms step_avg:154.96ms
step:857/1750 train_loss:3.6300 train_time:131258ms step_avg:154.97ms
step:858/1750 train_loss:3.4972 train_time:131420ms step_avg:154.98ms
step:859/1750 train_loss:3.5868 train_time:131579ms step_avg:154.98ms
step:860/1750 train_loss:3.6159 train_time:131736ms step_avg:154.98ms
step:861/1750 train_loss:3.6583 train_time:131899ms step_avg:154.99ms
step:862/1750 train_loss:3.6207 train_time:132062ms step_avg:155.00ms
step:863/1750 train_loss:3.6025 train_time:132222ms step_avg:155.01ms
step:864/1750 train_loss:3.4081 train_time:132380ms step_avg:155.01ms
step:865/1750 train_loss:3.6226 train_time:132537ms step_avg:155.01ms
step:866/1750 train_loss:3.8989 train_time:132698ms step_avg:155.02ms
step:867/1750 train_loss:3.4867 train_time:132856ms step_avg:155.02ms
step:868/1750 train_loss:3.6689 train_time:133014ms step_avg:155.03ms
step:869/1750 train_loss:3.6490 train_time:133172ms step_avg:155.03ms
step:870/1750 train_loss:3.4724 train_time:133332ms step_avg:155.04ms
step:871/1750 train_loss:3.4630 train_time:133490ms step_avg:155.04ms
step:872/1750 train_loss:3.6799 train_time:133650ms step_avg:155.05ms
step:873/1750 train_loss:3.4803 train_time:133809ms step_avg:155.05ms
step:874/1750 train_loss:3.2394 train_time:133969ms step_avg:155.06ms
step:875/1750 train_loss:3.6602 train_time:134128ms step_avg:155.06ms
step:875/1750 val_loss:3.5455 train_time:134169ms step_avg:155.11ms
step:876/1750 train_loss:3.4729 train_time:134287ms step_avg:155.07ms
step:877/1750 train_loss:3.6437 train_time:134448ms step_avg:155.07ms
step:878/1750 train_loss:3.4956 train_time:134606ms step_avg:155.08ms
step:879/1750 train_loss:3.6772 train_time:134766ms step_avg:155.08ms
step:880/1750 train_loss:3.3298 train_time:134924ms step_avg:155.09ms
step:881/1750 train_loss:3.5160 train_time:135083ms step_avg:155.09ms
step:882/1750 train_loss:3.7178 train_time:135240ms step_avg:155.09ms
step:883/1750 train_loss:3.8676 train_time:135400ms step_avg:155.10ms
step:884/1750 train_loss:3.5928 train_time:135560ms step_avg:155.10ms
step:885/1750 train_loss:3.5181 train_time:135718ms step_avg:155.11ms
step:886/1750 train_loss:3.5978 train_time:135877ms step_avg:155.11ms
step:887/1750 train_loss:4.1241 train_time:136038ms step_avg:155.12ms
step:888/1750 train_loss:3.8601 train_time:136201ms step_avg:155.13ms
step:889/1750 train_loss:3.5446 train_time:136360ms step_avg:155.13ms
step:890/1750 train_loss:3.5597 train_time:136517ms step_avg:155.13ms
step:891/1750 train_loss:3.3872 train_time:136677ms step_avg:155.14ms
step:892/1750 train_loss:3.7498 train_time:136836ms step_avg:155.14ms
step:893/1750 train_loss:3.4506 train_time:136993ms step_avg:155.14ms
step:894/1750 train_loss:3.6658 train_time:137154ms step_avg:155.15ms
step:895/1750 train_loss:3.7099 train_time:137311ms step_avg:155.15ms
step:896/1750 train_loss:3.5301 train_time:137470ms step_avg:155.16ms
step:897/1750 train_loss:3.5723 train_time:137631ms step_avg:155.16ms
step:898/1750 train_loss:3.6153 train_time:137791ms step_avg:155.17ms
step:899/1750 train_loss:3.5051 train_time:137947ms step_avg:155.17ms
step:900/1750 train_loss:3.4462 train_time:138106ms step_avg:155.18ms
step:901/1750 train_loss:3.6485 train_time:138265ms step_avg:155.18ms
step:902/1750 train_loss:3.6649 train_time:138422ms step_avg:155.18ms
step:903/1750 train_loss:3.5639 train_time:138584ms step_avg:155.19ms
step:904/1750 train_loss:3.5185 train_time:138743ms step_avg:155.19ms
step:905/1750 train_loss:3.5307 train_time:138902ms step_avg:155.20ms
step:906/1750 train_loss:3.7337 train_time:139062ms step_avg:155.20ms
step:907/1750 train_loss:3.5415 train_time:139221ms step_avg:155.21ms
step:908/1750 train_loss:3.5966 train_time:139380ms step_avg:155.21ms
step:909/1750 train_loss:3.4783 train_time:139540ms step_avg:155.22ms
step:910/1750 train_loss:3.5556 train_time:139703ms step_avg:155.23ms
step:911/1750 train_loss:3.6725 train_time:139865ms step_avg:155.23ms
step:912/1750 train_loss:3.6294 train_time:140028ms step_avg:155.24ms
step:913/1750 train_loss:3.4853 train_time:140189ms step_avg:155.25ms
step:914/1750 train_loss:3.7738 train_time:140350ms step_avg:155.25ms
step:915/1750 train_loss:3.5644 train_time:140513ms step_avg:155.26ms
step:916/1750 train_loss:3.6447 train_time:140674ms step_avg:155.27ms
step:917/1750 train_loss:3.6257 train_time:140834ms step_avg:155.27ms
step:918/1750 train_loss:4.8463 train_time:140998ms step_avg:155.28ms
step:919/1750 train_loss:3.5227 train_time:141162ms step_avg:155.29ms
step:920/1750 train_loss:3.6157 train_time:141321ms step_avg:155.30ms
step:921/1750 train_loss:3.5743 train_time:141483ms step_avg:155.31ms
step:922/1750 train_loss:3.6078 train_time:141645ms step_avg:155.31ms
step:923/1750 train_loss:3.6368 train_time:141805ms step_avg:155.32ms
step:924/1750 train_loss:3.7156 train_time:141967ms step_avg:155.32ms
step:925/1750 train_loss:3.6780 train_time:142126ms step_avg:155.33ms
step:926/1750 train_loss:3.5866 train_time:142286ms step_avg:155.33ms
step:927/1750 train_loss:3.5827 train_time:142447ms step_avg:155.34ms
step:928/1750 train_loss:3.8099 train_time:142609ms step_avg:155.35ms
step:929/1750 train_loss:3.6368 train_time:142770ms step_avg:155.35ms
step:930/1750 train_loss:3.4240 train_time:142931ms step_avg:155.36ms
step:931/1750 train_loss:3.5208 train_time:143090ms step_avg:155.36ms
step:932/1750 train_loss:3.6776 train_time:143251ms step_avg:155.37ms
step:933/1750 train_loss:3.4092 train_time:143412ms step_avg:155.38ms
step:934/1750 train_loss:3.6115 train_time:143577ms step_avg:155.39ms
step:935/1750 train_loss:3.4715 train_time:143742ms step_avg:155.40ms
step:936/1750 train_loss:3.5444 train_time:143906ms step_avg:155.41ms
step:937/1750 train_loss:3.6544 train_time:144069ms step_avg:155.41ms
step:938/1750 train_loss:3.5692 train_time:144229ms step_avg:155.42ms
step:939/1750 train_loss:3.7040 train_time:144394ms step_avg:155.43ms
step:940/1750 train_loss:3.5158 train_time:144554ms step_avg:155.43ms
step:941/1750 train_loss:3.5685 train_time:144714ms step_avg:155.44ms
step:942/1750 train_loss:3.3868 train_time:144875ms step_avg:155.44ms
step:943/1750 train_loss:3.7346 train_time:145039ms step_avg:155.45ms
step:944/1750 train_loss:3.4277 train_time:145326ms step_avg:155.60ms
step:945/1750 train_loss:3.4512 train_time:145493ms step_avg:155.61ms
step:946/1750 train_loss:5.1044 train_time:145655ms step_avg:155.61ms
step:947/1750 train_loss:3.6284 train_time:145816ms step_avg:155.62ms
step:948/1750 train_loss:3.5114 train_time:145978ms step_avg:155.63ms
step:949/1750 train_loss:3.4057 train_time:146276ms step_avg:155.78ms
step:950/1750 train_loss:3.4629 train_time:146435ms step_avg:155.78ms
step:951/1750 train_loss:3.4351 train_time:146599ms step_avg:155.79ms
step:952/1750 train_loss:3.5074 train_time:146760ms step_avg:155.80ms
step:953/1750 train_loss:3.5928 train_time:146923ms step_avg:155.80ms
step:954/1750 train_loss:3.4652 train_time:147085ms step_avg:155.81ms
step:955/1750 train_loss:3.5057 train_time:147245ms step_avg:155.81ms
step:956/1750 train_loss:3.4742 train_time:147407ms step_avg:155.82ms
step:957/1750 train_loss:3.5253 train_time:147569ms step_avg:155.83ms
step:958/1750 train_loss:3.5321 train_time:147732ms step_avg:155.84ms
step:959/1750 train_loss:3.5364 train_time:147893ms step_avg:155.84ms
step:960/1750 train_loss:3.4288 train_time:148054ms step_avg:155.85ms
step:961/1750 train_loss:3.6821 train_time:148213ms step_avg:155.85ms
step:962/1750 train_loss:3.6328 train_time:148374ms step_avg:155.86ms
step:963/1750 train_loss:3.5442 train_time:148537ms step_avg:155.86ms
step:964/1750 train_loss:3.4544 train_time:148700ms step_avg:155.87ms
step:965/1750 train_loss:3.5154 train_time:148859ms step_avg:155.87ms
step:966/1750 train_loss:3.7451 train_time:149020ms step_avg:155.88ms
step:967/1750 train_loss:3.5532 train_time:149181ms step_avg:155.88ms
step:968/1750 train_loss:3.5530 train_time:149341ms step_avg:155.89ms
step:969/1750 train_loss:3.6166 train_time:149503ms step_avg:155.90ms
