==================================================================================================== 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 # ----------------------------------------------------------------------------- # 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}' \sim 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 = A @ X X = a * X + b * B + c * A @ B 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.inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float() / dim)) self.seq_len_cached = None self.cos_cached = None self.sin_cached = None def forward(self, x): seq_len = x.shape[1] if seq_len != self.seq_len_cached: 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).to(x.device) 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 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 = nn.Linear(self.n_embd, self.n_embd, bias=False) self.c_k = nn.Linear(self.n_embd, self.n_embd, bias=False) self.c_v = nn.Linear(self.n_embd, self.n_embd, bias=False) # output projection self.c_proj = nn.Linear(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) def forward(self, x): 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) 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 = F.scaled_dot_product_attention(q.transpose(1, 2), k.transpose(1, 2), v.transpose(1, 2), is_causal=True) y = y.transpose(1, 2).contiguous().view_as(x) # re-assemble all head outputs side by side y = self.c_proj(y) return y class MLP(nn.Module): def __init__(self, config): super().__init__() self.c_fc = nn.Linear(config.n_embd, 4 * config.n_embd, bias=False) self.c_proj = nn.Linear(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) def forward(self, x): x = x + self.attn(F.rms_norm(x, (x.size(-1),))) x = x + self.mlp(F.rms_norm(x, (x.size(-1),))) return x # ----------------------------------------------------------------------------- # The main GPT-2 model @dataclass class GPTConfig: vocab_size : int = 50304 n_layer : int = 12 n_head : int = 6 # head dim 128 suggested by @Grad62304977 n_embd : int = 768 class GPT(nn.Module): def __init__(self, config): super().__init__() self.config = config 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 = nn.Linear(config.n_embd, config.vocab_size, bias=False) self.lm_head.weight.data.zero_() def forward(self, idx, targets=None, return_logits=True): # forward the GPT model itself x = self.transformer.wte(idx) # token embeddings of shape (b, t, n_embd) x = F.rms_norm(x, (x.size(-1),)) for block in self.transformer.h: x = block(x) x = F.rms_norm(x, (x.size(-1),)) if targets is not None: # if we are given some desired targets also calculate the loss logits = self.lm_head(x) logits = logits.float() # use tf32/fp32 for logits loss = F.cross_entropy(logits.view(-1, logits.size(-1)), targets.view(-1), ignore_index=-1) else: # inference-time mini-optimization: only forward the lm_head on the very last position logits = self.lm_head(x[:, [-1], :]) # note: using list [-1] to preserve the time dim logits = logits.float() # use tf32/fp32 for logits loss = None # there are performance reasons why not returning logits is prudent, if not needed if not return_logits: logits = None return logits, 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 # kick things off self.reset() def reset(self): self.current_shard = 0 self.current_position = self.process_rank * self.B * self.T self.tokens = _load_data_shard(self.files[self.current_shard]) 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): B = self.B T = self.T buf = self.tokens[self.current_position : self.current_position+B*T+1] buf = torch.tensor(buf.astype(np.int32), dtype=torch.long) x = (buf[:-1]).view(B, T) # inputs y = (buf[1:]).view(B, T) # targets # advance current position and load next shard if necessary self.current_position += B * T * self.num_processes if self.current_position + (B * T * self.num_processes + 1) > 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*64 # batch size, in sequences, across all devices device_batch_size : int = 64 # batch size, in sequences, per device sequence_length : int = 1024 # sequence length, in tokens num_iterations : int = 4578 # number of iterations to run warmup_iters : int = 0 warmdown_iters : int = 1308 # number of iterations of linear warmup/warmdown 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. # 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) if master_process: print(f"Training DataLoader: total number of tokens: {train_loader.ntok_total} across {len(train_loader.files)} files") print(f"Validation DataLoader: total number of tokens: {val_loader.ntok_total} across {len(val_loader.files)} files") 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() 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 ctx = torch.amp.autocast(device_type='cuda', dtype=torch.bfloat16) # 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=0.3, betas=(0.9, 0.95), fused=True) optimizer2 = torch.optim.Adam([raw_model.lm_head.weight], lr=0.002, betas=(0.9, 0.95), fused=True) optimizer3 = Muon(raw_model.transformer.h.parameters(), lr=0.02, momentum=0.95) optimizers = [optimizer1, optimizer2, optimizer3] # learning rate decay scheduler (linear warmup and warmdown) 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.warmdown_iters: return 1.0 # 3) linear warmdown else: decay_ratio = (args.num_iterations - it) / args.warmdown_iters return decay_ratio schedulers = [torch.optim.lr_scheduler.LambdaLR(opt, get_lr) for opt in optimizers] # begin logging 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') # log information about the hardware/software environment this is running on # and print the full `nvidia-smi` to file f.write(f"Running pytorch {torch.version.__version__} compiled for CUDA {torch.version.cuda}\nnvidia-smi:\n") import subprocess result = subprocess.run(['nvidia-smi'], stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True) f.write(f'{result.stdout}\n') f.write('='*100 + '\n') training_time_ms = 0 # start the clock torch.cuda.synchronize() t0 = time.time() # begin training train_loader.reset() for step in range(args.num_iterations + 1): last_step = (step == args.num_iterations) # This effectively ignores timing first 10 steps, which are slower for weird reasons. # Alternately, and slightly more correctly in terms of benchmarking, we could do 10 # steps with dummy data first, and then re-initialize the model and reset the loader. if step == 10: training_time_ms = 0 t0 = time.time() timed_steps = float('nan') if step <= 11 else (step - 10) + 1 # <= 11 to avoid bug in val # 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): x_val, y_val = val_loader.next_batch() with ctx: # of course, we'd like to use no_grad() here too, but that creates a torch.compile error for some reason _, loss = model(x_val, y_val, return_logits=False) val_loss += loss.detach() del loss dist.all_reduce(val_loss, op=dist.ReduceOp.AVG) val_loss /= val_steps # log val loss to console and to logfile if master_process: print(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') with open(logfile, "a") as f: f.write(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\n') # 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 with ctx: _, loss = model(x, y, return_logits=False) 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 # 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 if master_process: approx_time = training_time_ms + 1000 * (time.time() - t0) print(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") with open(logfile, "a") as f: f.write(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\n") 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 2.5.1+cu124 compiled for CUDA 12.4 nvidia-smi: Sun Nov 3 19:58:03 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 34C P0 141W / 700W | 5304MiB / 81559MiB | 4% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 1 NVIDIA H100 80GB HBM3 Off | 00000000:2A:00.0 Off | 0 | | N/A 35C P0 129W / 700W | 5352MiB / 81559MiB | 3% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 2 NVIDIA H100 80GB HBM3 Off | 00000000:3A:00.0 Off | 0 | | N/A 36C P0 125W / 700W | 5352MiB / 81559MiB | 2% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 3 NVIDIA H100 80GB HBM3 Off | 00000000:5D:00.0 Off | 0 | | N/A 32C P0 129W / 700W | 5352MiB / 81559MiB | 3% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 4 NVIDIA H100 80GB HBM3 Off | 00000000:9A:00.0 Off | 0 | | N/A 34C P0 141W / 700W | 5352MiB / 81559MiB | 6% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 5 NVIDIA H100 80GB HBM3 Off | 00000000:AB:00.0 Off | 0 | | N/A 37C P0 142W / 700W | 5352MiB / 81559MiB | 0% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 6 NVIDIA H100 80GB HBM3 Off | 00000000:BA:00.0 Off | 0 | | N/A 36C P0 142W / 700W | 5352MiB / 81559MiB | 2% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 7 NVIDIA H100 80GB HBM3 Off | 00000000:DB:00.0 Off | 0 | | N/A 35C P0 147W / 700W | 5112MiB / 81559MiB | 0% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 67806 C /usr/bin/python3 0MiB | | 1 N/A N/A 67807 C /usr/bin/python3 0MiB | | 2 N/A N/A 67808 C /usr/bin/python3 0MiB | | 3 N/A N/A 67809 C /usr/bin/python3 0MiB | | 4 N/A N/A 67810 C /usr/bin/python3 0MiB | | 5 N/A N/A 67811 C /usr/bin/python3 0MiB | | 6 N/A N/A 67812 C /usr/bin/python3 0MiB | | 7 N/A N/A 67813 C /usr/bin/python3 0MiB | +-----------------------------------------------------------------------------------------+ ==================================================================================================== step:0/4578 val_loss:10.8258 train_time:471ms step_avg:nanms step:1/4578 train_loss:10.8258 train_time:3347ms step_avg:nanms step:2/4578 train_loss:10.4264 train_time:3434ms step_avg:nanms step:3/4578 train_loss:9.9036 train_time:3568ms step_avg:nanms step:4/4578 train_loss:8.9782 train_time:3704ms step_avg:nanms step:5/4578 train_loss:8.0412 train_time:3842ms step_avg:nanms step:6/4578 train_loss:7.5182 train_time:3988ms step_avg:nanms step:7/4578 train_loss:7.0184 train_time:4118ms step_avg:nanms step:8/4578 train_loss:7.1936 train_time:4258ms step_avg:nanms step:9/4578 train_loss:6.9028 train_time:4403ms step_avg:nanms step:10/4578 train_loss:6.7643 train_time:4554ms step_avg:nanms step:11/4578 train_loss:6.7341 train_time:84ms step_avg:nanms step:12/4578 train_loss:6.6891 train_time:222ms step_avg:nanms step:13/4578 train_loss:6.5404 train_time:362ms step_avg:120.70ms step:14/4578 train_loss:6.5144 train_time:499ms step_avg:124.87ms step:15/4578 train_loss:6.4994 train_time:647ms step_avg:129.35ms step:16/4578 train_loss:6.4542 train_time:784ms step_avg:130.65ms step:17/4578 train_loss:6.4720 train_time:929ms step_avg:132.68ms step:18/4578 train_loss:6.5019 train_time:1077ms step_avg:134.57ms step:19/4578 train_loss:6.3335 train_time:1205ms step_avg:133.88ms step:20/4578 train_loss:6.3792 train_time:1346ms step_avg:134.63ms step:21/4578 train_loss:6.0579 train_time:1484ms step_avg:134.95ms step:22/4578 train_loss:6.3952 train_time:1626ms step_avg:135.52ms step:23/4578 train_loss:6.6283 train_time:1769ms step_avg:136.07ms step:24/4578 train_loss:6.2806 train_time:1908ms step_avg:136.32ms step:25/4578 train_loss:6.4435 train_time:2048ms step_avg:136.55ms step:26/4578 train_loss:6.1504 train_time:2189ms step_avg:136.82ms step:27/4578 train_loss:6.0603 train_time:2331ms step_avg:137.13ms step:28/4578 train_loss:6.2238 train_time:2467ms step_avg:137.07ms step:29/4578 train_loss:5.8878 train_time:2609ms step_avg:137.33ms step:30/4578 train_loss:6.1468 train_time:2749ms step_avg:137.47ms step:31/4578 train_loss:5.9902 train_time:2891ms step_avg:137.65ms step:32/4578 train_loss:5.9577 train_time:3032ms step_avg:137.81ms step:33/4578 train_loss:5.8020 train_time:3174ms step_avg:137.98ms step:34/4578 train_loss:6.0915 train_time:3310ms step_avg:137.91ms step:35/4578 train_loss:6.0166 train_time:3450ms step_avg:137.99ms step:36/4578 train_loss:6.1607 train_time:3591ms step_avg:138.13ms step:37/4578 train_loss:6.0773 train_time:3736ms step_avg:138.37ms step:38/4578 train_loss:5.9808 train_time:3874ms step_avg:138.36ms step:39/4578 train_loss:5.8647 train_time:4012ms step_avg:138.34ms step:40/4578 train_loss:5.8961 train_time:4151ms step_avg:138.35ms step:41/4578 train_loss:5.8117 train_time:4291ms step_avg:138.42ms step:42/4578 train_loss:5.8123 train_time:4431ms step_avg:138.47ms step:43/4578 train_loss:5.7165 train_time:4571ms step_avg:138.52ms step:44/4578 train_loss:5.7897 train_time:4710ms step_avg:138.54ms step:45/4578 train_loss:5.7751 train_time:4851ms step_avg:138.59ms step:46/4578 train_loss:5.9386 train_time:4992ms step_avg:138.66ms step:47/4578 train_loss:5.7310 train_time:5130ms step_avg:138.65ms step:48/4578 train_loss:5.5976 train_time:5273ms step_avg:138.77ms step:49/4578 train_loss:5.8028 train_time:5410ms step_avg:138.72ms step:50/4578 train_loss:5.6744 train_time:5550ms step_avg:138.75ms step:51/4578 train_loss:5.7978 train_time:5693ms step_avg:138.85ms step:52/4578 train_loss:5.6750 train_time:5830ms step_avg:138.81ms step:53/4578 train_loss:5.5255 train_time:5971ms step_avg:138.87ms step:54/4578 train_loss:5.6645 train_time:6111ms step_avg:138.88ms step:55/4578 train_loss:5.5355 train_time:6252ms step_avg:138.94ms step:56/4578 train_loss:5.8584 train_time:6395ms step_avg:139.01ms step:57/4578 train_loss:5.5132 train_time:6530ms step_avg:138.94ms step:58/4578 train_loss:5.4048 train_time:6671ms step_avg:138.98ms step:59/4578 train_loss:5.5517 train_time:6811ms step_avg:139.00ms step:60/4578 train_loss:5.4975 train_time:6951ms step_avg:139.03ms step:61/4578 train_loss:5.5916 train_time:7091ms step_avg:139.04ms step:62/4578 train_loss:5.3815 train_time:7233ms step_avg:139.09ms step:63/4578 train_loss:5.4760 train_time:7376ms step_avg:139.17ms step:64/4578 train_loss:5.4380 train_time:7514ms step_avg:139.14ms step:65/4578 train_loss:5.2123 train_time:7654ms step_avg:139.16ms step:66/4578 train_loss:5.2547 train_time:7792ms step_avg:139.15ms step:67/4578 train_loss:5.4417 train_time:7935ms step_avg:139.21ms step:68/4578 train_loss:5.3031 train_time:8072ms step_avg:139.17ms step:69/4578 train_loss:5.5388 train_time:8211ms step_avg:139.18ms step:70/4578 train_loss:5.1888 train_time:8352ms step_avg:139.19ms step:71/4578 train_loss:5.2521 train_time:8492ms step_avg:139.21ms step:72/4578 train_loss:5.4173 train_time:8632ms step_avg:139.22ms step:73/4578 train_loss:5.3452 train_time:8772ms step_avg:139.24ms step:74/4578 train_loss:5.2481 train_time:8912ms step_avg:139.25ms step:75/4578 train_loss:5.3636 train_time:9051ms step_avg:139.25ms step:76/4578 train_loss:5.3295 train_time:9191ms step_avg:139.26ms step:77/4578 train_loss:5.2757 train_time:9331ms step_avg:139.27ms step:78/4578 train_loss:5.3746 train_time:9472ms step_avg:139.29ms step:79/4578 train_loss:5.4590 train_time:9612ms step_avg:139.30ms step:80/4578 train_loss:5.2231 train_time:9752ms step_avg:139.32ms step:81/4578 train_loss:5.3148 train_time:9892ms step_avg:139.33ms step:82/4578 train_loss:5.0748 train_time:10032ms step_avg:139.33ms step:83/4578 train_loss:5.2599 train_time:10172ms step_avg:139.34ms step:84/4578 train_loss:5.2215 train_time:10312ms step_avg:139.35ms step:85/4578 train_loss:5.1973 train_time:10455ms step_avg:139.40ms step:86/4578 train_loss:5.0714 train_time:10600ms step_avg:139.48ms step:87/4578 train_loss:5.2601 train_time:10732ms step_avg:139.37ms step:88/4578 train_loss:5.1616 train_time:10872ms step_avg:139.38ms step:89/4578 train_loss:5.2234 train_time:11013ms step_avg:139.41ms step:90/4578 train_loss:5.1953 train_time:11152ms step_avg:139.40ms step:91/4578 train_loss:5.1039 train_time:11295ms step_avg:139.44ms step:92/4578 train_loss:5.1005 train_time:11432ms step_avg:139.41ms step:93/4578 train_loss:5.2141 train_time:11572ms step_avg:139.42ms step:94/4578 train_loss:5.0542 train_time:11712ms step_avg:139.42ms step:95/4578 train_loss:5.0576 train_time:11851ms step_avg:139.42ms step:96/4578 train_loss:5.0957 train_time:11991ms step_avg:139.43ms step:97/4578 train_loss:5.0051 train_time:12132ms step_avg:139.44ms step:98/4578 train_loss:5.0792 train_time:12270ms step_avg:139.43ms step:99/4578 train_loss:5.0092 train_time:12412ms step_avg:139.46ms step:100/4578 train_loss:5.1197 train_time:12551ms step_avg:139.45ms step:101/4578 train_loss:5.0968 train_time:12691ms step_avg:139.46ms step:102/4578 train_loss:4.9741 train_time:12831ms step_avg:139.47ms step:103/4578 train_loss:5.0998 train_time:12971ms step_avg:139.48ms step:104/4578 train_loss:5.0433 train_time:13111ms step_avg:139.48ms step:105/4578 train_loss:4.9137 train_time:13250ms step_avg:139.47ms step:106/4578 train_loss:4.9778 train_time:13389ms step_avg:139.47ms step:107/4578 train_loss:5.1866 train_time:13534ms step_avg:139.53ms step:108/4578 train_loss:4.9637 train_time:13672ms step_avg:139.51ms step:109/4578 train_loss:4.7614 train_time:13808ms step_avg:139.48ms step:110/4578 train_loss:4.9330 train_time:13951ms step_avg:139.51ms step:111/4578 train_loss:4.9214 train_time:14090ms step_avg:139.50ms step:112/4578 train_loss:4.8841 train_time:14229ms step_avg:139.50ms step:113/4578 train_loss:5.0065 train_time:14369ms step_avg:139.51ms step:114/4578 train_loss:4.9086 train_time:14509ms step_avg:139.51ms step:115/4578 train_loss:4.7789 train_time:14650ms step_avg:139.52ms step:116/4578 train_loss:4.9368 train_time:14790ms step_avg:139.53ms step:117/4578 train_loss:4.8412 train_time:14930ms step_avg:139.53ms step:118/4578 train_loss:4.7938 train_time:15071ms step_avg:139.55ms step:119/4578 train_loss:4.9543 train_time:15211ms step_avg:139.55ms step:120/4578 train_loss:4.8933 train_time:15351ms step_avg:139.55ms step:121/4578 train_loss:4.8092 train_time:15492ms step_avg:139.56ms step:122/4578 train_loss:4.7206 train_time:15638ms step_avg:139.62ms step:123/4578 train_loss:4.8468 train_time:15771ms step_avg:139.57ms step:124/4578 train_loss:4.7070 train_time:15910ms step_avg:139.56ms step:125/4578 train_loss:5.0148 train_time:16050ms step_avg:139.57ms step:125/4578 val_loss:4.8310 train_time:16105ms step_avg:140.04ms step:126/4578 train_loss:4.8736 train_time:16199ms step_avg:139.65ms step:127/4578 train_loss:4.8294 train_time:16348ms step_avg:139.72ms step:128/4578 train_loss:4.8768 train_time:16489ms step_avg:139.73ms step:129/4578 train_loss:4.7598 train_time:16639ms step_avg:139.82ms step:130/4578 train_loss:5.0642 train_time:16765ms step_avg:139.71ms step:131/4578 train_loss:4.8122 train_time:16909ms step_avg:139.74ms step:132/4578 train_loss:4.8106 train_time:17051ms step_avg:139.76ms step:133/4578 train_loss:4.7769 train_time:17185ms step_avg:139.71ms step:134/4578 train_loss:4.8157 train_time:17330ms step_avg:139.76ms step:135/4578 train_loss:4.7057 train_time:17470ms step_avg:139.76ms step:136/4578 train_loss:4.8151 train_time:17612ms step_avg:139.78ms step:137/4578 train_loss:4.6081 train_time:17750ms step_avg:139.76ms step:138/4578 train_loss:4.7709 train_time:17890ms step_avg:139.76ms step:139/4578 train_loss:4.7057 train_time:18029ms step_avg:139.76ms step:140/4578 train_loss:4.7442 train_time:18178ms step_avg:139.83ms step:141/4578 train_loss:4.8084 train_time:18320ms step_avg:139.85ms step:142/4578 train_loss:4.6876 train_time:18453ms step_avg:139.80ms step:143/4578 train_loss:4.7408 train_time:18597ms step_avg:139.83ms step:144/4578 train_loss:4.6072 train_time:18735ms step_avg:139.81ms step:145/4578 train_loss:4.7237 train_time:18876ms step_avg:139.82ms step:146/4578 train_loss:4.6742 train_time:19014ms step_avg:139.81ms step:147/4578 train_loss:4.5621 train_time:19157ms step_avg:139.83ms step:148/4578 train_loss:4.7062 train_time:19297ms step_avg:139.83ms step:149/4578 train_loss:4.7123 train_time:19443ms step_avg:139.87ms step:150/4578 train_loss:4.7207 train_time:19583ms step_avg:139.88ms step:151/4578 train_loss:4.7784 train_time:19722ms step_avg:139.87ms step:152/4578 train_loss:4.6449 train_time:19862ms step_avg:139.87ms step:153/4578 train_loss:4.6496 train_time:20001ms step_avg:139.87ms step:154/4578 train_loss:4.7309 train_time:20141ms step_avg:139.87ms step:155/4578 train_loss:4.6879 train_time:20286ms step_avg:139.90ms step:156/4578 train_loss:4.6220 train_time:20427ms step_avg:139.91ms step:157/4578 train_loss:4.6772 train_time:20573ms step_avg:139.95ms step:158/4578 train_loss:4.7779 train_time:20710ms step_avg:139.93ms step:159/4578 train_loss:4.5850 train_time:20849ms step_avg:139.92ms step:160/4578 train_loss:4.6382 train_time:20988ms step_avg:139.92ms step:161/4578 train_loss:4.4713 train_time:21128ms step_avg:139.92ms step:162/4578 train_loss:4.6548 train_time:21269ms step_avg:139.93ms step:163/4578 train_loss:4.6779 train_time:21413ms step_avg:139.95ms step:164/4578 train_loss:4.6777 train_time:21550ms step_avg:139.93ms step:165/4578 train_loss:4.4936 train_time:21690ms step_avg:139.93ms step:166/4578 train_loss:4.6014 train_time:21830ms step_avg:139.93ms step:167/4578 train_loss:4.7317 train_time:21969ms step_avg:139.93ms step:168/4578 train_loss:4.5137 train_time:22109ms step_avg:139.93ms step:169/4578 train_loss:4.5967 train_time:22250ms step_avg:139.94ms step:170/4578 train_loss:4.4733 train_time:22393ms step_avg:139.95ms step:171/4578 train_loss:4.3776 train_time:22531ms step_avg:139.94ms step:172/4578 train_loss:4.5249 train_time:22671ms step_avg:139.94ms step:173/4578 train_loss:4.5022 train_time:22810ms step_avg:139.94ms step:174/4578 train_loss:4.5522 train_time:22950ms step_avg:139.94ms step:175/4578 train_loss:4.7197 train_time:23091ms step_avg:139.94ms step:176/4578 train_loss:4.5608 train_time:23233ms step_avg:139.96ms step:177/4578 train_loss:4.4025 train_time:23371ms step_avg:139.95ms step:178/4578 train_loss:4.3816 train_time:23514ms step_avg:139.97ms step:179/4578 train_loss:4.4595 train_time:23650ms step_avg:139.94ms step:180/4578 train_loss:4.4469 train_time:23791ms step_avg:139.94ms step:181/4578 train_loss:4.4280 train_time:23932ms step_avg:139.95ms step:182/4578 train_loss:4.5780 train_time:24079ms step_avg:140.00ms step:183/4578 train_loss:4.4443 train_time:24210ms step_avg:139.94ms step:184/4578 train_loss:4.4247 train_time:24350ms step_avg:139.94ms step:185/4578 train_loss:4.4099 train_time:24493ms step_avg:139.96ms step:186/4578 train_loss:4.5159 train_time:24633ms step_avg:139.96ms step:187/4578 train_loss:4.4444 train_time:24770ms step_avg:139.94ms step:188/4578 train_loss:4.5794 train_time:24913ms step_avg:139.96ms step:189/4578 train_loss:4.4448 train_time:25252ms step_avg:141.07ms step:190/4578 train_loss:4.3807 train_time:25551ms step_avg:141.95ms step:191/4578 train_loss:4.5040 train_time:25685ms step_avg:141.91ms step:192/4578 train_loss:4.3494 train_time:25823ms step_avg:141.89ms step:193/4578 train_loss:4.2896 train_time:25962ms step_avg:141.87ms step:194/4578 train_loss:4.5209 train_time:26101ms step_avg:141.85ms step:195/4578 train_loss:4.4334 train_time:26239ms step_avg:141.83ms step:196/4578 train_loss:4.6294 train_time:26379ms step_avg:141.82ms step:197/4578 train_loss:4.4681 train_time:26524ms step_avg:141.84ms step:198/4578 train_loss:4.3227 train_time:26668ms step_avg:141.85ms step:199/4578 train_loss:4.4152 train_time:26808ms step_avg:141.84ms step:200/4578 train_loss:4.2852 train_time:26946ms step_avg:141.82ms step:201/4578 train_loss:4.3835 train_time:27086ms step_avg:141.81ms step:202/4578 train_loss:4.2677 train_time:27226ms step_avg:141.80ms step:203/4578 train_loss:4.4988 train_time:27369ms step_avg:141.81ms step:204/4578 train_loss:4.3408 train_time:27511ms step_avg:141.81ms step:205/4578 train_loss:4.4276 train_time:27649ms step_avg:141.79ms step:206/4578 train_loss:4.5047 train_time:27790ms step_avg:141.79ms step:207/4578 train_loss:4.1994 train_time:27930ms step_avg:141.77ms step:208/4578 train_loss:4.3388 