step:970/1750 train_loss:3.4042 train_time:149663ms step_avg:155.90ms
step:971/1750 train_loss:3.5684 train_time:149822ms step_avg:155.90ms
step:972/1750 train_loss:3.5157 train_time:149982ms step_avg:155.91ms
step:973/1750 train_loss:3.5700 train_time:150141ms step_avg:155.91ms
step:974/1750 train_loss:3.6237 train_time:150304ms step_avg:155.92ms
step:975/1750 train_loss:3.5031 train_time:150466ms step_avg:155.92ms
step:976/1750 train_loss:3.7001 train_time:150625ms step_avg:155.93ms
step:977/1750 train_loss:3.6091 train_time:150784ms step_avg:155.93ms
step:978/1750 train_loss:3.3944 train_time:150944ms step_avg:155.93ms
step:979/1750 train_loss:3.6671 train_time:151103ms step_avg:155.94ms
step:980/1750 train_loss:3.4498 train_time:151263ms step_avg:155.94ms
step:981/1750 train_loss:3.6113 train_time:151426ms step_avg:155.95ms
step:982/1750 train_loss:3.5873 train_time:151586ms step_avg:155.95ms
step:983/1750 train_loss:3.5540 train_time:151746ms step_avg:155.96ms
step:984/1750 train_loss:3.5270 train_time:151905ms step_avg:155.96ms
step:985/1750 train_loss:3.6194 train_time:152065ms step_avg:155.96ms
step:986/1750 train_loss:3.4549 train_time:152225ms step_avg:155.97ms
step:987/1750 train_loss:3.5195 train_time:152384ms step_avg:155.97ms
step:988/1750 train_loss:3.5088 train_time:152543ms step_avg:155.97ms
step:989/1750 train_loss:3.4560 train_time:152703ms step_avg:155.98ms
step:990/1750 train_loss:3.6917 train_time:152865ms step_avg:155.98ms
step:991/1750 train_loss:3.5107 train_time:153025ms step_avg:155.99ms
step:992/1750 train_loss:3.4697 train_time:153191ms step_avg:156.00ms
step:993/1750 train_loss:3.5439 train_time:153357ms step_avg:156.01ms
step:994/1750 train_loss:3.6350 train_time:153516ms step_avg:156.01ms
step:995/1750 train_loss:3.5715 train_time:153675ms step_avg:156.02ms
step:996/1750 train_loss:3.4938 train_time:153834ms step_avg:156.02ms
step:997/1750 train_loss:3.8167 train_time:153994ms step_avg:156.02ms
step:998/1750 train_loss:3.4885 train_time:154155ms step_avg:156.03ms
step:999/1750 train_loss:3.6246 train_time:154314ms step_avg:156.03ms
step:1000/1750 train_loss:3.4793 train_time:154476ms step_avg:156.04ms
step:1000/1750 val_loss:3.5076 train_time:154518ms step_avg:156.08ms
step:1001/1750 train_loss:3.5437 train_time:154637ms step_avg:156.04ms
step:1002/1750 train_loss:3.4226 train_time:154800ms step_avg:156.05ms
step:1003/1750 train_loss:3.5996 train_time:154964ms step_avg:156.06ms
step:1004/1750 train_loss:3.6425 train_time:155124ms step_avg:156.06ms
step:1005/1750 train_loss:3.4293 train_time:155286ms step_avg:156.07ms
step:1006/1750 train_loss:3.5033 train_time:155445ms step_avg:156.07ms
step:1007/1750 train_loss:3.4815 train_time:155605ms step_avg:156.07ms
step:1008/1750 train_loss:3.6005 train_time:155767ms step_avg:156.08ms
step:1009/1750 train_loss:3.6970 train_time:155930ms step_avg:156.09ms
step:1010/1750 train_loss:3.6022 train_time:156089ms step_avg:156.09ms
step:1011/1750 train_loss:3.5753 train_time:156249ms step_avg:156.09ms
step:1012/1750 train_loss:3.4324 train_time:156407ms step_avg:156.10ms
step:1013/1750 train_loss:3.5792 train_time:156569ms step_avg:156.10ms
step:1014/1750 train_loss:3.6658 train_time:156730ms step_avg:156.11ms
step:1015/1750 train_loss:3.3726 train_time:156891ms step_avg:156.11ms
step:1016/1750 train_loss:3.4563 train_time:157051ms step_avg:156.11ms
step:1017/1750 train_loss:3.4428 train_time:157212ms step_avg:156.12ms
step:1018/1750 train_loss:3.4336 train_time:157373ms step_avg:156.12ms
step:1019/1750 train_loss:3.5629 train_time:157534ms step_avg:156.13ms
step:1020/1750 train_loss:3.4335 train_time:157696ms step_avg:156.13ms
step:1021/1750 train_loss:3.3980 train_time:157855ms step_avg:156.14ms
step:1022/1750 train_loss:3.5180 train_time:158015ms step_avg:156.14ms
step:1023/1750 train_loss:3.5512 train_time:158175ms step_avg:156.15ms
step:1024/1750 train_loss:3.5186 train_time:158335ms step_avg:156.15ms
step:1025/1750 train_loss:3.5269 train_time:158496ms step_avg:156.15ms
step:1026/1750 train_loss:3.6749 train_time:158655ms step_avg:156.16ms
step:1027/1750 train_loss:3.3649 train_time:158815ms step_avg:156.16ms
step:1028/1750 train_loss:3.4404 train_time:158978ms step_avg:156.17ms
step:1029/1750 train_loss:3.3757 train_time:159141ms step_avg:156.17ms
step:1030/1750 train_loss:3.5809 train_time:159301ms step_avg:156.18ms
step:1031/1750 train_loss:3.5635 train_time:159461ms step_avg:156.18ms
step:1032/1750 train_loss:3.7423 train_time:159623ms step_avg:156.19ms
step:1033/1750 train_loss:3.5377 train_time:159785ms step_avg:156.19ms
step:1034/1750 train_loss:3.4593 train_time:159948ms step_avg:156.20ms
step:1035/1750 train_loss:3.4892 train_time:160108ms step_avg:156.20ms
step:1036/1750 train_loss:3.5330 train_time:160267ms step_avg:156.21ms
step:1037/1750 train_loss:3.8374 train_time:160427ms step_avg:156.21ms
step:1038/1750 train_loss:3.6664 train_time:160587ms step_avg:156.21ms
step:1039/1750 train_loss:3.5635 train_time:160751ms step_avg:156.22ms
step:1040/1750 train_loss:3.4618 train_time:160911ms step_avg:156.22ms
step:1041/1750 train_loss:3.5324 train_time:161074ms step_avg:156.23ms
step:1042/1750 train_loss:3.5759 train_time:161233ms step_avg:156.23ms
step:1043/1750 train_loss:3.4924 train_time:161394ms step_avg:156.24ms
step:1044/1750 train_loss:3.5089 train_time:161557ms step_avg:156.24ms
step:1045/1750 train_loss:3.5690 train_time:161720ms step_avg:156.25ms
step:1046/1750 train_loss:3.4820 train_time:161881ms step_avg:156.26ms
step:1047/1750 train_loss:3.6921 train_time:162044ms step_avg:156.26ms
step:1048/1750 train_loss:3.5484 train_time:162205ms step_avg:156.27ms
step:1049/1750 train_loss:3.4544 train_time:162366ms step_avg:156.27ms
step:1050/1750 train_loss:3.4438 train_time:162530ms step_avg:156.28ms
step:1051/1750 train_loss:3.5523 train_time:162693ms step_avg:156.29ms
step:1052/1750 train_loss:3.4078 train_time:162857ms step_avg:156.29ms
step:1053/1750 train_loss:3.7426 train_time:163018ms step_avg:156.30ms
step:1054/1750 train_loss:3.5885 train_time:163180ms step_avg:156.30ms
step:1055/1750 train_loss:3.4317 train_time:163340ms step_avg:156.31ms
step:1056/1750 train_loss:3.5566 train_time:163500ms step_avg:156.31ms
step:1057/1750 train_loss:3.6317 train_time:163663ms step_avg:156.32ms
step:1058/1750 train_loss:3.3569 train_time:163828ms step_avg:156.32ms
step:1059/1750 train_loss:3.4200 train_time:163994ms step_avg:156.33ms
step:1060/1750 train_loss:3.4909 train_time:164155ms step_avg:156.34ms
step:1061/1750 train_loss:3.4734 train_time:164314ms step_avg:156.34ms
step:1062/1750 train_loss:3.4312 train_time:164476ms step_avg:156.35ms
step:1063/1750 train_loss:3.5212 train_time:164638ms step_avg:156.35ms
step:1064/1750 train_loss:3.4429 train_time:164797ms step_avg:156.35ms
step:1065/1750 train_loss:3.4179 train_time:164961ms step_avg:156.36ms
step:1066/1750 train_loss:3.4579 train_time:165125ms step_avg:156.37ms
step:1067/1750 train_loss:3.3335 train_time:165290ms step_avg:156.38ms
step:1068/1750 train_loss:3.4858 train_time:165451ms step_avg:156.38ms
step:1069/1750 train_loss:3.3634 train_time:165614ms step_avg:156.39ms
step:1070/1750 train_loss:3.6216 train_time:165776ms step_avg:156.39ms
step:1071/1750 train_loss:3.5629 train_time:165942ms step_avg:156.40ms
step:1072/1750 train_loss:3.4991 train_time:166104ms step_avg:156.41ms
step:1073/1750 train_loss:3.5838 train_time:166263ms step_avg:156.41ms
step:1074/1750 train_loss:3.4957 train_time:166426ms step_avg:156.42ms
step:1075/1750 train_loss:3.4499 train_time:166591ms step_avg:156.42ms
step:1076/1750 train_loss:3.8512 train_time:166753ms step_avg:156.43ms
step:1077/1750 train_loss:3.4995 train_time:166913ms step_avg:156.43ms
step:1078/1750 train_loss:3.1461 train_time:167082ms step_avg:156.44ms
step:1079/1750 train_loss:3.5939 train_time:167244ms step_avg:156.45ms
step:1080/1750 train_loss:3.4844 train_time:167408ms step_avg:156.46ms
step:1081/1750 train_loss:3.5672 train_time:167569ms step_avg:156.46ms