train_time:28072ms step_avg:141.78ms step:209/4578 train_loss:4.3274 train_time:28215ms step_avg:141.78ms step:210/4578 train_loss:4.4841 train_time:28349ms step_avg:141.75ms step:211/4578 train_loss:4.4320 train_time:28490ms step_avg:141.74ms step:212/4578 train_loss:4.3024 train_time:28632ms step_avg:141.74ms step:213/4578 train_loss:4.3634 train_time:28769ms step_avg:141.72ms step:214/4578 train_loss:4.2713 train_time:28910ms step_avg:141.71ms step:215/4578 train_loss:4.3544 train_time:29055ms step_avg:141.73ms step:216/4578 train_loss:4.1654 train_time:29190ms step_avg:141.70ms step:217/4578 train_loss:4.2517 train_time:29332ms step_avg:141.70ms step:218/4578 train_loss:4.2490 train_time:29470ms step_avg:141.68ms step:219/4578 train_loss:4.3053 train_time:29611ms step_avg:141.68ms step:220/4578 train_loss:4.3086 train_time:29750ms step_avg:141.67ms step:221/4578 train_loss:4.3108 train_time:29896ms step_avg:141.69ms step:222/4578 train_loss:4.3338 train_time:30029ms step_avg:141.65ms step:223/4578 train_loss:4.2603 train_time:30169ms step_avg:141.64ms step:224/4578 train_loss:4.2229 train_time:30311ms step_avg:141.64ms step:225/4578 train_loss:4.5038 train_time:30449ms step_avg:141.62ms step:226/4578 train_loss:4.1279 train_time:30590ms step_avg:141.62ms step:227/4578 train_loss:4.2145 train_time:30729ms step_avg:141.61ms step:228/4578 train_loss:4.2100 train_time:30869ms step_avg:141.60ms step:229/4578 train_loss:4.3583 train_time:31009ms step_avg:141.59ms step:230/4578 train_loss:4.1440 train_time:31147ms step_avg:141.58ms step:231/4578 train_loss:4.2730 train_time:31288ms step_avg:141.58ms step:232/4578 train_loss:4.1320 train_time:31428ms step_avg:141.57ms step:233/4578 train_loss:4.1819 train_time:31569ms step_avg:141.57ms step:234/4578 train_loss:4.3241 train_time:31715ms step_avg:141.59ms step:235/4578 train_loss:4.2267 train_time:31849ms step_avg:141.55ms step:236/4578 train_loss:4.1263 train_time:31989ms step_avg:141.54ms step:237/4578 train_loss:4.2925 train_time:32128ms step_avg:141.53ms step:238/4578 train_loss:4.2899 train_time:32269ms step_avg:141.53ms step:239/4578 train_loss:4.1580 train_time:32409ms step_avg:141.53ms step:240/4578 train_loss:4.3004 train_time:32548ms step_avg:141.51ms step:241/4578 train_loss:4.3190 train_time:32690ms step_avg:141.52ms step:242/4578 train_loss:4.1730 train_time:32828ms step_avg:141.50ms step:243/4578 train_loss:4.3674 train_time:32969ms step_avg:141.50ms step:244/4578 train_loss:4.2254 train_time:33108ms step_avg:141.49ms step:245/4578 train_loss:4.2740 train_time:33248ms step_avg:141.48ms step:246/4578 train_loss:4.3434 train_time:33390ms step_avg:141.48ms step:247/4578 train_loss:4.2680 train_time:33529ms step_avg:141.47ms step:248/4578 train_loss:4.2060 train_time:33669ms step_avg:141.47ms step:249/4578 train_loss:4.3266 train_time:33809ms step_avg:141.46ms step:250/4578 train_loss:4.1242 train_time:33949ms step_avg:141.45ms step:250/4578 val_loss:4.2102 train_time:34005ms step_avg:141.69ms step:251/4578 train_loss:4.1648 train_time:34100ms step_avg:141.49ms step:252/4578 train_loss:4.2786 train_time:34246ms step_avg:141.51ms step:253/4578 train_loss:4.3323 train_time:34385ms step_avg:141.50ms step:254/4578 train_loss:4.1373 train_time:34524ms step_avg:141.49ms step:255/4578 train_loss:4.0899 train_time:34662ms step_avg:141.48ms step:256/4578 train_loss:4.2716 train_time:34804ms step_avg:141.48ms step:257/4578 train_loss:4.1917 train_time:34938ms step_avg:141.45ms step:258/4578 train_loss:4.1876 train_time:35079ms step_avg:141.45ms step:259/4578 train_loss:4.1603 train_time:35226ms step_avg:141.47ms step:260/4578 train_loss:4.2162 train_time:35364ms step_avg:141.46ms step:261/4578 train_loss:4.2481 train_time:35503ms step_avg:141.45ms step:262/4578 train_loss:4.2173 train_time:35642ms step_avg:141.44ms step:263/4578 train_loss:4.1753 train_time:35781ms step_avg:141.43ms step:264/4578 train_loss:4.0944 train_time:35921ms step_avg:141.42ms step:265/4578 train_loss:4.1757 train_time:36060ms step_avg:141.41ms step:266/4578 train_loss:4.0488 train_time:36204ms step_avg:141.42ms step:267/4578 train_loss:4.1090 train_time:36344ms step_avg:141.42ms step:268/4578 train_loss:4.1103 train_time:36485ms step_avg:141.41ms step:269/4578 train_loss:4.1301 train_time:36623ms step_avg:141.40ms step:270/4578 train_loss:4.0432 train_time:36765ms step_avg:141.40ms step:271/4578 train_loss:4.2764 train_time:36903ms step_avg:141.39ms step:272/4578 train_loss:4.1740 train_time:37042ms step_avg:141.38ms step:273/4578 train_loss:4.0885 train_time:37185ms step_avg:141.39ms step:274/4578 train_loss:4.1372 train_time:37325ms step_avg:141.38ms step:275/4578 train_loss:4.2189 train_time:37465ms step_avg:141.38ms step:276/4578 train_loss:4.2435 train_time:37605ms step_avg:141.37ms step:277/4578 train_loss:4.4167 train_time:37749ms step_avg:141.38ms step:278/4578 train_loss:4.2109 train_time:37885ms step_avg:141.36ms step:279/4578 train_loss:4.2697 train_time:38025ms step_avg:141.36ms step:280/4578 train_loss:4.1820 train_time:38164ms step_avg:141.35ms step:281/4578 train_loss:4.2966 train_time:38305ms step_avg:141.35ms step:282/4578 train_loss:4.1367 train_time:38449ms step_avg:141.36ms step:283/4578 train_loss:4.1436 train_time:38589ms step_avg:141.35ms step:284/4578 train_loss:4.0890 train_time:38725ms step_avg:141.33ms step:285/4578 train_loss:4.2348 train_time:38864ms step_avg:141.32ms step:286/4578 train_loss:4.2331 train_time:39004ms step_avg:141.32ms step:287/4578 train_loss:4.2647 train_time:39143ms step_avg:141.31ms step:288/4578 train_loss:4.0974 train_time:39286ms step_avg:141.31ms step:289/4578 train_loss:4.1992 train_time:39425ms step_avg:141.31ms step:290/4578 train_loss:4.0477 train_time:39565ms step_avg:141.30ms step:291/4578 train_loss:4.0503 train_time:39705ms step_avg:141.30ms step:292/4578 train_loss:4.1244 train_time:39844ms step_avg:141.29ms step:293/4578 train_loss:4.0499 train_time:39984ms step_avg:141.29ms step:294/4578 train_loss:4.0922 train_time:40123ms step_avg:141.28ms step:295/4578 train_loss:4.1330 train_time:40262ms step_avg:141.27ms step:296/4578 train_loss:4.0182 train_time:40402ms step_avg:141.27ms step:297/4578 train_loss:4.0255 train_time:40543ms step_avg:141.26ms step:298/4578 train_loss:4.0318 train_time:40682ms step_avg:141.26ms step:299/4578 train_loss:4.1370 train_time:40825ms step_avg:141.26ms step:300/4578 train_loss:4.0042 train_time:40961ms step_avg:141.25ms step:301/4578 train_loss:4.1424 train_time:41102ms step_avg:141.24ms step:302/4578 train_loss:4.1503 train_time:41243ms step_avg:141.24ms step:303/4578 train_loss:4.1006 train_time:41382ms step_avg:141.23ms step:304/4578 train_loss:4.1533 train_time:41523ms step_avg:141.23ms step:305/4578 train_loss:4.1358 train_time:41668ms step_avg:141.25ms step:306/4578 train_loss:4.6216 train_time:41802ms step_avg:141.22ms step:307/4578 train_loss:4.1035 train_time:41941ms step_avg:141.22ms step:308/4578 train_loss:4.0137 train_time:42083ms step_avg:141.22ms step:309/4578 train_loss:4.1657 train_time:42223ms step_avg:141.21ms step:310/4578 train_loss:4.0309 train_time:42362ms step_avg:141.21ms step:311/4578 train_loss:4.2431 train_time:42503ms step_avg:141.20ms step:312/4578 train_loss:4.0990 train_time:42642ms step_avg:141.20ms step:313/4578 train_loss:4.0440 train_time:42787ms step_avg:141.21ms step:314/4578 train_loss:4.1316 train_time:42923ms step_avg:141.19ms step:315/4578 train_loss:4.2574 train_time:43063ms step_avg:141.19ms step:316/4578 train_loss:4.1216 train_time:43203ms step_avg:141.18ms step:317/4578 train_loss:3.9632 train_time:43343ms step_avg:141.18ms step:318/4578 train_loss:4.0441 train_time:43484ms step_avg:141.18ms step:319/4578 train_loss:4.0781 train_time:43623ms step_avg:141.18ms step:320/4578 train_loss:4.0508 train_time:43764ms step_avg:141.17ms step:321/4578 train_loss:4.1630 train_time:43904ms step_avg:141.17ms step:322/4578 train_loss:4.1143 train_time:44043ms step_avg:141.16ms step:323/4578 train_loss:4.0938 train_time:44184ms step_avg:141.16ms step:324/4578 train_loss:4.1724 train_time:44323ms step_avg:141.16ms step:325/4578 train_loss:4.1214 train_time:44467ms step_avg:141.16ms step:326/4578 train_loss:4.1944 train_time:44603ms step_avg:141.15ms step:327/4578 train_loss:4.0595 train_time:44742ms step_avg:141.14ms step:328/4578 train_loss:4.5599 train_time:44889ms step_avg:141.16ms step:329/4578 train_loss:4.2354 train_time:45023ms step_avg:141.14ms step:330/4578 train_loss:3.9793 train_time:45162ms step_avg:141.13ms step:331/4578 train_loss:3.9210 train_time:45311ms step_avg:141.16ms step:332/4578 train_loss:4.1373 train_time:45441ms step_avg:141.12ms step:333/4578 train_loss:4.0659 train_time:45582ms step_avg:141.12ms step:334/4578 train_loss:4.0441 train_time:45730ms step_avg:141.14ms step:335/4578 train_loss:4.0030 train_time:45861ms step_avg:141.11ms step:336/4578 train_loss:4.1796 train_time:46002ms step_avg:141.11ms step:337/4578 train_loss:4.1195 train_time:46142ms step_avg:141.11ms step:338/4578 train_loss:4.6032 train_time:46282ms step_avg:141.10ms step:339/4578 train_loss:4.1055 train_time:46424ms step_avg:141.11ms step:340/4578 train_loss:4.0519 train_time:46562ms step_avg:141.10ms step:341/4578 train_loss:4.0923 train_time:46704ms step_avg:141.10ms step:342/4578 train_loss:4.0080 train_time:46843ms step_avg:141.09ms step:343/4578 train_loss:3.9760 train_time:46983ms step_avg:141.09ms step:344/4578 train_loss:4.0187 train_time:47124ms step_avg:141.09ms step:345/4578 train_loss:4.1556 train_time:47263ms step_avg:141.08ms step:346/4578 train_loss:3.9999 train_time:47404ms step_avg:141.08ms step:347/4578 train_loss:3.9318 train_time:47542ms step_avg:141.08ms step:348/4578 train_loss:3.9737 train_time:47682ms step_avg:141.07ms step:349/4578 train_loss:4.0180 train_time:47823ms step_avg:141.07ms step:350/4578 train_loss:3.9826 train_time:47961ms step_avg:141.06ms step:351/4578 train_loss:3.7127 train_time:48101ms step_avg:141.06ms step:352/4578 train_loss:3.9698 train_time:48241ms step_avg:141.06ms step:353/4578 train_loss:4.3214 train_time:48381ms step_avg:141.05ms step:354/4578 train_loss:3.8241 train_time:48522ms step_avg:141.05ms step:355/4578 train_loss:4.0818 train_time:48671ms step_avg:141.08ms step:356/4578 train_loss:3.9455 train_time:48802ms step_avg:141.05ms step:357/4578 train_loss:4.0555 train_time:48943ms step_avg:141.05ms step:358/4578 train_loss:3.9895 train_time:49082ms step_avg:141.04ms step:359/4578 train_loss:4.0083 train_time:49222ms step_avg:141.04ms step:360/4578 train_loss:4.0328 train_time:49365ms step_avg:141.04ms step:361/4578 train_loss:3.6226 train_time:49503ms step_avg:141.03ms step:362/4578 train_loss:4.1793 train_time:49642ms step_avg:141.03ms step:363/4578 train_loss:4.0773 train_time:49784ms step_avg:141.03ms step:364/4578 train_loss:4.0000 train_time:49922ms step_avg:141.02ms step:365/4578 train_loss:3.9082 train_time:50061ms step_avg:141.02ms step:366/4578 train_loss:4.0738 train_time:50202ms step_avg:141.02ms step:367/4578 train_loss:4.0250 train_time:50341ms step_avg:141.01ms step:368/4578 train_loss:4.0180 train_time:50485ms step_avg:141.02ms step:369/4578 train_loss:4.0008 train_time:50625ms step_avg:141.02ms step:370/4578 train_loss:3.9009 train_time:50768ms step_avg:141.02ms step:371/4578 train_loss:4.0444 train_time:50904ms step_avg:141.01ms step:372/4578 train_loss:3.9289 train_time:51042ms step_avg:141.00ms step:373/4578 train_loss:3.8569 train_time:51188ms step_avg:141.01ms step:374/4578 train_loss:4.0756 train_time:51324ms step_avg:141.00ms step:375/4578 train_loss:3.9941 train_time:51463ms step_avg:141.00ms step:375/4578 val_loss:3.9903 train_time:51518ms step_avg:141.15ms step:376/4578 train_loss:3.9667 train_time:51609ms step_avg:141.01ms step:377/4578 train_loss:4.0259 train_time:51756ms step_avg:141.02ms step:378/4578 train_loss:3.9513 train_time:52045ms step_avg:141.43ms step:379/4578 train_loss:4.0054 train_time:52183ms step_avg:141.42ms step:380/4578 train_loss:4.0282 train_time:52480ms step_avg:141.84ms step:381/4578 train_loss:4.1068 train_time:52616ms step_avg:141.82ms step:382/4578 train_loss:4.0012 train_time:52754ms step_avg:141.81ms step:383/4578 train_loss:3.9800 train_time:52893ms step_avg:141.80ms step:384/4578 train_loss:3.9475 train_time:53031ms step_avg:141.79ms step:385/4578 train_loss:4.0263 train_time:53169ms step_avg:141.78ms step:386/4578 train_loss:3.9416 train_time:53308ms step_avg:141.78ms step:387/4578 train_loss:4.0492 train_time:53454ms step_avg:141.79ms step:388/4578 train_loss:4.2390 train_time:53594ms step_avg:141.78ms step:389/4578 train_loss:3.9602 train_time:53733ms step_avg:141.78ms step:390/4578 train_loss:3.9474 train_time:53872ms step_avg:141.77ms step:391/4578 train_loss:4.0460 train_time:54010ms step_avg:141.76ms step:392/4578 train_loss:3.9680 train_time:54149ms step_avg:141.75ms step:393/4578 train_loss:4.0756 train_time:54289ms step_avg:141.75ms step:394/4578 train_loss:3.9176 train_time:54432ms step_avg:141.75ms step:395/4578 train_loss:4.0483 train_time:54570ms step_avg:141.74ms step:396/4578 train_loss:3.7949 train_time:54710ms step_avg:141.74ms step:397/4578 train_loss:3.9956 train_time:54854ms step_avg:141.74ms step:398/4578 train_loss:4.0374 train_time:54992ms step_avg:141.73ms step:399/4578 train_loss:4.0432 train_time:55130ms step_avg:141.72ms step:400/4578 train_loss:3.9391 train_time:55269ms step_avg:141.72ms step:401/4578 train_loss:3.9940 train_time:55411ms step_avg:141.72ms step:402/4578 train_loss:4.0674 train_time:55550ms step_avg:141.71ms step:403/4578 train_loss:3.9966 train_time:55691ms step_avg:141.71ms step:404/4578 train_loss:4.1078 train_time:55830ms step_avg:141.70ms step:405/4578 train_loss:3.8633 train_time:55971ms step_avg:141.70ms step:406/4578 train_loss:3.9432 train_time:56110ms step_avg:141.69ms step:407/4578 train_loss:4.2454 train_time:56249ms step_avg:141.68ms step:408/4578 train_loss:3.9487 train_time:56390ms step_avg:141.68ms step:409/4578 train_loss:3.9792 train_time:56530ms step_avg:141.68ms step:410/4578 train_loss:4.0197 train_time:56670ms step_avg:141.67ms step:411/4578 train_loss:3.9075 train_time:56810ms step_avg:141.67ms step:412/4578 train_loss:3.9297 train_time:56949ms step_avg:141.66ms step:413/4578 train_loss:4.3507 train_time:57089ms step_avg:141.66ms step:414/4578 train_loss:3.7874 train_time:57228ms step_avg:141.65ms step:415/4578 train_loss:4.1733 train_time:57373ms step_avg:141.66ms step:416/4578 train_loss:3.9146 train_time:57508ms step_avg:141.65ms step:417/4578 train_loss:3.9261 train_time:57649ms step_avg:141.64ms step:418/4578 train_loss:4.1093 train_time:57798ms step_avg:141.66ms step:419/4578 train_loss:3.8499 train_time:57929ms step_avg:141.64ms step:420/4578 train_loss:3.9645 train_time:58069ms step_avg:141.63ms step:421/4578 train_loss:3.8937 train_time:58209ms step_avg:141.63ms step:422/4578 train_loss:3.8077 train_time:58349ms step_avg:141.62ms step:423/4578 train_loss:3.9363 train_time:58490ms step_avg:141.62ms step:424/4578 train_loss:4.0254 train_time:58629ms step_avg:141.62ms step:425/4578 train_loss:3.7824 train_time:58769ms step_avg:141.61ms step:426/4578 train_loss:3.9667 train_time:58914ms step_avg:141.62ms step:427/4578 train_loss:3.8463 train_time:59051ms step_avg:141.61ms step:428/4578 train_loss:4.0541 train_time:59189ms step_avg:141.60ms step:429/4578 train_loss:3.9820 train_time:59330ms step_avg:141.60ms step:430/4578 train_loss:3.9123 train_time:59470ms step_avg:141.59ms step:431/4578 train_loss:3.8837 train_time:59610ms step_avg:141.59ms step:432/4578 train_loss:3.7913 train_time:59750ms step_avg:141.59ms step:433/4578 train_loss:3.9236 train_time:59891ms step_avg:141.59ms step:434/4578 train_loss:3.9813 train_time:60035ms step_avg:141.59ms step:435/4578 train_loss:3.9328 train_time:60170ms step_avg:141.58ms step:436/4578 train_loss:3.9688 train_time:60310ms step_avg:141.57ms step:437/4578 train_loss:3.9907 train_time:60455ms step_avg:141.58ms step:438/4578 train_loss:3.8759 train_time:60588ms step_avg:141.56ms step:439/4578 train_loss:3.8741 train_time:60728ms step_avg:141.56ms step:440/4578 train_loss:3.8600 train_time:60868ms step_avg:141.55ms step:441/4578 train_loss:4.0440 train_time:61008ms step_avg:141.55ms step:442/4578 train_loss:3.9291 train_time:61147ms step_avg:141.54ms step:443/4578 train_loss:3.9090 train_time:61288ms step_avg:141.54ms step:444/4578 train_loss:3.8033 train_time:61433ms step_avg:141.55ms step:445/4578 train_loss:4.0771 train_time:61567ms step_avg:141.53ms step:446/4578 train_loss:3.9987 train_time:61707ms step_avg:141.53ms step:447/4578 train_loss:3.9899 train_time:61853ms step_avg:141.54ms step:448/4578 train_loss:3.9111 train_time:61988ms step_avg:141.52ms step:449/4578 train_loss:4.0103 train_time:62127ms step_avg:141.52ms step:450/4578 train_loss:3.8450 train_time:62267ms step_avg:141.52ms step:451/4578 train_loss:3.8736 train_time:62410ms step_avg:141.52ms step:452/4578 train_loss:3.7449 train_time:62551ms step_avg:141.52ms step:453/4578 train_loss:3.8646 train_time:62692ms step_avg:141.52ms step:454/4578 train_loss:3.8342 train_time:62827ms step_avg:141.50ms step:455/4578 train_loss:3.8044 train_time:62968ms step_avg:141.50ms step:456/4578 train_loss:4.0116 train_time:63110ms step_avg:141.50ms step:457/4578 train_loss:3.8922 train_time:63247ms step_avg:141.49ms step:458/4578 train_loss:3.9580 train_time:63391ms step_avg:141.50ms step:459/4578 train_loss:3.9904 train_time:63527ms step_avg:141.49ms step:460/4578 train_loss:3.7958 train_time:63667ms step_avg:141.48ms step:461/4578 train_loss:3.9618 train_time:63808ms step_avg:141.48ms step:462/4578 train_loss:3.8677 train_time:63948ms step_avg:141.48ms step:463/4578 train_loss:3.8839 train_time:64089ms step_avg:141.48ms step:464/4578 train_loss:3.9379 train_time:64229ms step_avg:141.47ms step:465/4578 train_loss:3.8782 train_time:64368ms step_avg:141.47ms step:466/4578 train_loss:3.8833 train_time:64511ms step_avg:141.47ms step:467/4578 train_loss:3.9722 train_time:64647ms step_avg:141.46ms step:468/4578 train_loss:3.9819 train_time:64787ms step_avg:141.46ms step:469/4578 train_loss:3.9599 train_time:64927ms step_avg:141.45ms step:470/4578 train_loss:3.8555 train_time:65067ms step_avg:141.45ms step:471/4578 train_loss:3.9345 train_time:65208ms step_avg:141.45ms step:472/4578 train_loss:3.9830 train_time:65347ms step_avg:141.44ms step:473/4578 train_loss:3.9306 train_time:65488ms step_avg:141.44ms step:474/4578 train_loss:3.8873 train_time:65627ms step_avg:141.44ms step:475/4578 train_loss:3.7449 train_time:65768ms step_avg:141.44ms step:476/4578 train_loss:4.1841 train_time:65908ms step_avg:141.43ms step:477/4578 train_loss:3.9317 train_time:66048ms step_avg:141.43ms step:478/4578 train_loss:3.7504 train_time:66188ms step_avg:141.43ms step:479/4578 train_loss:3.9718 train_time:66327ms step_avg:141.42ms step:480/4578 train_loss:3.9334 train_time:66467ms step_avg:141.42ms step:481/4578 train_loss:4.0782 train_time:66607ms step_avg:141.42ms step:482/4578 train_loss:3.8859 train_time:66747ms step_avg:141.41ms step:483/4578 train_loss:3.6947 train_time:66888ms step_avg:141.41ms step:484/4578 train_loss:3.9704 train_time:67027ms step_avg:141.41ms step:485/4578 train_loss:3.8307 train_time:67167ms step_avg:141.41ms step:486/4578 train_loss:3.8360 train_time:67308ms step_avg:141.40ms step:487/4578 train_loss:3.7600 train_time:67446ms step_avg:141.40ms step:488/4578 train_loss:3.8357 train_time:67587ms step_avg:141.40ms step:489/4578 train_loss:4.0351 train_time:67727ms step_avg:141.39ms step:490/4578 train_loss:3.8794 train_time:67868ms step_avg:141.39ms step:491/4578 train_loss:3.7668 train_time:68007ms step_avg:141.39ms step:492/4578 train_loss:3.7753 train_time:68147ms step_avg:141.38ms step:493/4578 train_loss:3.8972 train_time:68288ms step_avg:141.38ms step:494/4578 train_loss:3.7506 train_time:68427ms step_avg:141.38ms step:495/4578 train_loss:3.8759 train_time:68568ms step_avg:141.38ms step:496/4578 train_loss:3.8154 train_time:68708ms step_avg:141.37ms step:497/4578 train_loss:3.6977 train_time:68847ms step_avg:141.37ms step:498/4578 train_loss:3.8941 train_time:68988ms step_avg:141.37ms step:499/4578 train_loss:3.9610 train_time:69128ms step_avg:141.37ms step:500/4578 train_loss:3.9993 train_time:69268ms step_avg:141.36ms step:500/4578 val_loss:3.8718 train_time:69323ms step_avg:141.48ms step:501/4578 train_loss:3.9179 train_time:69418ms step_avg:141.38ms step:502/4578 train_loss:3.9623 train_time:69565ms step_avg:141.39ms step:503/4578 train_loss:3.9051 train_time:69704ms step_avg:141.39ms step:504/4578 train_loss:3.9423 train_time:69842ms step_avg:141.38ms step:505/4578 train_loss:3.8891 train_time:69980ms step_avg:141.37ms step:506/4578 train_loss:3.9837 train_time:70117ms step_avg:141.37ms step:507/4578 train_loss:3.8079 train_time:70256ms step_avg:141.36ms step:508/4578 train_loss:3.9287 train_time:70398ms step_avg:141.36ms step:509/4578 train_loss:3.9964 train_time:70543ms step_avg:141.37ms step:510/4578 train_loss:3.9369 train_time:70683ms step_avg:141.37ms step:511/4578 train_loss:3.7519 train_time:70823ms step_avg:141.36ms step:512/4578 train_loss:3.9414 train_time:70964ms step_avg:141.36ms step:513/4578 train_loss:3.8790 train_time:71101ms step_avg:141.35ms step:514/4578 train_loss:3.8501 train_time:71240ms step_avg:141.35ms step:515/4578 train_loss:3.9320 train_time:71381ms step_avg:141.35ms step:516/4578 train_loss:3.9011 train_time:71523ms step_avg:141.35ms step:517/4578 train_loss:4.2601 train_time:71662ms step_avg:141.35ms step:518/4578 train_loss:3.8519 train_time:71805ms step_avg:141.35ms step:519/4578 train_loss:3.9532 train_time:71941ms step_avg:141.34ms step:520/4578 train_loss:3.8548 train_time:72081ms step_avg:141.33ms step:521/4578 train_loss:3.8532 train_time:72220ms step_avg:141.33ms step:522/4578 train_loss:3.8020 train_time:72360ms step_avg:141.33ms step:523/4578 train_loss:3.8230 train_time:72503ms step_avg:141.33ms step:524/4578 train_loss:4.4667 train_time:72643ms step_avg:141.33ms step:525/4578 train_loss:3.9072 train_time:72783ms step_avg:141.33ms step:526/4578 train_loss:3.8509 train_time:72923ms step_avg:141.32ms step:527/4578 train_loss:3.8595 train_time:73062ms step_avg:141.32ms step:528/4578 train_loss:3.8186 train_time:73205ms step_avg:141.32ms step:529/4578 train_loss:3.7893 train_time:73342ms step_avg:141.31ms step:530/4578 train_loss:4.0034 train_time:73487ms step_avg:141.32ms step:531/4578 train_loss:3.8118 train_time:73623ms step_avg:141.31ms step:532/4578 train_loss:4.0829 train_time:73763ms step_avg:141.31ms step:533/4578 train_loss:3.8945 train_time:73903ms step_avg:141.31ms step:534/4578 train_loss:3.8199 train_time:74042ms step_avg:141.30ms step:535/4578 train_loss:3.8428 train_time:74185ms step_avg:141.30ms step:536/4578 train_loss:3.7805 train_time:74325ms step_avg:141.30ms step:537/4578 train_loss:3.9085 train_time:74461ms step_avg:141.29ms step:538/4578 train_loss:3.8960 train_time:74602ms step_avg:141.29ms step:539/4578 train_loss:3.7972 train_time:74742ms step_avg:141.29ms step:540/4578 train_loss:4.2904 train_time:74882ms step_avg:141.29ms step:541/4578 train_loss:3.8372 train_time:75022ms step_avg:141.28ms step:542/4578 train_loss:3.9461 train_time:75160ms step_avg:141.28ms step:543/4578 train_loss:3.7703 train_time:75300ms step_avg:141.28ms step:544/4578 train_loss:3.7460 train_time:75440ms step_avg:141.27ms step:545/4578 train_loss:3.8336 train_time:75580ms step_avg:141.27ms step:546/4578 train_loss:3.7600 train_time:75721ms step_avg:141.27ms step:547/4578 train_loss:3.8129 train_time:75860ms step_avg:141.27ms step:548/4578 train_loss:3.8154 train_time:76001ms step_avg:141.26ms step:549/4578 train_loss:3.7920 