step:1082/1750 train_loss:3.6523 train_time:167731ms step_avg:156.47ms
step:1083/1750 train_loss:3.5629 train_time:167892ms step_avg:156.47ms
step:1084/1750 train_loss:3.5317 train_time:168055ms step_avg:156.48ms
step:1085/1750 train_loss:3.4942 train_time:168215ms step_avg:156.48ms
step:1086/1750 train_loss:3.6921 train_time:168378ms step_avg:156.48ms
step:1087/1750 train_loss:3.5739 train_time:168539ms step_avg:156.49ms
step:1088/1750 train_loss:3.4276 train_time:168704ms step_avg:156.50ms
step:1089/1750 train_loss:3.4360 train_time:168872ms step_avg:156.51ms
step:1090/1750 train_loss:3.5463 train_time:169036ms step_avg:156.51ms
step:1091/1750 train_loss:3.3490 train_time:169197ms step_avg:156.52ms
step:1092/1750 train_loss:3.5500 train_time:169359ms step_avg:156.52ms
step:1093/1750 train_loss:3.6714 train_time:169521ms step_avg:156.53ms
step:1094/1750 train_loss:3.5112 train_time:169682ms step_avg:156.53ms
step:1095/1750 train_loss:3.4807 train_time:169843ms step_avg:156.54ms
step:1096/1750 train_loss:3.4911 train_time:170008ms step_avg:156.54ms
step:1097/1750 train_loss:3.5509 train_time:170171ms step_avg:156.55ms
step:1098/1750 train_loss:3.6234 train_time:170335ms step_avg:156.56ms
step:1099/1750 train_loss:3.5885 train_time:170498ms step_avg:156.56ms
step:1100/1750 train_loss:3.4994 train_time:170662ms step_avg:156.57ms
step:1101/1750 train_loss:3.3446 train_time:170825ms step_avg:156.58ms
step:1102/1750 train_loss:3.3705 train_time:170990ms step_avg:156.58ms
step:1103/1750 train_loss:3.5177 train_time:171156ms step_avg:156.59ms
step:1104/1750 train_loss:3.3837 train_time:171315ms step_avg:156.60ms
step:1105/1750 train_loss:4.1258 train_time:171478ms step_avg:156.60ms
step:1106/1750 train_loss:3.2904 train_time:171638ms step_avg:156.60ms
step:1107/1750 train_loss:3.6309 train_time:171799ms step_avg:156.61ms
step:1108/1750 train_loss:3.4094 train_time:171959ms step_avg:156.61ms
step:1109/1750 train_loss:3.5691 train_time:172118ms step_avg:156.61ms
step:1110/1750 train_loss:3.4971 train_time:172278ms step_avg:156.62ms
step:1111/1750 train_loss:3.5513 train_time:172440ms step_avg:156.62ms
step:1112/1750 train_loss:3.6304 train_time:172603ms step_avg:156.63ms
step:1113/1750 train_loss:3.5023 train_time:172770ms step_avg:156.64ms
step:1114/1750 train_loss:3.4426 train_time:172935ms step_avg:156.64ms
step:1115/1750 train_loss:3.3140 train_time:173097ms step_avg:156.65ms
step:1116/1750 train_loss:3.4963 train_time:173257ms step_avg:156.65ms
step:1117/1750 train_loss:3.6636 train_time:173421ms step_avg:156.66ms
step:1118/1750 train_loss:3.6907 train_time:173584ms step_avg:156.66ms
step:1119/1750 train_loss:3.5432 train_time:173745ms step_avg:156.67ms
step:1120/1750 train_loss:3.5565 train_time:173906ms step_avg:156.67ms
step:1121/1750 train_loss:3.4505 train_time:174068ms step_avg:156.68ms
step:1122/1750 train_loss:3.5249 train_time:174228ms step_avg:156.68ms
step:1123/1750 train_loss:3.6610 train_time:174387ms step_avg:156.68ms
step:1124/1750 train_loss:3.4215 train_time:174549ms step_avg:156.69ms
step:1125/1750 train_loss:3.2945 train_time:174710ms step_avg:156.69ms
step:1125/1750 val_loss:3.4787 train_time:174752ms step_avg:156.73ms
step:1126/1750 train_loss:3.5449 train_time:174871ms step_avg:156.69ms
step:1127/1750 train_loss:3.7483 train_time:175035ms step_avg:156.70ms
step:1128/1750 train_loss:3.3041 train_time:175198ms step_avg:156.71ms
step:1129/1750 train_loss:3.6295 train_time:175361ms step_avg:156.71ms
step:1130/1750 train_loss:3.4525 train_time:175525ms step_avg:156.72ms
step:1131/1750 train_loss:3.4665 train_time:175691ms step_avg:156.73ms
step:1132/1750 train_loss:3.4346 train_time:175849ms step_avg:156.73ms
step:1133/1750 train_loss:3.5671 train_time:176135ms step_avg:156.84ms
step:1134/1750 train_loss:3.5168 train_time:176303ms step_avg:156.85ms
step:1135/1750 train_loss:3.5887 train_time:176463ms step_avg:156.86ms
step:1136/1750 train_loss:3.6255 train_time:176625ms step_avg:156.86ms
step:1137/1750 train_loss:3.5320 train_time:176785ms step_avg:156.86ms
step:1138/1750 train_loss:3.4278 train_time:176948ms step_avg:156.87ms
step:1139/1750 train_loss:3.7254 train_time:177252ms step_avg:157.00ms
step:1140/1750 train_loss:3.5311 train_time:177411ms step_avg:157.00ms
step:1141/1750 train_loss:3.6655 train_time:177573ms step_avg:157.01ms
step:1142/1750 train_loss:3.5272 train_time:177734ms step_avg:157.01ms
step:1143/1750 train_loss:3.4356 train_time:177896ms step_avg:157.01ms
step:1144/1750 train_loss:3.5154 train_time:178059ms step_avg:157.02ms
step:1145/1750 train_loss:3.6638 train_time:178219ms step_avg:157.02ms
step:1146/1750 train_loss:3.6229 train_time:178381ms step_avg:157.03ms
step:1147/1750 train_loss:3.5515 train_time:178541ms step_avg:157.03ms
step:1148/1750 train_loss:3.5661 train_time:178703ms step_avg:157.03ms
step:1149/1750 train_loss:3.3998 train_time:178864ms step_avg:157.04ms
step:1150/1750 train_loss:3.4357 train_time:179025ms step_avg:157.04ms
step:1151/1750 train_loss:3.3864 train_time:179189ms step_avg:157.05ms
step:1152/1750 train_loss:3.4763 train_time:179352ms step_avg:157.05ms
step:1153/1750 train_loss:3.4938 train_time:179513ms step_avg:157.05ms
step:1154/1750 train_loss:3.5916 train_time:179673ms step_avg:157.06ms
step:1155/1750 train_loss:3.3963 train_time:179835ms step_avg:157.06ms
step:1156/1750 train_loss:3.6086 train_time:180004ms step_avg:157.07ms
step:1157/1750 train_loss:3.5687 train_time:180165ms step_avg:157.07ms
step:1158/1750 train_loss:3.3306 train_time:180326ms step_avg:157.08ms
step:1159/1750 train_loss:3.4073 train_time:180486ms step_avg:157.08ms
step:1160/1750 train_loss:3.4020 train_time:180647ms step_avg:157.08ms
step:1161/1750 train_loss:3.1626 train_time:180810ms step_avg:157.09ms
step:1162/1750 train_loss:3.4922 train_time:180970ms step_avg:157.09ms
step:1163/1750 train_loss:3.4609 train_time:181133ms step_avg:157.10ms
step:1164/1750 train_loss:3.3622 train_time:181294ms step_avg:157.10ms
step:1165/1750 train_loss:3.3182 train_time:181452ms step_avg:157.10ms
step:1166/1750 train_loss:3.4551 train_time:181615ms step_avg:157.11ms
step:1167/1750 train_loss:3.4753 train_time:181777ms step_avg:157.11ms
step:1168/1750 train_loss:3.7923 train_time:181941ms step_avg:157.12ms
step:1169/1750 train_loss:3.4424 train_time:182106ms step_avg:157.12ms
step:1170/1750 train_loss:3.4606 train_time:182268ms step_avg:157.13ms
step:1171/1750 train_loss:3.3807 train_time:182429ms step_avg:157.13ms
step:1172/1750 train_loss:3.4908 train_time:182592ms step_avg:157.14ms
step:1173/1750 train_loss:3.6122 train_time:182760ms step_avg:157.15ms
step:1174/1750 train_loss:3.4522 train_time:182930ms step_avg:157.16ms
step:1175/1750 train_loss:3.4449 train_time:183096ms step_avg:157.16ms
step:1176/1750 train_loss:3.4958 train_time:183263ms step_avg:157.17ms
step:1177/1750 train_loss:3.5204 train_time:183431ms step_avg:157.18ms
step:1178/1750 train_loss:3.5733 train_time:183594ms step_avg:157.19ms
step:1179/1750 train_loss:3.4707 train_time:183755ms step_avg:157.19ms
step:1180/1750 train_loss:3.4263 train_time:183928ms step_avg:157.20ms
step:1181/1750 train_loss:3.4092 train_time:184089ms step_avg:157.21ms
step:1182/1750 train_loss:3.4522 train_time:184251ms step_avg:157.21ms
step:1183/1750 train_loss:3.4030 train_time:184416ms step_avg:157.22ms
step:1184/1750 train_loss:3.5793 train_time:184581ms step_avg:157.22ms
step:1185/1750 train_loss:3.6156 train_time:184747ms step_avg:157.23ms
step:1186/1750 train_loss:3.4307 train_time:184914ms step_avg:157.24ms
step:1187/1750 train_loss:3.4783 train_time:185086ms step_avg:157.25ms
step:1188/1750 train_loss:3.5145 train_time:185248ms step_avg:157.26ms
step:1189/1750 train_loss:3.3411 train_time:185413ms step_avg:157.26ms
step:1190/1750 train_loss:3.5151 train_time:185577ms step_avg:157.27ms
step:1191/1750 train_loss:3.6534 train_time:185743ms step_avg:157.28ms
step:1192/1750 train_loss:3.4666 train_time:185904ms step_avg:157.28ms
step:1193/1750 train_loss:3.3440 train_time:186067ms step_avg:157.28ms
step:1194/1750 train_loss:3.6306 train_time:186231ms step_avg:157.29ms
step:1195/1750 train_loss:3.4468 train_time:186399ms step_avg:157.30ms
step:1196/1750 train_loss:3.4567 train_time:186570ms step_avg:157.31ms
step:1197/1750 train_loss:3.3592 train_time:186737ms step_avg:157.32ms
step:1198/1750 train_loss:3.3648 train_time:186909ms step_avg:157.33ms
step:1199/1750 train_loss:3.4102 train_time:187072ms step_avg:157.34ms
step:1200/1750 train_loss:3.5110 train_time:187235ms step_avg:157.34ms
step:1201/1750 train_loss:3.5579 train_time:187399ms step_avg:157.35ms
step:1202/1750 train_loss:3.6375 train_time:187572ms step_avg:157.36ms
step:1203/1750 train_loss:3.4743 train_time:187736ms step_avg:157.36ms
step:1204/1750 train_loss:3.3837 train_time:187904ms step_avg:157.37ms
step:1205/1750 train_loss:3.4983 train_time:188066ms step_avg:157.38ms
step:1206/1750 train_loss:3.5471 train_time:188230ms step_avg:157.38ms
step:1207/1750 train_loss:3.5911 train_time:188394ms step_avg:157.39ms
step:1208/1750 train_loss:3.4704 train_time:188554ms step_avg:157.39ms
step:1209/1750 train_loss:3.3054 train_time:188719ms step_avg:157.40ms
step:1210/1750 train_loss:3.3726 train_time:188883ms step_avg:157.40ms
step:1211/1750 train_loss:3.4723 train_time:189047ms step_avg:157.41ms
step:1212/1750 train_loss:3.4664 train_time:189211ms step_avg:157.41ms
step:1213/1750 train_loss:3.4817 train_time:189374ms step_avg:157.42ms
step:1214/1750 train_loss:3.3397 train_time:189540ms step_avg:157.43ms
step:1215/1750 train_loss:3.4611 train_time:189704ms step_avg:157.43ms
step:1216/1750 train_loss:3.4010 train_time:189866ms step_avg:157.43ms
step:1217/1750 train_loss:3.3975 train_time:190031ms step_avg:157.44ms
step:1218/1750 train_loss:3.4863 train_time:190194ms step_avg:157.45ms
step:1219/1750 train_loss:3.3344 train_time:190363ms step_avg:157.45ms
step:1220/1750 train_loss:3.5422 train_time:190525ms step_avg:157.46ms
step:1221/1750 train_loss:3.5791 train_time:190686ms step_avg:157.46ms
step:1222/1750 train_loss:3.5193 train_time:190847ms step_avg:157.46ms
step:1223/1750 train_loss:3.3568 train_time:191011ms step_avg:157.47ms
step:1224/1750 train_loss:3.3268 train_time:191177ms step_avg:157.48ms
step:1225/1750 train_loss:3.4424 train_time:191339ms step_avg:157.48ms
step:1226/1750 train_loss:3.3937 train_time:191504ms step_avg:157.49ms
step:1227/1750 train_loss:3.3368 train_time:191667ms step_avg:157.49ms
step:1228/1750 train_loss:3.5222 train_time:191827ms step_avg:157.49ms
step:1229/1750 train_loss:3.4372 train_time:191991ms step_avg:157.50ms
step:1230/1750 train_loss:3.4666 train_time:192161ms step_avg:157.51ms
step:1231/1750 train_loss:3.6480 train_time:192326ms step_avg:157.52ms
step:1232/1750 train_loss:3.5637 train_time:192492ms step_avg:157.52ms
step:1233/1750 train_loss:3.4981 train_time:192655ms step_avg:157.53ms
step:1234/1750 train_loss:3.6553 train_time:192820ms step_avg:157.53ms
step:1235/1750 train_loss:3.3984 train_time:192984ms step_avg:157.54ms
step:1236/1750 train_loss:3.3555 train_time:193145ms step_avg:157.54ms
step:1237/1750 train_loss:3.3387 train_time:193310ms step_avg:157.55ms
step:1238/1750 train_loss:3.3588 train_time:193481ms step_avg:157.56ms
step:1239/1750 train_loss:3.4009 train_time:193643ms step_avg:157.56ms
step:1240/1750 train_loss:3.4450 train_time:193806ms step_avg:157.57ms
step:1241/1750 train_loss:3.4957 train_time:193970ms step_avg:157.57ms
step:1242/1750 train_loss:3.3704 train_time:194132ms step_avg:157.57ms
step:1243/1750 train_loss:3.4772 train_time:194297ms step_avg:157.58ms
step:1244/1750 train_loss:3.4789 train_time:194458ms step_avg:157.58ms
step:1245/1750 train_loss:3.4927 train_time:194622ms step_avg:157.59ms
step:1246/1750 train_loss:3.3081 train_time:194784ms step_avg:157.59ms
step:1247/1750 train_loss:3.4480 train_time:194945ms step_avg:157.59ms
step:1248/1750 train_loss:3.5110 train_time:195107ms step_avg:157.60ms
step:1249/1750 train_loss:3.4896 train_time:195268ms step_avg:157.60ms
step:1250/1750 train_loss:3.3769 train_time:195429ms step_avg:157.60ms
step:1250/1750 val_loss:3.4267 train_time:195474ms step_avg:157.64ms
step:1251/1750 train_loss:3.5692 train_time:195594ms step_avg:157.61ms
step:1252/1750 train_loss:3.4457 train_time:195754ms step_avg:157.61ms
step:1253/1750 train_loss:3.3757 train_time:195916ms step_avg:157.62ms
step:1254/1750 train_loss:3.4819 train_time:196080ms step_avg:157.62ms
step:1255/1750 train_loss:3.5884 train_time:196250ms step_avg:157.63ms
step:1256/1750 train_loss:3.3823 train_time:196415ms step_avg:157.64ms
step:1257/1750 train_loss:3.4415 train_time:196577ms step_avg:157.64ms
step:1258/1750 train_loss:3.4247 train_time:196743ms step_avg:157.65ms
step:1259/1750 train_loss:3.4061 train_time:196905ms step_avg:157.65ms
step:1260/1750 train_loss:3.2797 train_time:197064ms step_avg:157.65ms
step:1261/1750 train_loss:3.3679 train_time:197229ms step_avg:157.66ms
step:1262/1750 train_loss:3.3962 train_time:197394ms step_avg:157.66ms
step:1263/1750 train_loss:3.3025 train_time:197558ms step_avg:157.67ms
step:1264/1750 train_loss:3.5156 train_time:197720ms step_avg:157.67ms
step:1265/1750 train_loss:3.5023 train_time:197881ms step_avg:157.67ms
step:1266/1750 train_loss:3.5087 train_time:198044ms step_avg:157.68ms
step:1267/1750 train_loss:3.4374 train_time:198208ms step_avg:157.68ms
step:1268/1750 train_loss:3.4741 train_time:198372ms step_avg:157.69ms
step:1269/1750 train_loss:3.3242 train_time:198540ms step_avg:157.70ms
step:1270/1750 train_loss:3.1677 train_time:198700ms step_avg:157.70ms
step:1271/1750 train_loss:3.4709 train_time:198864ms step_avg:157.70ms
step:1272/1750 train_loss:3.4285 train_time:199023ms step_avg:157.70ms
step:1273/1750 train_loss:3.4642 train_time:199184ms step_avg:157.71ms
step:1274/1750 train_loss:3.4252 train_time:199346ms step_avg:157.71ms
step:1275/1750 train_loss:3.5045 train_time:199510ms step_avg:157.72ms
step:1276/1750 train_loss:3.5399 train_time:199670ms step_avg:157.72ms
step:1277/1750 train_loss:3.4806 train_time:199837ms step_avg:157.72ms
step:1278/1750 train_loss:3.4731 train_time:199996ms step_avg:157.73ms
step:1279/1750 train_loss:3.3254 train_time:200161ms step_avg:157.73ms
step:1280/1750 train_loss:3.4421 train_time:200330ms step_avg:157.74ms
step:1281/1750 train_loss:3.4950 train_time:200492ms step_avg:157.74ms
step:1282/1750 train_loss:3.5375 train_time:200652ms step_avg:157.75ms
step:1283/1750 train_loss:3.4084 train_time:200817ms step_avg:157.75ms
step:1284/1750 train_loss:3.4373 train_time:200979ms step_avg:157.75ms
step:1285/1750 train_loss:3.4306 train_time:201141ms step_avg:157.76ms
step:1286/1750 train_loss:3.4000 train_time:201303ms step_avg:157.76ms
step:1287/1750 train_loss:3.5625 train_time:201467ms step_avg:157.77ms
step:1288/1750 train_loss:3.3719 train_time:201632ms step_avg:157.77ms
step:1289/1750 train_loss:3.4553 train_time:201801ms step_avg:157.78ms
step:1290/1750 train_loss:3.5266 train_time:201968ms step_avg:157.79ms
step:1291/1750 train_loss:3.4542 train_time:202133ms step_avg:157.79ms
step:1292/1750 train_loss:3.5450 train_time:202298ms step_avg:157.80ms
step:1293/1750 train_loss:3.5884 train_time:202462ms step_avg:157.80ms
step:1294/1750 train_loss:3.5383 train_time:202626ms step_avg:157.81ms
step:1295/1750 train_loss:3.3555 train_time:202787ms step_avg:157.81ms
step:1296/1750 train_loss:3.4485 train_time:202952ms step_avg:157.82ms
step:1297/1750 train_loss:3.3458 train_time:203116ms step_avg:157.82ms
step:1298/1750 train_loss:3.3501 train_time:203280ms step_avg:157.83ms
step:1299/1750 train_loss:3.4599 train_time:203444ms step_avg:157.83ms
step:1300/1750 train_loss:3.4773 train_time:203604ms step_avg:157.83ms
step:1301/1750 train_loss:3.4762 train_time:203766ms step_avg:157.84ms
step:1302/1750 train_loss:3.6416 train_time:203935ms step_avg:157.84ms
step:1303/1750 train_loss:3.3734 train_time:204103ms step_avg:157.85ms
step:1304/1750 train_loss:3.5839 train_time:204267ms step_avg:157.86ms