train_time:76140ms step_avg:141.26ms step:550/4578 train_loss:3.8911 train_time:76283ms step_avg:141.26ms step:551/4578 train_loss:3.7689 train_time:76420ms step_avg:141.26ms step:552/4578 train_loss:3.7842 train_time:76559ms step_avg:141.25ms step:553/4578 train_loss:4.1160 train_time:76700ms step_avg:141.25ms step:554/4578 train_loss:3.9140 train_time:76840ms step_avg:141.25ms step:555/4578 train_loss:3.8772 train_time:76981ms step_avg:141.25ms step:556/4578 train_loss:3.8210 train_time:77120ms step_avg:141.25ms step:557/4578 train_loss:3.8552 train_time:77264ms step_avg:141.25ms step:558/4578 train_loss:3.5157 train_time:77401ms step_avg:141.24ms step:559/4578 train_loss:3.7711 train_time:77540ms step_avg:141.24ms step:560/4578 train_loss:3.8178 train_time:77680ms step_avg:141.24ms step:561/4578 train_loss:3.8685 train_time:77820ms step_avg:141.23ms step:562/4578 train_loss:3.7751 train_time:77962ms step_avg:141.23ms step:563/4578 train_loss:3.7218 train_time:78100ms step_avg:141.23ms step:564/4578 train_loss:3.9242 train_time:78240ms step_avg:141.23ms step:565/4578 train_loss:3.7397 train_time:78380ms step_avg:141.23ms step:566/4578 train_loss:3.8491 train_time:78522ms step_avg:141.23ms step:567/4578 train_loss:3.7981 train_time:78806ms step_avg:141.48ms step:568/4578 train_loss:3.7607 train_time:78943ms step_avg:141.48ms step:569/4578 train_loss:3.8460 train_time:79082ms step_avg:141.47ms step:570/4578 train_loss:3.8131 train_time:79386ms step_avg:141.76ms step:571/4578 train_loss:3.8485 train_time:79522ms step_avg:141.75ms step:572/4578 train_loss:3.9355 train_time:79660ms step_avg:141.74ms step:573/4578 train_loss:3.8811 train_time:79798ms step_avg:141.74ms step:574/4578 train_loss:3.8952 train_time:79936ms step_avg:141.73ms step:575/4578 train_loss:3.9409 train_time:80075ms step_avg:141.73ms step:576/4578 train_loss:3.9002 train_time:80215ms step_avg:141.72ms step:577/4578 train_loss:3.9190 train_time:80363ms step_avg:141.73ms step:578/4578 train_loss:3.8480 train_time:80505ms step_avg:141.73ms step:579/4578 train_loss:3.8375 train_time:80645ms step_avg:141.73ms step:580/4578 train_loss:3.8262 train_time:80784ms step_avg:141.73ms step:581/4578 train_loss:3.7628 train_time:80922ms step_avg:141.72ms step:582/4578 train_loss:3.7927 train_time:81061ms step_avg:141.71ms step:583/4578 train_loss:4.0255 train_time:81202ms step_avg:141.71ms step:584/4578 train_loss:3.7869 train_time:81345ms step_avg:141.72ms step:585/4578 train_loss:3.7532 train_time:81484ms step_avg:141.71ms step:586/4578 train_loss:3.9393 train_time:81626ms step_avg:141.71ms step:587/4578 train_loss:3.6973 train_time:81766ms step_avg:141.71ms step:588/4578 train_loss:3.8299 train_time:81904ms step_avg:141.70ms step:589/4578 train_loss:3.8161 train_time:82053ms step_avg:141.72ms step:590/4578 train_loss:4.1632 train_time:82185ms step_avg:141.70ms step:591/4578 train_loss:3.9414 train_time:82328ms step_avg:141.70ms step:592/4578 train_loss:3.6759 train_time:82469ms step_avg:141.70ms step:593/4578 train_loss:3.7025 train_time:82605ms step_avg:141.69ms step:594/4578 train_loss:3.6921 train_time:82743ms step_avg:141.68ms step:595/4578 train_loss:3.7304 train_time:82883ms step_avg:141.68ms step:596/4578 train_loss:4.0959 train_time:83023ms step_avg:141.68ms step:597/4578 train_loss:3.8158 train_time:83164ms step_avg:141.68ms step:598/4578 train_loss:3.7519 train_time:83303ms step_avg:141.67ms step:599/4578 train_loss:3.8205 train_time:83443ms step_avg:141.67ms step:600/4578 train_loss:3.6418 train_time:83586ms step_avg:141.67ms step:601/4578 train_loss:3.7628 train_time:83727ms step_avg:141.67ms step:602/4578 train_loss:3.7968 train_time:83863ms step_avg:141.66ms step:603/4578 train_loss:3.8206 train_time:84003ms step_avg:141.66ms step:604/4578 train_loss:3.9456 train_time:84144ms step_avg:141.66ms step:605/4578 train_loss:3.7979 train_time:84283ms step_avg:141.65ms step:606/4578 train_loss:3.7773 train_time:84425ms step_avg:141.65ms step:607/4578 train_loss:3.7360 train_time:84565ms step_avg:141.65ms step:608/4578 train_loss:3.9842 train_time:84706ms step_avg:141.65ms step:609/4578 train_loss:3.8076 train_time:84845ms step_avg:141.64ms step:610/4578 train_loss:3.7841 train_time:84984ms step_avg:141.64ms step:611/4578 train_loss:3.8812 train_time:85124ms step_avg:141.64ms step:612/4578 train_loss:3.7807 train_time:85263ms step_avg:141.63ms step:613/4578 train_loss:3.7705 train_time:85404ms step_avg:141.63ms step:614/4578 train_loss:3.9366 train_time:85543ms step_avg:141.63ms step:615/4578 train_loss:3.8844 train_time:85684ms step_avg:141.63ms step:616/4578 train_loss:3.8615 train_time:85824ms step_avg:141.62ms step:617/4578 train_loss:3.7808 train_time:85963ms step_avg:141.62ms step:618/4578 train_loss:3.7335 train_time:86103ms step_avg:141.62ms step:619/4578 train_loss:3.8404 train_time:86243ms step_avg:141.61ms step:620/4578 train_loss:3.7370 train_time:86386ms step_avg:141.62ms step:621/4578 train_loss:3.7660 train_time:86522ms step_avg:141.61ms step:622/4578 train_loss:4.0714 train_time:86661ms step_avg:141.60ms step:623/4578 train_loss:3.7541 train_time:86809ms step_avg:141.61ms step:624/4578 train_loss:3.7841 train_time:86943ms step_avg:141.60ms step:625/4578 train_loss:3.8612 train_time:87083ms step_avg:141.60ms step:625/4578 val_loss:3.7954 train_time:87139ms step_avg:141.69ms step:626/4578 train_loss:3.8914 train_time:87235ms step_avg:141.62ms step:627/4578 train_loss:3.9103 train_time:87382ms step_avg:141.62ms step:628/4578 train_loss:3.8971 train_time:87521ms step_avg:141.62ms step:629/4578 train_loss:3.9330 train_time:87660ms step_avg:141.62ms step:630/4578 train_loss:3.7609 train_time:87798ms step_avg:141.61ms step:631/4578 train_loss:3.8851 train_time:87937ms step_avg:141.60ms step:632/4578 train_loss:3.9134 train_time:88074ms step_avg:141.60ms step:633/4578 train_loss:3.8188 train_time:88217ms step_avg:141.60ms step:634/4578 train_loss:3.7561 train_time:88365ms step_avg:141.61ms step:635/4578 train_loss:3.8539 train_time:88502ms step_avg:141.60ms step:636/4578 train_loss:4.1103 train_time:88643ms step_avg:141.60ms step:637/4578 train_loss:3.7054 train_time:88785ms step_avg:141.60ms step:638/4578 train_loss:3.5326 train_time:88921ms step_avg:141.59ms step:639/4578 train_loss:3.7465 train_time:89060ms step_avg:141.59ms step:640/4578 train_loss:3.7782 train_time:89208ms step_avg:141.60ms step:641/4578 train_loss:3.7383 train_time:89341ms step_avg:141.59ms step:642/4578 train_loss:3.7476 train_time:89481ms step_avg:141.58ms step:643/4578 train_loss:3.7852 train_time:89622ms step_avg:141.58ms step:644/4578 train_loss:3.7875 train_time:89770ms step_avg:141.59ms step:645/4578 train_loss:3.7230 train_time:89898ms step_avg:141.57ms step:646/4578 train_loss:3.9425 train_time:90039ms step_avg:141.57ms step:647/4578 train_loss:3.8508 train_time:90188ms step_avg:141.58ms step:648/4578 train_loss:3.8369 train_time:90319ms step_avg:141.57ms step:649/4578 train_loss:3.8645 train_time:90459ms step_avg:141.56ms step:650/4578 train_loss:3.9328 train_time:90600ms step_avg:141.56ms step:651/4578 train_loss:3.7881 train_time:90743ms step_avg:141.56ms step:652/4578 train_loss:3.9277 train_time:90878ms step_avg:141.56ms step:653/4578 train_loss:3.7548 train_time:91018ms step_avg:141.55ms step:654/4578 train_loss:3.8265 train_time:91165ms step_avg:141.56ms step:655/4578 train_loss:3.5956 train_time:91299ms step_avg:141.55ms step:656/4578 train_loss:3.7411 train_time:91441ms step_avg:141.55ms step:657/4578 train_loss:3.7488 train_time:91584ms step_avg:141.55ms step:658/4578 train_loss:3.6787 train_time:91721ms step_avg:141.54ms step:659/4578 train_loss:3.8548 train_time:91859ms step_avg:141.54ms step:660/4578 train_loss:3.7574 train_time:92000ms step_avg:141.54ms step:661/4578 train_loss:3.8477 train_time:92139ms step_avg:141.53ms step:662/4578 train_loss:3.9228 train_time:92280ms step_avg:141.53ms step:663/4578 train_loss:3.8380 train_time:92419ms step_avg:141.53ms step:664/4578 train_loss:3.7142 train_time:92561ms step_avg:141.53ms step:665/4578 train_loss:3.7931 train_time:92699ms step_avg:141.52ms step:666/4578 train_loss:3.6611 train_time:92844ms step_avg:141.53ms step:667/4578 train_loss:3.9511 train_time:92978ms step_avg:141.52ms step:668/4578 train_loss:3.7877 train_time:93121ms step_avg:141.52ms step:669/4578 train_loss:3.7964 train_time:93258ms step_avg:141.52ms step:670/4578 train_loss:3.6447 train_time:93412ms step_avg:141.53ms step:671/4578 train_loss:3.7635 train_time:93538ms step_avg:141.51ms step:672/4578 train_loss:3.7192 train_time:93680ms step_avg:141.51ms step:673/4578 train_loss:3.7436 train_time:93817ms step_avg:141.50ms step:674/4578 train_loss:4.0138 train_time:93957ms step_avg:141.50ms step:675/4578 train_loss:3.8100 train_time:94100ms step_avg:141.50ms step:676/4578 train_loss:3.8731 train_time:94237ms step_avg:141.50ms step:677/4578 train_loss:3.6621 train_time:94377ms step_avg:141.49ms step:678/4578 train_loss:3.7633 train_time:94519ms step_avg:141.50ms step:679/4578 train_loss:3.7120 train_time:94660ms step_avg:141.49ms step:680/4578 train_loss:3.8512 train_time:94798ms step_avg:141.49ms step:681/4578 train_loss:3.7482 train_time:94937ms step_avg:141.49ms step:682/4578 train_loss:3.7793 train_time:95076ms step_avg:141.48ms step:683/4578 train_loss:3.8621 train_time:95217ms step_avg:141.48ms step:684/4578 train_loss:3.8949 train_time:95356ms step_avg:141.48ms step:685/4578 train_loss:3.8015 train_time:95498ms step_avg:141.48ms step:686/4578 train_loss:3.8683 train_time:95638ms step_avg:141.48ms step:687/4578 train_loss:3.8035 train_time:95778ms step_avg:141.47ms step:688/4578 train_loss:3.8478 train_time:95918ms step_avg:141.47ms step:689/4578 train_loss:3.4514 train_time:96058ms step_avg:141.47ms step:690/4578 train_loss:3.5828 train_time:96203ms step_avg:141.48ms step:691/4578 train_loss:3.7236 train_time:96338ms step_avg:141.47ms step:692/4578 train_loss:3.6063 train_time:96476ms step_avg:141.46ms step:693/4578 train_loss:3.8203 train_time:96617ms step_avg:141.46ms step:694/4578 train_loss:3.8340 train_time:96763ms step_avg:141.47ms step:695/4578 train_loss:3.7185 train_time:96904ms step_avg:141.47ms step:696/4578 train_loss:3.7099 train_time:97038ms step_avg:141.45ms step:697/4578 train_loss:4.0210 train_time:97177ms step_avg:141.45ms step:698/4578 train_loss:3.7717 train_time:97319ms step_avg:141.45ms step:699/4578 train_loss:3.8106 train_time:97458ms step_avg:141.45ms step:700/4578 train_loss:3.9723 train_time:97600ms step_avg:141.45ms step:701/4578 train_loss:3.7354 train_time:97753ms step_avg:141.47ms step:702/4578 train_loss:3.7064 train_time:97879ms step_avg:141.44ms step:703/4578 train_loss:3.6929 train_time:98020ms step_avg:141.44ms step:704/4578 train_loss:3.6505 train_time:98160ms step_avg:141.44ms step:705/4578 train_loss:3.7373 train_time:98303ms step_avg:141.44ms step:706/4578 train_loss:3.7312 train_time:98440ms step_avg:141.44ms step:707/4578 train_loss:3.7456 train_time:98580ms step_avg:141.43ms step:708/4578 train_loss:3.8111 train_time:98720ms step_avg:141.43ms step:709/4578 train_loss:3.7642 train_time:98860ms step_avg:141.43ms step:710/4578 train_loss:3.7424 train_time:99000ms step_avg:141.43ms step:711/4578 train_loss:3.7136 train_time:99143ms step_avg:141.43ms step:712/4578 train_loss:3.7564 train_time:99279ms step_avg:141.42ms step:713/4578 train_loss:3.8142 train_time:99422ms step_avg:141.43ms step:714/4578 train_loss:3.8323 train_time:99565ms step_avg:141.43ms step:715/4578 train_loss:3.7424 train_time:99699ms step_avg:141.42ms step:716/4578 train_loss:3.7346 train_time:99845ms step_avg:141.42ms step:717/4578 train_loss:3.7547 train_time:99978ms step_avg:141.41ms step:718/4578 train_loss:3.9061 train_time:100118ms step_avg:141.41ms step:719/4578 train_loss:3.7583 train_time:100261ms step_avg:141.41ms step:720/4578 train_loss:3.8378 