step:1305/1750 train_loss:3.3368 train_time:204428ms step_avg:157.86ms
step:1306/1750 train_loss:3.5132 train_time:204596ms step_avg:157.87ms
step:1307/1750 train_loss:3.5301 train_time:204758ms step_avg:157.87ms
step:1308/1750 train_loss:3.3640 train_time:204922ms step_avg:157.88ms
step:1309/1750 train_loss:3.3737 train_time:205088ms step_avg:157.88ms
step:1310/1750 train_loss:3.3832 train_time:205249ms step_avg:157.88ms
step:1311/1750 train_loss:3.3627 train_time:205412ms step_avg:157.89ms
step:1312/1750 train_loss:3.4498 train_time:205577ms step_avg:157.89ms
step:1313/1750 train_loss:3.4118 train_time:205742ms step_avg:157.90ms
step:1314/1750 train_loss:3.1102 train_time:205908ms step_avg:157.90ms
step:1315/1750 train_loss:3.3456 train_time:206070ms step_avg:157.91ms
step:1316/1750 train_loss:3.4624 train_time:206232ms step_avg:157.91ms
step:1317/1750 train_loss:3.4865 train_time:206396ms step_avg:157.92ms
step:1318/1750 train_loss:3.3611 train_time:206564ms step_avg:157.92ms
step:1319/1750 train_loss:3.4953 train_time:206728ms step_avg:157.93ms
step:1320/1750 train_loss:3.5258 train_time:206895ms step_avg:157.93ms
step:1321/1750 train_loss:3.4318 train_time:207060ms step_avg:157.94ms
step:1322/1750 train_loss:3.3891 train_time:207350ms step_avg:158.04ms
step:1323/1750 train_loss:3.3930 train_time:207523ms step_avg:158.05ms
step:1324/1750 train_loss:3.5011 train_time:207688ms step_avg:158.06ms
step:1325/1750 train_loss:3.5577 train_time:207858ms step_avg:158.07ms
step:1326/1750 train_loss:3.2883 train_time:208025ms step_avg:158.07ms
step:1327/1750 train_loss:3.2337 train_time:208186ms step_avg:158.08ms
step:1328/1750 train_loss:3.5595 train_time:208351ms step_avg:158.08ms
step:1329/1750 train_loss:3.3653 train_time:208657ms step_avg:158.19ms
step:1330/1750 train_loss:3.4957 train_time:208823ms step_avg:158.20ms
step:1331/1750 train_loss:3.4038 train_time:208983ms step_avg:158.20ms
step:1332/1750 train_loss:3.8129 train_time:209149ms step_avg:158.21ms
step:1333/1750 train_loss:3.5391 train_time:209315ms step_avg:158.21ms
step:1334/1750 train_loss:3.4398 train_time:209478ms step_avg:158.22ms
step:1335/1750 train_loss:3.3664 train_time:209643ms step_avg:158.22ms
step:1336/1750 train_loss:3.3625 train_time:209814ms step_avg:158.23ms
step:1337/1750 train_loss:3.6167 train_time:209981ms step_avg:158.24ms
step:1338/1750 train_loss:3.5878 train_time:210144ms step_avg:158.24ms
step:1339/1750 train_loss:3.4082 train_time:210311ms step_avg:158.25ms
step:1340/1750 train_loss:3.3501 train_time:210475ms step_avg:158.25ms
step:1341/1750 train_loss:3.6607 train_time:210637ms step_avg:158.25ms
step:1342/1750 train_loss:3.4229 train_time:210805ms step_avg:158.26ms
step:1343/1750 train_loss:3.4310 train_time:210967ms step_avg:158.26ms
step:1344/1750 train_loss:3.4868 train_time:211133ms step_avg:158.27ms
step:1345/1750 train_loss:3.4505 train_time:211300ms step_avg:158.28ms
step:1346/1750 train_loss:3.3621 train_time:211464ms step_avg:158.28ms
step:1347/1750 train_loss:3.3383 train_time:211626ms step_avg:158.28ms
step:1348/1750 train_loss:3.4106 train_time:211788ms step_avg:158.29ms
step:1349/1750 train_loss:3.3429 train_time:211950ms step_avg:158.29ms
step:1350/1750 train_loss:3.4538 train_time:212117ms step_avg:158.30ms
step:1351/1750 train_loss:3.3149 train_time:212279ms step_avg:158.30ms
step:1352/1750 train_loss:3.3711 train_time:212443ms step_avg:158.30ms
step:1353/1750 train_loss:3.4801 train_time:212610ms step_avg:158.31ms
step:1354/1750 train_loss:3.3205 train_time:212775ms step_avg:158.31ms
step:1355/1750 train_loss:3.2551 train_time:212937ms step_avg:158.32ms
step:1356/1750 train_loss:3.5821 train_time:213101ms step_avg:158.32ms
step:1357/1750 train_loss:3.4962 train_time:213267ms step_avg:158.33ms
step:1358/1750 train_loss:3.2434 train_time:213431ms step_avg:158.33ms
step:1359/1750 train_loss:3.5127 train_time:213596ms step_avg:158.34ms
step:1360/1750 train_loss:3.4200 train_time:213762ms step_avg:158.34ms
step:1361/1750 train_loss:3.2163 train_time:213931ms step_avg:158.35ms
step:1362/1750 train_loss:3.4582 train_time:214095ms step_avg:158.35ms
step:1363/1750 train_loss:3.3436 train_time:214264ms step_avg:158.36ms
step:1364/1750 train_loss:3.3779 train_time:214425ms step_avg:158.36ms
step:1365/1750 train_loss:3.3779 train_time:214585ms step_avg:158.37ms
step:1366/1750 train_loss:3.4947 train_time:214749ms step_avg:158.37ms
step:1367/1750 train_loss:3.4587 train_time:214914ms step_avg:158.37ms
step:1368/1750 train_loss:3.4138 train_time:215079ms step_avg:158.38ms
step:1369/1750 train_loss:3.3305 train_time:215249ms step_avg:158.39ms
step:1370/1750 train_loss:3.6741 train_time:215415ms step_avg:158.39ms
step:1371/1750 train_loss:3.3874 train_time:215579ms step_avg:158.40ms
step:1372/1750 train_loss:3.4334 train_time:215746ms step_avg:158.40ms
step:1373/1750 train_loss:3.4330 train_time:215909ms step_avg:158.41ms
step:1374/1750 train_loss:3.2261 train_time:216075ms step_avg:158.41ms
step:1375/1750 train_loss:3.6158 train_time:216239ms step_avg:158.42ms
step:1375/1750 val_loss:3.3786 train_time:216280ms step_avg:158.45ms
step:1376/1750 train_loss:3.4063 train_time:216402ms step_avg:158.42ms
step:1377/1750 train_loss:3.5439 train_time:216566ms step_avg:158.42ms
step:1378/1750 train_loss:3.5410 train_time:216728ms step_avg:158.43ms
step:1379/1750 train_loss:3.1894 train_time:216894ms step_avg:158.43ms
step:1380/1750 train_loss:3.3850 train_time:217057ms step_avg:158.44ms
step:1381/1750 train_loss:3.7776 train_time:217224ms step_avg:158.44ms
step:1382/1750 train_loss:3.2833 train_time:217386ms step_avg:158.44ms
step:1383/1750 train_loss:3.4618 train_time:217552ms step_avg:158.45ms
step:1384/1750 train_loss:3.5452 train_time:217717ms step_avg:158.45ms
step:1385/1750 train_loss:3.4698 train_time:217877ms step_avg:158.46ms
step:1386/1750 train_loss:3.4149 train_time:218041ms step_avg:158.46ms
step:1387/1750 train_loss:3.2617 train_time:218203ms step_avg:158.46ms
step:1388/1750 train_loss:3.4077 train_time:218366ms step_avg:158.47ms
step:1389/1750 train_loss:3.3818 train_time:218530ms step_avg:158.47ms
step:1390/1750 train_loss:3.6364 train_time:218692ms step_avg:158.47ms
step:1391/1750 train_loss:3.3529 train_time:218856ms step_avg:158.48ms
step:1392/1750 train_loss:3.3572 train_time:219020ms step_avg:158.48ms
step:1393/1750 train_loss:3.3112 train_time:219183ms step_avg:158.48ms
step:1394/1750 train_loss:3.5748 train_time:219346ms step_avg:158.49ms
step:1395/1750 train_loss:3.4626 train_time:219507ms step_avg:158.49ms
step:1396/1750 train_loss:3.4718 train_time:219668ms step_avg:158.49ms
step:1397/1750 train_loss:3.3678 train_time:219829ms step_avg:158.49ms
step:1398/1750 train_loss:3.3218 train_time:219990ms step_avg:158.49ms
step:1399/1750 train_loss:3.3927 train_time:220154ms step_avg:158.50ms
step:1400/1750 train_loss:3.3811 train_time:220318ms step_avg:158.50ms
step:1401/1750 train_loss:3.4087 train_time:220480ms step_avg:158.50ms
step:1402/1750 train_loss:3.3583 train_time:220644ms step_avg:158.51ms
step:1403/1750 train_loss:3.5638 train_time:220811ms step_avg:158.51ms
step:1404/1750 train_loss:3.3415 train_time:220973ms step_avg:158.52ms
step:1405/1750 train_loss:3.3728 train_time:221138ms step_avg:158.52ms
step:1406/1750 train_loss:3.3764 train_time:221303ms step_avg:158.53ms
step:1407/1750 train_loss:3.2405 train_time:221466ms step_avg:158.53ms
step:1408/1750 train_loss:3.3687 train_time:221628ms step_avg:158.53ms
step:1409/1750 train_loss:3.3605 train_time:221798ms step_avg:158.54ms
step:1410/1750 train_loss:3.3494 train_time:221960ms step_avg:158.54ms
step:1411/1750 train_loss:3.4296 train_time:222120ms step_avg:158.54ms
step:1412/1750 train_loss:3.3927 train_time:222283ms step_avg:158.55ms
step:1413/1750 train_loss:3.4217 train_time:222445ms step_avg:158.55ms
step:1414/1750 train_loss:3.4013 train_time:222609ms step_avg:158.55ms
step:1415/1750 train_loss:3.4722 train_time:222777ms step_avg:158.56ms
step:1416/1750 train_loss:3.2900 train_time:222945ms step_avg:158.57ms
step:1417/1750 train_loss:3.3462 train_time:223109ms step_avg:158.57ms
step:1418/1750 train_loss:3.4545 train_time:223272ms step_avg:158.57ms
step:1419/1750 train_loss:3.4082 train_time:223438ms step_avg:158.58ms
step:1420/1750 train_loss:3.4260 train_time:223604ms step_avg:158.58ms
step:1421/1750 train_loss:3.4413 train_time:223770ms step_avg:158.59ms
step:1422/1750 train_loss:3.4024 train_time:223935ms step_avg:158.59ms
step:1423/1750 train_loss:3.3810 train_time:224098ms step_avg:158.60ms
step:1424/1750 train_loss:3.3904 train_time:224266ms step_avg:158.60ms
step:1425/1750 train_loss:3.2488 train_time:224437ms step_avg:158.61ms
step:1426/1750 train_loss:3.3913 train_time:224599ms step_avg:158.62ms
step:1427/1750 train_loss:3.3364 train_time:224766ms step_avg:158.62ms
step:1428/1750 train_loss:3.4436 train_time:224929ms step_avg:158.62ms
step:1429/1750 train_loss:3.4208 train_time:225089ms step_avg:158.63ms
step:1430/1750 train_loss:3.3208 train_time:225257ms step_avg:158.63ms
step:1431/1750 train_loss:3.3812 train_time:225422ms step_avg:158.64ms
step:1432/1750 train_loss:3.4014 train_time:225588ms step_avg:158.64ms
step:1433/1750 train_loss:3.2007 train_time:225758ms step_avg:158.65ms
step:1434/1750 train_loss:3.3506 train_time:225926ms step_avg:158.66ms
step:1435/1750 train_loss:3.1829 train_time:226090ms step_avg:158.66ms
step:1436/1750 train_loss:3.2874 train_time:226257ms step_avg:158.67ms
step:1437/1750 train_loss:3.4693 train_time:226419ms step_avg:158.67ms
step:1438/1750 train_loss:3.4393 train_time:226579ms step_avg:158.67ms
step:1439/1750 train_loss:3.3801 train_time:226745ms step_avg:158.67ms
step:1440/1750 train_loss:3.2445 train_time:226907ms step_avg:158.68ms
step:1441/1750 train_loss:3.4103 train_time:227073ms step_avg:158.68ms
step:1442/1750 train_loss:3.4462 train_time:227241ms step_avg:158.69ms
step:1443/1750 train_loss:3.5396 train_time:227415ms step_avg:158.70ms
step:1444/1750 train_loss:3.5130 train_time:227580ms step_avg:158.70ms
step:1445/1750 train_loss:3.3990 train_time:227743ms step_avg:158.71ms
step:1446/1750 train_loss:3.2650 train_time:227909ms step_avg:158.71ms
step:1447/1750 train_loss:3.3583 train_time:228078ms step_avg:158.72ms
step:1448/1750 train_loss:3.3530 train_time:228242ms step_avg:158.72ms
step:1449/1750 train_loss:3.4656 train_time:228407ms step_avg:158.73ms
step:1450/1750 train_loss:3.4512 train_time:228572ms step_avg:158.73ms
step:1451/1750 train_loss:3.2714 train_time:228736ms step_avg:158.73ms
step:1452/1750 train_loss:3.3926 train_time:228902ms step_avg:158.74ms
step:1453/1750 train_loss:3.3171 train_time:229064ms step_avg:158.74ms
step:1454/1750 train_loss:3.3466 train_time:229229ms step_avg:158.75ms
step:1455/1750 train_loss:3.3904 train_time:229400ms step_avg:158.75ms
step:1456/1750 train_loss:3.3397 train_time:229564ms step_avg:158.76ms
step:1457/1750 train_loss:3.2158 train_time:229726ms step_avg:158.76ms
step:1458/1750 train_loss:3.4799 train_time:229891ms step_avg:158.76ms
step:1459/1750 train_loss:3.3303 train_time:230060ms step_avg:158.77ms
step:1460/1750 train_loss:3.3756 train_time:230224ms step_avg:158.78ms
step:1461/1750 train_loss:3.4894 train_time:230392ms step_avg:158.78ms
step:1462/1750 train_loss:3.3224 train_time:230558ms step_avg:158.79ms
step:1463/1750 train_loss:3.5254 train_time:230726ms step_avg:158.79ms
step:1464/1750 train_loss:3.4174 train_time:230892ms step_avg:158.80ms
step:1465/1750 train_loss:3.4207 train_time:231058ms step_avg:158.80ms
step:1466/1750 train_loss:3.3425 train_time:231220ms step_avg:158.80ms
step:1467/1750 train_loss:3.4493 train_time:231386ms step_avg:158.81ms
step:1468/1750 train_loss:3.3467 train_time:231550ms step_avg:158.81ms
step:1469/1750 train_loss:3.3290 train_time:231718ms step_avg:158.82ms
step:1470/1750 train_loss:3.3929 train_time:231888ms step_avg:158.83ms
step:1471/1750 train_loss:3.3125 train_time:232060ms step_avg:158.84ms
step:1472/1750 train_loss:3.2969 train_time:232230ms step_avg:158.84ms
step:1473/1750 train_loss:3.4959 train_time:232392ms step_avg:158.85ms
step:1474/1750 train_loss:3.3712 train_time:232561ms step_avg:158.85ms
step:1475/1750 train_loss:3.2023 train_time:232730ms step_avg:158.86ms
step:1476/1750 train_loss:3.3216 train_time:232893ms step_avg:158.86ms
step:1477/1750 train_loss:3.2995 train_time:233063ms step_avg:158.87ms
step:1478/1750 train_loss:3.3646 train_time:233232ms step_avg:158.88ms
step:1479/1750 train_loss:3.4487 train_time:233398ms step_avg:158.88ms
step:1480/1750 train_loss:3.3301 train_time:233561ms step_avg:158.88ms
step:1481/1750 train_loss:3.5073 train_time:233727ms step_avg:158.89ms
step:1482/1750 train_loss:3.4260 train_time:233900ms step_avg:158.90ms
step:1483/1750 train_loss:3.3369 train_time:234074ms step_avg:158.91ms
step:1484/1750 train_loss:3.3203 train_time:234244ms step_avg:158.92ms
step:1485/1750 train_loss:3.3379 train_time:234409ms step_avg:158.92ms
step:1486/1750 train_loss:3.2785 train_time:234578ms step_avg:158.93ms
step:1487/1750 train_loss:3.3933 train_time:234743ms step_avg:158.93ms
step:1488/1750 train_loss:3.2925 train_time:234911ms step_avg:158.94ms
step:1489/1750 train_loss:3.3778 train_time:235076ms step_avg:158.94ms
step:1490/1750 train_loss:3.3097 train_time:235240ms step_avg:158.95ms
step:1491/1750 train_loss:3.2187 train_time:235407ms step_avg:158.95ms
step:1492/1750 train_loss:3.3162 train_time:235571ms step_avg:158.95ms
step:1493/1750 train_loss:3.4921 train_time:235735ms step_avg:158.96ms
step:1494/1750 train_loss:3.3517 train_time:235897ms step_avg:158.96ms
step:1495/1750 train_loss:3.0873 train_time:236065ms step_avg:158.97ms
step:1496/1750 train_loss:3.4110 train_time:236230ms step_avg:158.97ms
step:1497/1750 train_loss:3.3651 train_time:236397ms step_avg:158.98ms
step:1498/1750 train_loss:3.3997 train_time:236564ms step_avg:158.98ms
step:1499/1750 train_loss:3.3713 train_time:236734ms step_avg:158.99ms
step:1500/1750 train_loss:3.3472 train_time:236906ms step_avg:159.00ms
step:1500/1750 val_loss:3.3345 train_time:236950ms step_avg:159.03ms
step:1501/1750 train_loss:3.1429 train_time:237078ms step_avg:159.01ms
step:1502/1750 train_loss:3.4147 train_time:237252ms step_avg:159.02ms
step:1503/1750 train_loss:3.2970 train_time:237416ms step_avg:159.02ms
step:1504/1750 train_loss:3.3032 train_time:237581ms step_avg:159.02ms
step:1505/1750 train_loss:3.2662 train_time:237746ms step_avg:159.03ms
step:1506/1750 train_loss:3.3307 train_time:237913ms step_avg:159.03ms
step:1507/1750 train_loss:3.2218 train_time:238087ms step_avg:159.04ms
step:1508/1750 train_loss:3.5363 train_time:238254ms step_avg:159.05ms
step:1509/1750 train_loss:3.3294 train_time:238417ms step_avg:159.05ms
step:1510/1750 train_loss:3.3277 train_time:238581ms step_avg:159.05ms
step:1511/1750 train_loss:3.4621 train_time:238870ms step_avg:159.14ms
step:1512/1750 train_loss:3.4730 train_time:239040ms step_avg:159.15ms
step:1513/1750 train_loss:3.3233 train_time:239207ms step_avg:159.15ms
step:1514/1750 train_loss:3.1416 train_time:239374ms step_avg:159.16ms
step:1515/1750 train_loss:3.2923 train_time:239538ms step_avg:159.16ms
step:1516/1750 train_loss:3.3008 train_time:239707ms step_avg:159.17ms
step:1517/1750 train_loss:3.3537 train_time:239871ms step_avg:159.17ms
step:1518/1750 train_loss:3.2577 train_time:240037ms step_avg:159.18ms
step:1519/1750 train_loss:3.5556 train_time:240372ms step_avg:159.29ms
step:1520/1750 train_loss:3.1801 train_time:240541ms step_avg:159.30ms
step:1521/1750 train_loss:3.2622 train_time:240704ms step_avg:159.30ms
step:1522/1750 train_loss:3.4024 train_time:240870ms step_avg:159.31ms
step:1523/1750 train_loss:3.2761 train_time:241033ms step_avg:159.31ms
step:1524/1750 train_loss:3.3953 train_time:241198ms step_avg:159.31ms
step:1525/1750 train_loss:3.3787 train_time:241368ms step_avg:159.32ms
step:1526/1750 train_loss:3.3216 train_time:241538ms step_avg:159.33ms
step:1527/1750 train_loss:3.3360 train_time:241703ms step_avg:159.33ms