train_time:100405ms step_avg:141.42ms step:721/4578 train_loss:3.9982 train_time:100539ms step_avg:141.41ms step:722/4578 train_loss:3.6252 train_time:100685ms step_avg:141.41ms step:723/4578 train_loss:3.8984 train_time:100818ms step_avg:141.40ms step:724/4578 train_loss:3.9427 train_time:100958ms step_avg:141.40ms step:725/4578 train_loss:3.7354 train_time:101099ms step_avg:141.40ms step:726/4578 train_loss:3.8096 train_time:101237ms step_avg:141.39ms step:727/4578 train_loss:3.7100 train_time:101377ms step_avg:141.39ms step:728/4578 train_loss:3.7304 train_time:101517ms step_avg:141.39ms step:729/4578 train_loss:3.9025 train_time:101658ms step_avg:141.39ms step:730/4578 train_loss:3.8539 train_time:101798ms step_avg:141.39ms step:731/4578 train_loss:3.8463 train_time:101938ms step_avg:141.38ms step:732/4578 train_loss:3.7350 train_time:102078ms step_avg:141.38ms step:733/4578 train_loss:3.7602 train_time:102220ms step_avg:141.38ms step:734/4578 train_loss:3.9906 train_time:102359ms step_avg:141.38ms step:735/4578 train_loss:3.7235 train_time:102499ms step_avg:141.38ms step:736/4578 train_loss:3.7909 train_time:102639ms step_avg:141.38ms step:737/4578 train_loss:3.9161 train_time:102778ms step_avg:141.37ms step:738/4578 train_loss:3.8262 train_time:102918ms step_avg:141.37ms step:739/4578 train_loss:3.7755 train_time:103064ms step_avg:141.38ms step:740/4578 train_loss:3.6665 train_time:103197ms step_avg:141.37ms step:741/4578 train_loss:4.2984 train_time:103338ms step_avg:141.36ms step:742/4578 train_loss:3.6721 train_time:103478ms step_avg:141.36ms step:743/4578 train_loss:3.7456 train_time:103617ms step_avg:141.36ms step:744/4578 train_loss:3.7504 train_time:103760ms step_avg:141.36ms step:745/4578 train_loss:3.8077 train_time:103898ms step_avg:141.36ms step:746/4578 train_loss:3.7912 train_time:104038ms step_avg:141.36ms step:747/4578 train_loss:3.7723 train_time:104177ms step_avg:141.35ms step:748/4578 train_loss:3.8023 train_time:104320ms step_avg:141.36ms step:749/4578 train_loss:3.7249 train_time:104458ms step_avg:141.35ms step:750/4578 train_loss:3.7322 train_time:104600ms step_avg:141.35ms step:750/4578 val_loss:3.7413 train_time:104653ms step_avg:141.42ms step:751/4578 train_loss:3.7654 train_time:104748ms step_avg:141.36ms step:752/4578 train_loss:3.7294 train_time:104895ms step_avg:141.37ms step:753/4578 train_loss:3.7676 train_time:105036ms step_avg:141.37ms step:754/4578 train_loss:3.7864 train_time:105174ms step_avg:141.36ms step:755/4578 train_loss:3.7541 train_time:105316ms step_avg:141.36ms step:756/4578 train_loss:3.8345 train_time:105587ms step_avg:141.54ms step:757/4578 train_loss:3.6637 train_time:105723ms step_avg:141.53ms step:758/4578 train_loss:3.9004 train_time:105862ms step_avg:141.53ms step:759/4578 train_loss:3.8174 train_time:106000ms step_avg:141.52ms step:760/4578 train_loss:3.7496 train_time:106299ms step_avg:141.73ms step:761/4578 train_loss:3.8583 train_time:106433ms step_avg:141.72ms step:762/4578 train_loss:3.5740 train_time:106571ms step_avg:141.72ms step:763/4578 train_loss:3.7251 train_time:106710ms step_avg:141.71ms step:764/4578 train_loss:3.8343 train_time:106849ms step_avg:141.71ms step:765/4578 train_loss:3.4935 train_time:106987ms step_avg:141.70ms step:766/4578 train_loss:3.9113 train_time:107127ms step_avg:141.70ms step:767/4578 train_loss:3.7620 train_time:107272ms step_avg:141.71ms step:768/4578 train_loss:3.7335 train_time:107418ms step_avg:141.71ms step:769/4578 train_loss:3.7505 train_time:107555ms step_avg:141.71ms step:770/4578 train_loss:3.7688 train_time:107697ms step_avg:141.71ms step:771/4578 train_loss:3.8212 train_time:107833ms step_avg:141.70ms step:772/4578 train_loss:4.0506 train_time:107973ms step_avg:141.70ms step:773/4578 train_loss:3.6279 train_time:108112ms step_avg:141.69ms step:774/4578 train_loss:3.8242 train_time:108255ms step_avg:141.70ms step:775/4578 train_loss:3.8007 train_time:108406ms step_avg:141.71ms step:776/4578 train_loss:3.7755 train_time:108537ms step_avg:141.69ms step:777/4578 train_loss:3.5795 train_time:108675ms step_avg:141.69ms step:778/4578 train_loss:3.5716 train_time:108820ms step_avg:141.69ms step:779/4578 train_loss:3.6456 train_time:108956ms step_avg:141.69ms step:780/4578 train_loss:3.7341 train_time:109098ms step_avg:141.69ms step:781/4578 train_loss:3.7681 train_time:109236ms step_avg:141.68ms step:782/4578 train_loss:3.8357 train_time:109377ms step_avg:141.68ms step:783/4578 train_loss:3.7473 train_time:109515ms step_avg:141.68ms step:784/4578 train_loss:3.7417 train_time:109658ms step_avg:141.68ms step:785/4578 train_loss:3.7496 train_time:109795ms step_avg:141.67ms step:786/4578 train_loss:3.7238 train_time:109935ms step_avg:141.67ms step:787/4578 train_loss:3.6227 train_time:110082ms step_avg:141.68ms step:788/4578 train_loss:3.8797 train_time:110216ms step_avg:141.67ms step:789/4578 train_loss:3.6723 train_time:110356ms step_avg:141.66ms step:790/4578 train_loss:3.7323 train_time:110495ms step_avg:141.66ms step:791/4578 train_loss:3.7919 train_time:110637ms step_avg:141.66ms step:792/4578 train_loss:3.9289 train_time:110775ms step_avg:141.66ms step:793/4578 train_loss:3.9418 train_time:110914ms step_avg:141.65ms step:794/4578 train_loss:3.6539 train_time:111055ms step_avg:141.65ms step:795/4578 train_loss:3.7689 train_time:111196ms step_avg:141.65ms step:796/4578 train_loss:3.8275 train_time:111335ms step_avg:141.65ms step:797/4578 train_loss:3.9488 train_time:111480ms step_avg:141.65ms step:798/4578 train_loss:3.6849 train_time:111616ms step_avg:141.64ms step:799/4578 train_loss:3.8352 train_time:111761ms step_avg:141.65ms step:800/4578 train_loss:3.7280 train_time:111895ms step_avg:141.64ms step:801/4578 train_loss:3.7100 train_time:112039ms step_avg:141.64ms step:802/4578 train_loss:3.8052 train_time:112176ms step_avg:141.64ms step:803/4578 train_loss:3.6691 train_time:112316ms step_avg:141.63ms step:804/4578 train_loss:3.6951 train_time:112458ms step_avg:141.64ms step:805/4578 train_loss:3.8009 train_time:112595ms step_avg:141.63ms step:806/4578 train_loss:3.6996 train_time:112743ms step_avg:141.64ms step:807/4578 train_loss:3.7179 train_time:112883ms step_avg:141.63ms step:808/4578 train_loss:3.8077 train_time:113020ms step_avg:141.63ms step:809/4578 train_loss:3.7325 train_time:113156ms step_avg:141.62ms step:810/4578 train_loss:3.6541 train_time:113296ms step_avg:141.62ms step:811/4578 train_loss:3.7359 train_time:113436ms step_avg:141.62ms step:812/4578 train_loss:3.7679 train_time:113576ms step_avg:141.62ms step:813/4578 train_loss:3.7676 train_time:113717ms step_avg:141.62ms step:814/4578 train_loss:3.8011 train_time:113856ms step_avg:141.61ms step:815/4578 train_loss:3.7427 train_time:113998ms step_avg:141.61ms step:816/4578 train_loss:3.7264 train_time:114138ms step_avg:141.61ms step:817/4578 train_loss:3.8286 train_time:114277ms step_avg:141.61ms step:818/4578 train_loss:3.9306 train_time:114418ms step_avg:141.61ms step:819/4578 train_loss:3.6941 train_time:114557ms step_avg:141.60ms step:820/4578 train_loss:3.8936 train_time:114698ms step_avg:141.60ms step:821/4578 train_loss:3.6803 train_time:114838ms step_avg:141.60ms step:822/4578 train_loss:3.7197 train_time:114978ms step_avg:141.60ms step:823/4578 train_loss:3.8419 train_time:115118ms step_avg:141.60ms step:824/4578 train_loss:3.7506 train_time:115258ms step_avg:141.59ms step:825/4578 train_loss:3.6792 train_time:115397ms step_avg:141.59ms step:826/4578 train_loss:3.7790 train_time:115537ms step_avg:141.59ms step:827/4578 train_loss:3.6725 train_time:115676ms step_avg:141.59ms step:828/4578 train_loss:3.9075 train_time:115821ms step_avg:141.59ms step:829/4578 train_loss:3.7907 train_time:115960ms step_avg:141.59ms step:830/4578 train_loss:3.8509 train_time:116096ms step_avg:141.58ms step:831/4578 train_loss:3.7049 train_time:116236ms step_avg:141.58ms step:832/4578 train_loss:3.7524 train_time:116378ms step_avg:141.58ms step:833/4578 train_loss:3.6883 train_time:116515ms step_avg:141.57ms step:834/4578 train_loss:3.8081 train_time:116656ms step_avg:141.57ms step:835/4578 train_loss:3.6584 train_time:116795ms step_avg:141.57ms step:836/4578 train_loss:3.6313 train_time:116936ms step_avg:141.57ms step:837/4578 train_loss:3.8875 train_time:117075ms step_avg:141.57ms step:838/4578 train_loss:3.5877 train_time:117218ms step_avg:141.57ms step:839/4578 train_loss:3.7572 train_time:117357ms step_avg:141.56ms step:840/4578 train_loss:3.5983 train_time:117501ms step_avg:141.57ms step:841/4578 train_loss:3.6439 train_time:117637ms step_avg:141.56ms step:842/4578 train_loss:3.7336 train_time:117783ms step_avg:141.57ms step:843/4578 train_loss:3.7499 train_time:117922ms step_avg:141.56ms step:844/4578 train_loss:3.7506 train_time:118065ms step_avg:141.56ms step:845/4578 train_loss:3.6016 train_time:118195ms step_avg:141.55ms step:846/4578 train_loss:3.8398 train_time:118337ms step_avg:141.55ms step:847/4578 train_loss:3.7058 train_time:118481ms step_avg:141.55ms step:848/4578 train_loss:3.6595 train_time:118616ms step_avg:141.55ms step:849/4578 train_loss:3.7995 train_time:118757ms step_avg:141.55ms step:850/4578 train_loss:3.6673 train_time:118901ms step_avg:141.55ms step:851/4578 train_loss:3.6198 train_time:119037ms step_avg:141.54ms step:852/4578 train_loss:3.9144 train_time:119176ms step_avg:141.54ms step:853/4578 train_loss:3.6226 train_time:119317ms step_avg:141.54ms step:854/4578 train_loss:3.7339 train_time:119457ms step_avg:141.54ms step:855/4578 train_loss:3.8178 train_time:119597ms step_avg:141.53ms step:856/4578 train_loss:3.7012 train_time:119736ms step_avg:141.53ms step:857/4578 train_loss:3.7228 train_time:119876ms step_avg:141.53ms step:858/4578 train_loss:3.7709 train_time:120017ms step_avg:141.53ms step:859/4578 train_loss:3.6562 train_time:120158ms step_avg:141.53ms step:860/4578 train_loss:3.7401 train_time:120296ms step_avg:141.52ms step:861/4578 train_loss:3.7618 train_time:120436ms step_avg:141.52ms step:862/4578 train_loss:3.8138 train_time:120578ms step_avg:141.52ms step:863/4578 train_loss:3.7648 train_time:120716ms step_avg:141.52ms step:864/4578 train_loss:3.7478 train_time:120855ms step_avg:141.52ms step:865/4578 train_loss:3.5708 train_time:120995ms step_avg:141.51ms step:866/4578 train_loss:3.7661 train_time:121135ms step_avg:141.51ms step:867/4578 train_loss:4.0291 train_time:121274ms step_avg:141.51ms step:868/4578 train_loss:3.6238 train_time:121414ms step_avg:141.51ms step:869/4578 train_loss:3.8016 train_time:121555ms step_avg:141.51ms step:870/4578 train_loss:3.7892 train_time:121697ms step_avg:141.51ms step:871/4578 train_loss:3.6214 train_time:121835ms step_avg:141.50ms step:872/4578 train_loss:3.5928 train_time:121975ms step_avg:141.50ms step:873/4578 train_loss:3.8298 train_time:122116ms step_avg:141.50ms step:874/4578 train_loss:3.6211 train_time:122257ms step_avg:141.50ms step:875/4578 train_loss:3.3500 train_time:122394ms step_avg:141.50ms step:875/4578 val_loss:3.6978 train_time:122450ms step_avg:141.56ms step:876/4578 train_loss:3.8152 train_time:122545ms step_avg:141.51ms step:877/4578 train_loss:3.6223 train_time:122689ms step_avg:141.51ms step:878/4578 train_loss:3.7956 train_time:122828ms step_avg:141.51ms step:879/4578 train_loss:3.6591 train_time:122967ms step_avg:141.50ms step:880/4578 train_loss:3.8297 train_time:123105ms step_avg:141.50ms step:881/4578 train_loss:3.4950 train_time:123244ms step_avg:141.50ms step:882/4578 train_loss:3.6686 train_time:123382ms step_avg:141.49ms step:883/4578 train_loss:3.8641 train_time:123527ms step_avg:141.50ms step:884/4578 train_loss:4.0165 train_time:123671ms step_avg:141.50ms step:885/4578 train_loss:3.7408 train_time:123809ms step_avg:141.50ms step:886/4578 train_loss:3.6591 train_time:123952ms step_avg:141.50ms step:887/4578 train_loss:3.7479 train_time:124088ms step_avg:141.49ms step:888/4578 