step:1528/1750 train_loss:3.4590 train_time:241868ms step_avg:159.33ms
step:1529/1750 train_loss:3.4614 train_time:242030ms step_avg:159.33ms
step:1530/1750 train_loss:3.2848 train_time:242192ms step_avg:159.34ms
step:1531/1750 train_loss:3.2445 train_time:242357ms step_avg:159.34ms
step:1532/1750 train_loss:3.3995 train_time:242522ms step_avg:159.34ms
step:1533/1750 train_loss:3.3249 train_time:242690ms step_avg:159.35ms
step:1534/1750 train_loss:3.3281 train_time:242859ms step_avg:159.36ms
step:1535/1750 train_loss:3.3296 train_time:243026ms step_avg:159.36ms
step:1536/1750 train_loss:3.2741 train_time:243196ms step_avg:159.37ms
step:1537/1750 train_loss:3.3194 train_time:243359ms step_avg:159.37ms
step:1538/1750 train_loss:3.4764 train_time:243529ms step_avg:159.38ms
step:1539/1750 train_loss:3.4447 train_time:243699ms step_avg:159.38ms
step:1540/1750 train_loss:3.3255 train_time:243862ms step_avg:159.39ms
step:1541/1750 train_loss:3.2802 train_time:244025ms step_avg:159.39ms
step:1542/1750 train_loss:3.2982 train_time:244195ms step_avg:159.40ms
step:1543/1750 train_loss:3.1977 train_time:244360ms step_avg:159.40ms
step:1544/1750 train_loss:3.3397 train_time:244522ms step_avg:159.40ms
step:1545/1750 train_loss:3.3150 train_time:244688ms step_avg:159.41ms
step:1546/1750 train_loss:3.3074 train_time:244858ms step_avg:159.41ms
step:1547/1750 train_loss:3.2698 train_time:245026ms step_avg:159.42ms
step:1548/1750 train_loss:3.3117 train_time:245196ms step_avg:159.43ms
step:1549/1750 train_loss:3.3821 train_time:245360ms step_avg:159.43ms
step:1550/1750 train_loss:3.3409 train_time:245524ms step_avg:159.43ms
step:1551/1750 train_loss:3.2529 train_time:245691ms step_avg:159.44ms
step:1552/1750 train_loss:3.2660 train_time:245857ms step_avg:159.44ms
step:1553/1750 train_loss:3.2745 train_time:246021ms step_avg:159.44ms
step:1554/1750 train_loss:3.4065 train_time:246184ms step_avg:159.45ms
step:1555/1750 train_loss:3.3833 train_time:246352ms step_avg:159.45ms
step:1556/1750 train_loss:3.3282 train_time:246514ms step_avg:159.45ms
step:1557/1750 train_loss:3.3709 train_time:246675ms step_avg:159.45ms
step:1558/1750 train_loss:3.3062 train_time:246841ms step_avg:159.46ms
step:1559/1750 train_loss:3.1793 train_time:247017ms step_avg:159.47ms
step:1560/1750 train_loss:3.4770 train_time:247180ms step_avg:159.47ms
step:1561/1750 train_loss:3.2728 train_time:247346ms step_avg:159.48ms
step:1562/1750 train_loss:3.2568 train_time:247510ms step_avg:159.48ms
step:1563/1750 train_loss:3.3700 train_time:247677ms step_avg:159.48ms
step:1564/1750 train_loss:3.2002 train_time:247846ms step_avg:159.49ms
step:1565/1750 train_loss:3.2125 train_time:248017ms step_avg:159.50ms
step:1566/1750 train_loss:3.4094 train_time:248180ms step_avg:159.50ms
step:1567/1750 train_loss:3.2775 train_time:248344ms step_avg:159.50ms
step:1568/1750 train_loss:3.2862 train_time:248514ms step_avg:159.51ms
step:1569/1750 train_loss:3.3746 train_time:248690ms step_avg:159.52ms
step:1570/1750 train_loss:3.3320 train_time:248860ms step_avg:159.53ms
step:1571/1750 train_loss:3.2059 train_time:249026ms step_avg:159.53ms
step:1572/1750 train_loss:3.2482 train_time:249192ms step_avg:159.53ms
step:1573/1750 train_loss:3.3615 train_time:249360ms step_avg:159.54ms
step:1574/1750 train_loss:3.2118 train_time:249523ms step_avg:159.54ms
step:1575/1750 train_loss:3.3757 train_time:249686ms step_avg:159.54ms
step:1576/1750 train_loss:3.2796 train_time:249853ms step_avg:159.55ms
step:1577/1750 train_loss:3.3308 train_time:250021ms step_avg:159.55ms
step:1578/1750 train_loss:3.3192 train_time:250187ms step_avg:159.56ms
step:1579/1750 train_loss:3.2875 train_time:250357ms step_avg:159.56ms
step:1580/1750 train_loss:3.2496 train_time:250524ms step_avg:159.57ms
step:1581/1750 train_loss:3.4513 train_time:250697ms step_avg:159.58ms
step:1582/1750 train_loss:3.2778 train_time:250870ms step_avg:159.59ms
step:1583/1750 train_loss:3.4247 train_time:251043ms step_avg:159.60ms
step:1584/1750 train_loss:3.2459 train_time:251207ms step_avg:159.60ms
step:1585/1750 train_loss:3.4092 train_time:251377ms step_avg:159.60ms
step:1586/1750 train_loss:3.1937 train_time:251544ms step_avg:159.61ms
step:1587/1750 train_loss:3.3964 train_time:251708ms step_avg:159.61ms
step:1588/1750 train_loss:3.2759 train_time:251876ms step_avg:159.62ms
step:1589/1750 train_loss:3.4388 train_time:252040ms step_avg:159.62ms
step:1590/1750 train_loss:3.2841 train_time:252206ms step_avg:159.62ms
step:1591/1750 train_loss:3.2955 train_time:252374ms step_avg:159.63ms
step:1592/1750 train_loss:3.3671 train_time:252540ms step_avg:159.63ms
step:1593/1750 train_loss:3.3333 train_time:252713ms step_avg:159.64ms
step:1594/1750 train_loss:3.3047 train_time:252878ms step_avg:159.65ms
step:1595/1750 train_loss:3.4581 train_time:253044ms step_avg:159.65ms
step:1596/1750 train_loss:3.1623 train_time:253220ms step_avg:159.66ms
step:1597/1750 train_loss:3.3344 train_time:253390ms step_avg:159.67ms
step:1598/1750 train_loss:3.3844 train_time:253560ms step_avg:159.67ms
step:1599/1750 train_loss:3.4476 train_time:253733ms step_avg:159.68ms
step:1600/1750 train_loss:3.2730 train_time:253901ms step_avg:159.69ms
step:1601/1750 train_loss:3.5762 train_time:254065ms step_avg:159.69ms
step:1602/1750 train_loss:3.4546 train_time:254233ms step_avg:159.69ms
step:1603/1750 train_loss:3.2245 train_time:254406ms step_avg:159.70ms
step:1604/1750 train_loss:3.2645 train_time:254572ms step_avg:159.71ms
step:1605/1750 train_loss:3.1575 train_time:254746ms step_avg:159.72ms
step:1606/1750 train_loss:3.4707 train_time:254923ms step_avg:159.73ms
step:1607/1750 train_loss:3.3022 train_time:255086ms step_avg:159.73ms
step:1608/1750 train_loss:3.3057 train_time:255256ms step_avg:159.73ms
step:1609/1750 train_loss:3.2485 train_time:255425ms step_avg:159.74ms
step:1610/1750 train_loss:3.7545 train_time:255604ms step_avg:159.75ms
step:1611/1750 train_loss:3.5063 train_time:255773ms step_avg:159.76ms
step:1612/1750 train_loss:3.3892 train_time:255946ms step_avg:159.77ms
step:1613/1750 train_loss:3.2666 train_time:256122ms step_avg:159.78ms
step:1614/1750 train_loss:3.2970 train_time:256289ms step_avg:159.78ms
step:1615/1750 train_loss:3.3129 train_time:256459ms step_avg:159.79ms
step:1616/1750 train_loss:3.2844 train_time:256639ms step_avg:159.80ms
step:1617/1750 train_loss:3.3550 train_time:256815ms step_avg:159.81ms
step:1618/1750 train_loss:3.2822 train_time:256979ms step_avg:159.81ms
step:1619/1750 train_loss:3.1874 train_time:257146ms step_avg:159.82ms
step:1620/1750 train_loss:3.4589 train_time:257311ms step_avg:159.82ms
step:1621/1750 train_loss:3.3823 train_time:257482ms step_avg:159.83ms
step:1622/1750 train_loss:3.1567 train_time:257649ms step_avg:159.83ms
step:1623/1750 train_loss:3.2579 train_time:257816ms step_avg:159.84ms
step:1624/1750 train_loss:3.2141 train_time:257980ms step_avg:159.84ms
step:1625/1750 train_loss:3.3201 train_time:258145ms step_avg:159.84ms
step:1625/1750 val_loss:3.2984 train_time:258187ms step_avg:159.87ms
step:1626/1750 train_loss:3.2399 train_time:258311ms step_avg:159.85ms
step:1627/1750 train_loss:3.2375 train_time:258474ms step_avg:159.85ms
step:1628/1750 train_loss:3.3611 train_time:258640ms step_avg:159.85ms
step:1629/1750 train_loss:3.2468 train_time:258806ms step_avg:159.86ms
step:1630/1750 train_loss:3.3185 train_time:258975ms step_avg:159.86ms
step:1631/1750 train_loss:3.1781 train_time:259153ms step_avg:159.87ms
step:1632/1750 train_loss:3.1488 train_time:259317ms step_avg:159.87ms
step:1633/1750 train_loss:3.2978 train_time:259485ms step_avg:159.88ms
step:1634/1750 train_loss:3.3118 train_time:259649ms step_avg:159.88ms
step:1635/1750 train_loss:3.2460 train_time:259821ms step_avg:159.89ms
step:1636/1750 train_loss:3.3264 train_time:259987ms step_avg:159.89ms
step:1637/1750 train_loss:3.3782 train_time:260155ms step_avg:159.90ms
step:1638/1750 train_loss:3.4021 train_time:260325ms step_avg:159.90ms