train_loss:4.2534 train_time:124228ms step_avg:141.49ms step:889/4578 train_loss:4.0090 train_time:124368ms step_avg:141.49ms step:890/4578 train_loss:3.6903 train_time:124508ms step_avg:141.49ms step:891/4578 train_loss:3.7046 train_time:124649ms step_avg:141.49ms step:892/4578 train_loss:3.5338 train_time:124789ms step_avg:141.48ms step:893/4578 train_loss:3.8758 train_time:124929ms step_avg:141.48ms step:894/4578 train_loss:3.5983 train_time:125068ms step_avg:141.48ms step:895/4578 train_loss:3.8631 train_time:125208ms step_avg:141.48ms step:896/4578 train_loss:3.8655 train_time:125347ms step_avg:141.48ms step:897/4578 train_loss:3.6685 train_time:125488ms step_avg:141.47ms step:898/4578 train_loss:3.7104 train_time:125630ms step_avg:141.47ms step:899/4578 train_loss:3.7657 train_time:125769ms step_avg:141.47ms step:900/4578 train_loss:3.6545 train_time:125909ms step_avg:141.47ms step:901/4578 train_loss:3.5948 train_time:126049ms step_avg:141.47ms step:902/4578 train_loss:3.8001 train_time:126188ms step_avg:141.47ms step:903/4578 train_loss:3.8069 train_time:126329ms step_avg:141.47ms step:904/4578 train_loss:3.7088 train_time:126468ms step_avg:141.46ms step:905/4578 train_loss:3.6732 train_time:126609ms step_avg:141.46ms step:906/4578 train_loss:3.6674 train_time:126748ms step_avg:141.46ms step:907/4578 train_loss:3.8981 train_time:126888ms step_avg:141.46ms step:908/4578 train_loss:3.6846 train_time:127029ms step_avg:141.46ms step:909/4578 train_loss:3.7247 train_time:127168ms step_avg:141.45ms step:910/4578 train_loss:3.6310 train_time:127309ms step_avg:141.45ms step:911/4578 train_loss:3.7181 train_time:127449ms step_avg:141.45ms step:912/4578 train_loss:3.7840 train_time:127588ms step_avg:141.45ms step:913/4578 train_loss:3.7932 train_time:127728ms step_avg:141.45ms step:914/4578 train_loss:3.6585 train_time:127870ms step_avg:141.45ms step:915/4578 train_loss:3.9084 train_time:128008ms step_avg:141.44ms step:916/4578 train_loss:3.7102 train_time:128148ms step_avg:141.44ms step:917/4578 train_loss:3.7956 train_time:128289ms step_avg:141.44ms step:918/4578 train_loss:3.7704 train_time:128427ms step_avg:141.44ms step:919/4578 train_loss:5.0107 train_time:128567ms step_avg:141.44ms step:920/4578 train_loss:3.6919 train_time:128708ms step_avg:141.44ms step:921/4578 train_loss:3.7448 train_time:128847ms step_avg:141.44ms step:922/4578 train_loss:3.7074 train_time:128993ms step_avg:141.44ms step:923/4578 train_loss:3.7567 train_time:129129ms step_avg:141.43ms step:924/4578 train_loss:3.7685 train_time:129279ms step_avg:141.44ms step:925/4578 train_loss:3.8614 train_time:129409ms step_avg:141.43ms step:926/4578 train_loss:3.8323 train_time:129548ms step_avg:141.43ms step:927/4578 train_loss:3.7308 train_time:129688ms step_avg:141.43ms step:928/4578 train_loss:3.7172 train_time:129837ms step_avg:141.44ms step:929/4578 train_loss:3.9443 train_time:129969ms step_avg:141.42ms step:930/4578 train_loss:3.7865 train_time:130108ms step_avg:141.42ms step:931/4578 train_loss:3.5780 train_time:130246ms step_avg:141.42ms step:932/4578 train_loss:3.6679 train_time:130393ms step_avg:141.42ms step:933/4578 train_loss:3.8540 train_time:130527ms step_avg:141.42ms step:934/4578 train_loss:3.5757 train_time:130666ms step_avg:141.41ms step:935/4578 train_loss:3.7450 train_time:130806ms step_avg:141.41ms step:936/4578 train_loss:3.6179 train_time:130946ms step_avg:141.41ms step:937/4578 train_loss:3.6870 train_time:131086ms step_avg:141.41ms step:938/4578 train_loss:3.7809 train_time:131226ms step_avg:141.41ms step:939/4578 train_loss:3.7106 train_time:131366ms step_avg:141.41ms step:940/4578 train_loss:3.8764 train_time:131507ms step_avg:141.41ms step:941/4578 train_loss:3.6555 train_time:131647ms step_avg:141.40ms step:942/4578 train_loss:3.7159 train_time:131798ms step_avg:141.41ms step:943/4578 train_loss:3.5164 train_time:131927ms step_avg:141.40ms step:944/4578 train_loss:3.8656 train_time:132067ms step_avg:141.40ms step:945/4578 train_loss:3.5808 train_time:132402ms step_avg:141.61ms step:946/4578 train_loss:3.6004 train_time:132537ms step_avg:141.60ms step:947/4578 train_loss:5.2230 train_time:132678ms step_avg:141.60ms step:948/4578 train_loss:3.7677 train_time:132819ms step_avg:141.60ms step:949/4578 train_loss:3.6711 train_time:132957ms step_avg:141.59ms step:950/4578 train_loss:3.5639 train_time:133255ms step_avg:141.76ms step:951/4578 train_loss:3.6212 train_time:133391ms step_avg:141.75ms step:952/4578 train_loss:3.5794 train_time:133529ms step_avg:141.75ms step:953/4578 train_loss:3.6546 train_time:133667ms step_avg:141.75ms step:954/4578 train_loss:3.7269 train_time:133805ms step_avg:141.74ms step:955/4578 train_loss:3.6128 train_time:133943ms step_avg:141.74ms step:956/4578 train_loss:3.6425 train_time:134084ms step_avg:141.74ms step:957/4578 train_loss:3.6132 train_time:134227ms step_avg:141.74ms step:958/4578 train_loss:3.6732 train_time:134369ms step_avg:141.74ms step:959/4578 train_loss:3.6684 train_time:134510ms step_avg:141.74ms step:960/4578 train_loss:3.6804 train_time:134649ms step_avg:141.74ms step:961/4578 train_loss:3.5684 train_time:134788ms step_avg:141.73ms step:962/4578 train_loss:3.8267 train_time:134927ms step_avg:141.73ms step:963/4578 train_loss:3.7734 train_time:135065ms step_avg:141.73ms step:964/4578 train_loss:3.7342 train_time:135209ms step_avg:141.73ms step:965/4578 train_loss:3.6170 train_time:135350ms step_avg:141.73ms step:966/4578 train_loss:3.6544 train_time:135489ms step_avg:141.72ms step:967/4578 train_loss:3.8759 train_time:135630ms step_avg:141.72ms step:968/4578 train_loss:3.7072 train_time:135778ms step_avg:141.73ms step:969/4578 train_loss:3.6955 train_time:135907ms step_avg:141.72ms step:970/4578 train_loss:3.7498 train_time:136046ms step_avg:141.71ms step:971/4578 train_loss:3.5638 train_time:136187ms step_avg:141.71ms step:972/4578 train_loss:3.7149 train_time:136329ms step_avg:141.71ms step:973/4578 train_loss:3.6731 train_time:136469ms step_avg:141.71ms step:974/4578 train_loss:3.7132 train_time:136611ms step_avg:141.71ms step:975/4578 train_loss:3.7848 train_time:136748ms step_avg:141.71ms step:976/4578 train_loss:3.6614 train_time:136887ms step_avg:141.71ms step:977/4578 train_loss:3.8568 train_time:137032ms step_avg:141.71ms step:978/4578 train_loss:3.7371 train_time:137167ms step_avg:141.70ms step:979/4578 train_loss:3.5739 train_time:137308ms step_avg:141.70ms step:980/4578 train_loss:3.8622 train_time:137448ms step_avg:141.70ms step:981/4578 train_loss:3.5907 train_time:137587ms step_avg:141.70ms step:982/4578 train_loss:3.7571 train_time:137726ms step_avg:141.69ms step:983/4578 train_loss:3.7346 train_time:137865ms step_avg:141.69ms step:984/4578 train_loss:3.7462 train_time:138006ms step_avg:141.69ms step:985/4578 train_loss:3.6757 train_time:138148ms step_avg:141.69ms step:986/4578 train_loss:3.7670 train_time:138288ms step_avg:141.69ms step:987/4578 train_loss:3.5914 train_time:138429ms step_avg:141.69ms step:988/4578 train_loss:3.6702 train_time:138569ms step_avg:141.69ms step:989/4578 train_loss:3.6828 train_time:138708ms step_avg:141.68ms step:990/4578 train_loss:3.6043 train_time:138848ms step_avg:141.68ms step:991/4578 train_loss:3.8166 train_time:138989ms step_avg:141.68ms step:992/4578 train_loss:3.6429 train_time:139129ms step_avg:141.68ms step:993/4578 train_loss:3.6120 train_time:139267ms step_avg:141.68ms step:994/4578 train_loss:3.6887 train_time:139408ms step_avg:141.67ms step:995/4578 train_loss:3.7688 train_time:139548ms step_avg:141.67ms step:996/4578 train_loss:3.7209 train_time:139687ms step_avg:141.67ms step:997/4578 train_loss:3.6281 train_time:139827ms step_avg:141.67ms step:998/4578 train_loss:3.9752 train_time:139966ms step_avg:141.67ms step:999/4578 train_loss:3.6410 train_time:140106ms step_avg:141.66ms step:1000/4578 train_loss:3.7608 train_time:140246ms step_avg:141.66ms step:1000/4578 val_loss:3.6582 train_time:140302ms step_avg:141.72ms step:1001/4578 train_loss:3.6352 train_time:140400ms step_avg:141.67ms step:1002/4578 train_loss:3.6825 train_time:140547ms step_avg:141.68ms step:1003/4578 train_loss:3.5670 train_time:140685ms step_avg:141.68ms step:1004/4578 train_loss:3.7536 train_time:140824ms step_avg:141.67ms step:1005/4578 train_loss:3.8023 train_time:140962ms step_avg:141.67ms step:1006/4578 train_loss:3.5781 train_time:141100ms step_avg:141.67ms step:1007/4578 train_loss:3.6578 train_time:141240ms step_avg:141.67ms step:1008/4578 train_loss:3.6229 train_time:141380ms step_avg:141.66ms step:1009/4578 train_loss:3.7468 train_time:141527ms step_avg:141.67ms step:1010/4578 train_loss:3.8481 train_time:141675ms step_avg:141.68ms step:1011/4578 train_loss:3.7450 train_time:141807ms step_avg:141.67ms step:1012/4578 train_loss:3.7043 train_time:141946ms step_avg:141.66ms step:1013/4578 train_loss:3.5657 train_time:142085ms step_avg:141.66ms step:1014/4578 train_loss:3.7026 train_time:142226ms step_avg:141.66ms step:1015/4578 train_loss:3.8223 train_time:142365ms step_avg:141.66ms step:1016/4578 train_loss:3.5259 train_time:142508ms step_avg:141.66ms step:1017/4578 train_loss:3.6235 train_time:142649ms step_avg:141.66ms step:1018/4578 train_loss:3.6255 train_time:142788ms step_avg:141.66ms step:1019/4578 train_loss:3.5626 train_time:142929ms step_avg:141.65ms step:1020/4578 train_loss:3.7098 train_time:143068ms step_avg:141.65ms step:1021/4578 train_loss:3.6172 train_time:143211ms step_avg:141.65ms step:1022/4578 train_loss:3.5575 train_time:143347ms step_avg:141.65ms step:1023/4578 train_loss:3.6594 train_time:143488ms step_avg:141.65ms step:1024/4578 train_loss:3.6878 train_time:143629ms step_avg:141.65ms step:1025/4578 train_loss:3.6656 train_time:143769ms step_avg:141.64ms step:1026/4578 train_loss:3.6737 train_time:143909ms step_avg:141.64ms step:1027/4578 train_loss:3.8391 train_time:144048ms step_avg:141.64ms step:1028/4578 train_loss:3.5236 train_time:144188ms step_avg:141.64ms step:1029/4578 train_loss:3.5841 train_time:144327ms step_avg:141.64ms step:1030/4578 train_loss:3.5362 train_time:144467ms step_avg:141.63ms step:1031/4578 train_loss:3.7002 train_time:144608ms step_avg:141.63ms step:1032/4578 train_loss:3.6893 train_time:144748ms step_avg:141.63ms step:1033/4578 train_loss:3.8668 train_time:144887ms step_avg:141.63ms step:1034/4578 train_loss:3.6883 train_time:145027ms step_avg:141.63ms step:1035/4578 train_loss:3.6151 train_time:145166ms step_avg:141.63ms step:1036/4578 train_loss:3.6223 train_time:145307ms step_avg:141.62ms step:1037/4578 train_loss:3.6861 train_time:145447ms step_avg:141.62ms step:1038/4578 train_loss:3.9929 train_time:145586ms step_avg:141.62ms step:1039/4578 train_loss:3.8104 train_time:145726ms step_avg:141.62ms step:1040/4578 train_loss:3.7076 train_time:145867ms step_avg:141.62ms step:1041/4578 train_loss:3.6057 train_time:146006ms step_avg:141.62ms step:1042/4578 train_loss:3.6737 train_time:146146ms step_avg:141.61ms step:1043/4578 train_loss:3.7106 train_time:146285ms step_avg:141.61ms step:1044/4578 train_loss:3.6396 train_time:146426ms step_avg:141.61ms step:1045/4578 train_loss:3.6465 train_time:146566ms step_avg:141.61ms step:1046/4578 train_loss:3.7258 train_time:146706ms step_avg:141.61ms step:1047/4578 train_loss:3.6321 train_time:146847ms step_avg:141.61ms step:1048/4578 train_loss:3.8388 train_time:146986ms step_avg:141.60ms step:1049/4578 train_loss:3.6900 train_time:147126ms step_avg:141.60ms step:1050/4578 train_loss:3.6142 train_time:147264ms step_avg:141.60ms step:1051/4578 train_loss:3.5877 train_time:147405ms step_avg:141.60ms step:1052/4578 train_loss:3.7071 train_time:147545ms step_avg:141.60ms step:1053/4578 train_loss:3.5842 train_time:147684ms step_avg:141.60ms step:1054/4578 train_loss:3.9012 train_time:147826ms step_avg:141.60ms step:1055/4578 train_loss:3.7374 train_time:147966ms step_avg:141.59ms