step:1639/1750 train_loss:3.5698 train_time:260495ms step_avg:159.91ms
step:1640/1750 train_loss:3.3546 train_time:260664ms step_avg:159.92ms
step:1641/1750 train_loss:3.3004 train_time:260833ms step_avg:159.92ms
step:1642/1750 train_loss:3.4098 train_time:260998ms step_avg:159.93ms
step:1643/1750 train_loss:3.2760 train_time:261171ms step_avg:159.93ms
step:1644/1750 train_loss:3.3103 train_time:261335ms step_avg:159.94ms
step:1645/1750 train_loss:3.3150 train_time:261497ms step_avg:159.94ms
step:1646/1750 train_loss:3.0735 train_time:261664ms step_avg:159.94ms
step:1647/1750 train_loss:3.3193 train_time:261831ms step_avg:159.95ms
step:1648/1750 train_loss:3.2163 train_time:261997ms step_avg:159.95ms
step:1649/1750 train_loss:3.2904 train_time:262160ms step_avg:159.95ms
step:1650/1750 train_loss:3.2701 train_time:262326ms step_avg:159.96ms
step:1651/1750 train_loss:3.3456 train_time:262492ms step_avg:159.96ms
step:1652/1750 train_loss:3.2557 train_time:262658ms step_avg:159.96ms
step:1653/1750 train_loss:3.3941 train_time:262828ms step_avg:159.97ms
step:1654/1750 train_loss:3.3811 train_time:262991ms step_avg:159.97ms
step:1655/1750 train_loss:3.1775 train_time:263162ms step_avg:159.98ms
step:1656/1750 train_loss:3.3322 train_time:263337ms step_avg:159.99ms
step:1657/1750 train_loss:3.2506 train_time:263505ms step_avg:159.99ms
step:1658/1750 train_loss:3.2242 train_time:263668ms step_avg:159.99ms
step:1659/1750 train_loss:3.3077 train_time:263834ms step_avg:160.00ms
step:1660/1750 train_loss:3.3455 train_time:263999ms step_avg:160.00ms
step:1661/1750 train_loss:3.2572 train_time:264166ms step_avg:160.00ms
step:1662/1750 train_loss:3.3614 train_time:264332ms step_avg:160.01ms
step:1663/1750 train_loss:3.3445 train_time:264501ms step_avg:160.01ms
step:1664/1750 train_loss:3.4076 train_time:264678ms step_avg:160.02ms
step:1665/1750 train_loss:3.3317 train_time:264845ms step_avg:160.03ms
step:1666/1750 train_loss:3.5002 train_time:265011ms step_avg:160.03ms
step:1667/1750 train_loss:3.2033 train_time:265177ms step_avg:160.03ms
step:1668/1750 train_loss:3.2906 train_time:265347ms step_avg:160.04ms
step:1669/1750 train_loss:3.2122 train_time:265512ms step_avg:160.04ms
step:1670/1750 train_loss:3.2210 train_time:265679ms step_avg:160.05ms
step:1671/1750 train_loss:3.3737 train_time:265847ms step_avg:160.05ms
step:1672/1750 train_loss:3.5737 train_time:266013ms step_avg:160.06ms
step:1673/1750 train_loss:3.2789 train_time:266182ms step_avg:160.06ms
step:1674/1750 train_loss:3.2565 train_time:266350ms step_avg:160.07ms
step:1675/1750 train_loss:3.1279 train_time:266519ms step_avg:160.07ms
step:1676/1750 train_loss:3.3523 train_time:266690ms step_avg:160.08ms
step:1677/1750 train_loss:3.2808 train_time:266857ms step_avg:160.08ms
step:1678/1750 train_loss:3.2984 train_time:267027ms step_avg:160.09ms
step:1679/1750 train_loss:3.3038 train_time:267191ms step_avg:160.09ms
step:1680/1750 train_loss:3.0872 train_time:267366ms step_avg:160.10ms
step:1681/1750 train_loss:3.3000 train_time:267534ms step_avg:160.10ms
step:1682/1750 train_loss:3.2875 train_time:267703ms step_avg:160.11ms
step:1683/1750 train_loss:3.3085 train_time:267870ms step_avg:160.11ms
step:1684/1750 train_loss:3.3417 train_time:268032ms step_avg:160.11ms
step:1685/1750 train_loss:3.2429 train_time:268196ms step_avg:160.12ms
step:1686/1750 train_loss:3.3681 train_time:268365ms step_avg:160.12ms
step:1687/1750 train_loss:3.2435 train_time:268533ms step_avg:160.13ms
step:1688/1750 train_loss:3.3149 train_time:268707ms step_avg:160.14ms
step:1689/1750 train_loss:3.2229 train_time:268877ms step_avg:160.14ms
step:1690/1750 train_loss:3.0754 train_time:269050ms step_avg:160.15ms
step:1691/1750 train_loss:3.3056 train_time:269214ms step_avg:160.15ms
step:1692/1750 train_loss:3.2953 train_time:269378ms step_avg:160.15ms
step:1693/1750 train_loss:3.2189 train_time:269544ms step_avg:160.16ms
step:1694/1750 train_loss:3.6118 train_time:269717ms step_avg:160.16ms
step:1695/1750 train_loss:3.3348 train_time:269888ms step_avg:160.17ms
step:1696/1750 train_loss:3.3380 train_time:270055ms step_avg:160.17ms
step:1697/1750 train_loss:3.2523 train_time:270218ms step_avg:160.18ms
step:1698/1750 train_loss:3.1280 train_time:270387ms step_avg:160.18ms
step:1699/1750 train_loss:3.2325 train_time:270554ms step_avg:160.19ms
step:1700/1750 train_loss:3.2479 train_time:270844ms step_avg:160.26ms
step:1701/1750 train_loss:3.3176 train_time:271015ms step_avg:160.27ms
step:1702/1750 train_loss:3.2428 train_time:271180ms step_avg:160.27ms
step:1703/1750 train_loss:3.4195 train_time:271342ms step_avg:160.27ms
step:1704/1750 train_loss:3.2141 train_time:271509ms step_avg:160.28ms
step:1705/1750 train_loss:3.4360 train_time:271674ms step_avg:160.28ms
step:1706/1750 train_loss:3.2526 train_time:271838ms step_avg:160.28ms
step:1707/1750 train_loss:3.0565 train_time:272008ms step_avg:160.29ms
step:1708/1750 train_loss:3.3904 train_time:272173ms step_avg:160.29ms
step:1709/1750 train_loss:3.2991 train_time:272485ms step_avg:160.38ms
step:1710/1750 train_loss:3.2804 train_time:272657ms step_avg:160.39ms
step:1711/1750 train_loss:3.2869 train_time:272825ms step_avg:160.39ms
step:1712/1750 train_loss:3.3174 train_time:272993ms step_avg:160.40ms
step:1713/1750 train_loss:3.3407 train_time:273162ms step_avg:160.40ms
step:1714/1750 train_loss:3.2286 train_time:273333ms step_avg:160.41ms
step:1715/1750 train_loss:3.2843 train_time:273512ms step_avg:160.42ms
step:1716/1750 train_loss:3.0978 train_time:273676ms step_avg:160.42ms
step:1717/1750 train_loss:3.2523 train_time:273841ms step_avg:160.42ms
step:1718/1750 train_loss:3.2569 train_time:274008ms step_avg:160.43ms
step:1719/1750 train_loss:3.2132 train_time:274176ms step_avg:160.43ms
step:1720/1750 train_loss:3.3767 train_time:274352ms step_avg:160.44ms
step:1721/1750 train_loss:3.1648 train_time:274530ms step_avg:160.45ms
step:1722/1750 train_loss:3.3136 train_time:274696ms step_avg:160.45ms
step:1723/1750 train_loss:3.4040 train_time:274873ms step_avg:160.46ms
step:1724/1750 train_loss:3.2589 train_time:275042ms step_avg:160.47ms
step:1725/1750 train_loss:3.4911 train_time:275215ms step_avg:160.48ms
step:1726/1750 train_loss:3.2605 train_time:275388ms step_avg:160.48ms
step:1727/1750 train_loss:3.3339 train_time:275553ms step_avg:160.48ms
step:1728/1750 train_loss:3.2996 train_time:275719ms step_avg:160.49ms
step:1729/1750 train_loss:3.2765 train_time:275890ms step_avg:160.49ms
step:1730/1750 train_loss:3.6556 train_time:276062ms step_avg:160.50ms
step:1731/1750 train_loss:3.2951 train_time:276226ms step_avg:160.50ms
step:1732/1750 train_loss:3.4304 train_time:276392ms step_avg:160.51ms
step:1733/1750 train_loss:3.2062 train_time:276554ms step_avg:160.51ms
step:1734/1750 train_loss:3.2478 train_time:276722ms step_avg:160.51ms
step:1735/1750 train_loss:3.2698 train_time:276890ms step_avg:160.52ms
step:1736/1750 train_loss:3.2556 train_time:277060ms step_avg:160.52ms
step:1737/1750 train_loss:3.3853 train_time:277232ms step_avg:160.53ms
step:1738/1750 train_loss:3.2207 train_time:277409ms step_avg:160.54ms
step:1739/1750 train_loss:3.2871 train_time:277582ms step_avg:160.54ms
step:1740/1750 train_loss:3.3715 train_time:277753ms step_avg:160.55ms
step:1741/1750 train_loss:3.1647 train_time:277919ms step_avg:160.55ms
step:1742/1750 train_loss:3.0605 train_time:278089ms step_avg:160.56ms
step:1743/1750 train_loss:2.9539 train_time:278263ms step_avg:160.57ms
step:1744/1750 train_loss:3.2929 train_time:278429ms step_avg:160.57ms
step:1745/1750 train_loss:3.3124 train_time:278590ms step_avg:160.57ms
step:1746/1750 train_loss:3.2688 train_time:278755ms step_avg:160.57ms
step:1747/1750 train_loss:3.2991 train_time:278927ms step_avg:160.58ms
step:1748/1750 train_loss:3.4997 train_time:279108ms step_avg:160.59ms
step:1749/1750 train_loss:3.2289 train_time:279276ms step_avg:160.60ms
step:1750/1750 train_loss:3.2830 train_time:279449ms step_avg:160.60ms
step:1750/1750 val_loss:3.2773 train_time:279496ms step_avg:160.63ms