step:1056/4578 train_loss:3.5934 train_time:148106ms step_avg:141.59ms step:1057/4578 train_loss:3.6991 train_time:148245ms step_avg:141.59ms step:1058/4578 train_loss:3.7724 train_time:148385ms step_avg:141.59ms step:1059/4578 train_loss:3.4989 train_time:148526ms step_avg:141.59ms step:1060/4578 train_loss:3.6286 train_time:148667ms step_avg:141.59ms step:1061/4578 train_loss:3.6417 train_time:148808ms step_avg:141.59ms step:1062/4578 train_loss:3.6121 train_time:148945ms step_avg:141.58ms step:1063/4578 train_loss:3.5806 train_time:149085ms step_avg:141.58ms step:1064/4578 train_loss:3.6800 train_time:149225ms step_avg:141.58ms step:1065/4578 train_loss:3.5874 train_time:149364ms step_avg:141.58ms step:1066/4578 train_loss:3.5686 train_time:149504ms step_avg:141.58ms step:1067/4578 train_loss:3.5939 train_time:149644ms step_avg:141.57ms step:1068/4578 train_loss:3.5068 train_time:149784ms step_avg:141.57ms step:1069/4578 train_loss:3.6184 train_time:149925ms step_avg:141.57ms step:1070/4578 train_loss:3.4944 train_time:150063ms step_avg:141.57ms step:1071/4578 train_loss:3.7523 train_time:150204ms step_avg:141.57ms step:1072/4578 train_loss:3.6990 train_time:150344ms step_avg:141.57ms step:1073/4578 train_loss:3.6501 train_time:150501ms step_avg:141.58ms step:1074/4578 train_loss:3.7170 train_time:150627ms step_avg:141.57ms step:1075/4578 train_loss:3.6614 train_time:150767ms step_avg:141.57ms step:1076/4578 train_loss:3.5973 train_time:150906ms step_avg:141.56ms step:1077/4578 train_loss:3.9871 train_time:151047ms step_avg:141.56ms step:1078/4578 train_loss:3.6670 train_time:151187ms step_avg:141.56ms step:1079/4578 train_loss:3.3487 train_time:151327ms step_avg:141.56ms step:1080/4578 train_loss:3.7323 train_time:151477ms step_avg:141.57ms step:1081/4578 train_loss:3.6441 train_time:151605ms step_avg:141.55ms step:1082/4578 train_loss:3.7148 train_time:151745ms step_avg:141.55ms step:1083/4578 train_loss:3.8149 train_time:151899ms step_avg:141.56ms step:1084/4578 train_loss:3.7088 train_time:152025ms step_avg:141.55ms step:1085/4578 train_loss:3.6851 train_time:152164ms step_avg:141.55ms step:1086/4578 train_loss:3.6431 train_time:152313ms step_avg:141.56ms step:1087/4578 train_loss:3.8426 train_time:152446ms step_avg:141.55ms step:1088/4578 train_loss:3.7307 train_time:152584ms step_avg:141.54ms step:1089/4578 train_loss:3.5655 train_time:152733ms step_avg:141.55ms step:1090/4578 train_loss:3.5838 train_time:152865ms step_avg:141.54ms step:1091/4578 train_loss:3.6981 train_time:153005ms step_avg:141.54ms step:1092/4578 train_loss:3.5021 train_time:153157ms step_avg:141.55ms step:1093/4578 train_loss:3.6991 train_time:153284ms step_avg:141.54ms step:1094/4578 train_loss:3.8268 train_time:153424ms step_avg:141.54ms step:1095/4578 train_loss:3.6752 train_time:153564ms step_avg:141.53ms step:1096/4578 train_loss:3.6133 train_time:153704ms step_avg:141.53ms step:1097/4578 train_loss:3.6427 train_time:153844ms step_avg:141.53ms step:1098/4578 train_loss:3.6912 train_time:153984ms step_avg:141.53ms step:1099/4578 train_loss:3.7605 train_time:154124ms step_avg:141.53ms step:1100/4578 train_loss:3.7196 train_time:154264ms step_avg:141.53ms step:1101/4578 train_loss:3.6495 train_time:154404ms step_avg:141.53ms step:1102/4578 train_loss:3.5006 train_time:154546ms step_avg:141.53ms step:1103/4578 train_loss:3.5759 train_time:154684ms step_avg:141.52ms step:1104/4578 train_loss:3.6590 train_time:154824ms step_avg:141.52ms step:1105/4578 train_loss:3.5323 train_time:154963ms step_avg:141.52ms step:1106/4578 train_loss:4.2861 train_time:155104ms step_avg:141.52ms step:1107/4578 train_loss:3.4408 train_time:155246ms step_avg:141.52ms step:1108/4578 train_loss:3.7752 train_time:155383ms step_avg:141.51ms step:1109/4578 train_loss:3.5644 train_time:155524ms step_avg:141.51ms step:1110/4578 train_loss:3.7105 train_time:155663ms step_avg:141.51ms step:1111/4578 train_loss:3.6428 train_time:155803ms step_avg:141.51ms step:1112/4578 train_loss:3.6850 train_time:155944ms step_avg:141.51ms step:1113/4578 train_loss:3.7758 train_time:156084ms step_avg:141.51ms step:1114/4578 train_loss:3.6279 train_time:156225ms step_avg:141.51ms step:1115/4578 train_loss:3.5735 train_time:156364ms step_avg:141.51ms step:1116/4578 train_loss:3.4780 train_time:156506ms step_avg:141.51ms step:1117/4578 train_loss:3.6494 train_time:156646ms step_avg:141.50ms step:1118/4578 train_loss:3.7977 train_time:156785ms step_avg:141.50ms step:1119/4578 train_loss:3.8396 train_time:156925ms step_avg:141.50ms step:1120/4578 train_loss:3.6718 train_time:157064ms step_avg:141.50ms step:1121/4578 train_loss:3.7039 train_time:157204ms step_avg:141.50ms step:1122/4578 train_loss:3.6047 train_time:157344ms step_avg:141.50ms step:1123/4578 train_loss:3.6580 train_time:157484ms step_avg:141.49ms step:1124/4578 train_loss:3.8007 train_time:157624ms step_avg:141.49ms step:1125/4578 train_loss:3.5756 train_time:157763ms step_avg:141.49ms step:1125/4578 val_loss:3.6322 train_time:157819ms step_avg:141.54ms step:1126/4578 train_loss:3.4854 train_time:157911ms step_avg:141.50ms step:1127/4578 train_loss:3.6905 train_time:158064ms step_avg:141.51ms step:1128/4578 train_loss:3.9061 train_time:158201ms step_avg:141.50ms step:1129/4578 train_loss:3.4439 train_time:158340ms step_avg:141.50ms step:1130/4578 train_loss:3.7705 train_time:158479ms step_avg:141.50ms step:1131/4578 train_loss:3.5985 train_time:158618ms step_avg:141.50ms step:1132/4578 train_loss:3.6307 train_time:158758ms step_avg:141.50ms step:1133/4578 train_loss:3.5839 train_time:158902ms step_avg:141.50ms step:1134/4578 train_loss:3.7433 train_time:159177ms step_avg:141.62ms step:1135/4578 train_loss:3.6760 train_time:159315ms step_avg:141.61ms step:1136/4578 train_loss:3.7286 train_time:159455ms step_avg:141.61ms step:1137/4578 train_loss:3.7612 train_time:159594ms step_avg:141.61ms step:1138/4578 train_loss:3.6792 train_time:159733ms step_avg:141.61ms step:1139/4578 train_loss:3.5741 train_time:159872ms step_avg:141.60ms step:1140/4578 train_loss:3.8775 train_time:160166ms step_avg:141.74ms step:1141/4578 train_loss:3.6816 train_time:160301ms step_avg:141.73ms step:1142/4578 train_loss:3.7838 train_time:160441ms step_avg:141.73ms step:1143/4578 train_loss:3.6683 train_time:160579ms step_avg:141.73ms step:1144/4578 train_loss:3.5828 train_time:160718ms step_avg:141.73ms step:1145/4578 train_loss:3.6899 train_time:160858ms step_avg:141.72ms step:1146/4578 train_loss:3.8077 train_time:160998ms step_avg:141.72ms step:1147/4578 train_loss:3.7774 train_time:161146ms step_avg:141.73ms step:1148/4578 train_loss:3.6951 train_time:161286ms step_avg:141.73ms step:1149/4578 train_loss:3.7130 train_time:161425ms step_avg:141.73ms step:1150/4578 train_loss:3.5654 train_time:161564ms step_avg:141.72ms step:1151/4578 train_loss:3.5849 train_time:161704ms step_avg:141.72ms step:1152/4578 train_loss:3.5602 train_time:161843ms step_avg:141.72ms step:1153/4578 train_loss:3.7018 train_time:161984ms step_avg:141.72ms step:1154/4578 train_loss:3.6719 train_time:162128ms step_avg:141.72ms step:1155/4578 train_loss:3.7332 train_time:162268ms step_avg:141.72ms step:1156/4578 train_loss:3.5861 train_time:162408ms step_avg:141.72ms step:1157/4578 train_loss:3.7535 train_time:162547ms step_avg:141.71ms step:1158/4578 train_loss:3.7068 train_time:162687ms step_avg:141.71ms step:1159/4578 train_loss:3.5250 train_time:162827ms step_avg:141.71ms step:1160/4578 train_loss:3.5588 train_time:162966ms step_avg:141.71ms step:1161/4578 train_loss:3.5395 train_time:163117ms step_avg:141.72ms step:1162/4578 train_loss:3.3651 train_time:163254ms step_avg:141.71ms step:1163/4578 train_loss:3.6669 train_time:163396ms step_avg:141.71ms step:1164/4578 train_loss:3.6286 train_time:163528ms step_avg:141.71ms step:1165/4578 train_loss:3.4969 train_time:163667ms step_avg:141.70ms step:1166/4578 train_loss:3.4937 train_time:163810ms step_avg:141.70ms step:1167/4578 train_loss:3.6017 train_time:163950ms step_avg:141.70ms step:1168/4578 train_loss:3.6073 train_time:164091ms step_avg:141.70ms step:1169/4578 train_loss:3.9303 train_time:164236ms step_avg:141.70ms step:1170/4578 train_loss:3.6132 train_time:164369ms step_avg:141.70ms step:1171/4578 train_loss:3.6295 train_time:164514ms step_avg:141.70ms step:1172/4578 train_loss:3.5353 train_time:164651ms step_avg:141.70ms step:1173/4578 train_loss:3.6269 train_time:164791ms step_avg:141.70ms step:1174/4578 train_loss:3.7654 train_time:164929ms step_avg:141.69ms step:1175/4578 train_loss:3.6030 train_time:165070ms step_avg:141.69ms step:1176/4578 train_loss:3.6258 train_time:165208ms step_avg:141.69ms step:1177/4578 train_loss:3.6770 train_time:165348ms step_avg:141.69ms step:1178/4578 train_loss:3.6654 train_time:165500ms step_avg:141.70ms step:1179/4578 train_loss:3.7144 train_time:165628ms step_avg:141.68ms step:1180/4578 train_loss:3.6245 train_time:165767ms step_avg:141.68ms step:1181/4578 train_loss:3.6252 train_time:165907ms step_avg:141.68ms step:1182/4578 train_loss:3.5733 train_time:166048ms step_avg:141.68ms step:1183/4578 train_loss:3.6340 train_time:166190ms step_avg:141.68ms step:1184/4578 train_loss:3.5592 train_time:166328ms step_avg:141.68ms step:1185/4578 train_loss:3.7206 train_time:166467ms step_avg:141.67ms step:1186/4578 train_loss:3.7819 train_time:166607ms step_avg:141.67ms step:1187/4578 train_loss:3.5765 train_time:166747ms step_avg:141.67ms step:1188/4578 train_loss:3.6367 train_time:166889ms step_avg:141.67ms step:1189/4578 train_loss:3.6619 train_time:167028ms step_avg:141.67ms step:1190/4578 train_loss:3.5010 train_time:167167ms step_avg:141.67ms step:1191/4578 train_loss:3.6774 train_time:167308ms step_avg:141.67ms step:1192/4578 train_loss:3.8146 train_time:167448ms step_avg:141.67ms step:1193/4578 train_loss:3.6176 train_time:167589ms step_avg:141.66ms step:1194/4578 train_loss:3.5098 train_time:167729ms step_avg:141.66ms step:1195/4578 train_loss:3.7958 train_time:167869ms step_avg:141.66ms step:1196/4578 train_loss:3.5979 train_time:168007ms step_avg:141.66ms step:1197/4578 train_loss:3.6111 train_time:168147ms step_avg:141.66ms step:1198/4578 train_loss:3.5046 train_time:168288ms step_avg:141.66ms step:1199/4578 train_loss:3.5179 train_time:168428ms step_avg:141.66ms step:1200/4578 train_loss:3.5713 train_time:168568ms step_avg:141.65ms step:1201/4578 train_loss:3.6570 train_time:168707ms step_avg:141.65ms step:1202/4578 train_loss:3.7290 train_time:168846ms step_avg:141.65ms step:1203/4578 train_loss:3.8023 train_time:168986ms step_avg:141.65ms step:1204/4578 train_loss:3.6453 train_time:169127ms step_avg:141.65ms step:1205/4578 train_loss:3.5672 train_time:169267ms step_avg:141.65ms step:1206/4578 train_loss:3.6550 train_time:169407ms step_avg:141.64ms step:1207/4578 train_loss:3.7019 train_time:169547ms step_avg:141.64ms step:1208/4578 train_loss:3.7512 train_time:169688ms step_avg:141.64ms step:1209/4578 train_loss:3.6240 train_time:169827ms step_avg:141.64ms step:1210/4578 train_loss:3.4863 train_time:169971ms step_avg:141.64ms step:1211/4578 train_loss:3.5313 train_time:170109ms step_avg:141.64ms step:1212/4578 train_loss:3.6298 train_time:170248ms step_avg:141.64ms step:1213/4578 train_loss:3.6454 train_time:170388ms step_avg:141.64ms step:1214/4578 train_loss:3.6783 train_time:170528ms step_avg:141.63ms step:1215/4578 train_loss:3.5734 train_time:170667ms step_avg:141.63ms step:1216/4578 train_loss:3.6243 train_time:170808ms step_avg:141.63ms step:1217/4578 train_loss:3.5689 train_time:170948ms step_avg:141.63ms step:1218/4578 train_loss:3.5656 train_time:171088ms step_avg:141.63ms step:1219/4578 train_loss:3.6562 train_time:171228ms step_avg:141.63ms step:1220/4578 train_loss:3.5091 train_time:171368ms step_avg:141.63ms step:1221/4578 train_loss:3.7289 train_time:171508ms step_avg:141.62ms step:1222/4578 train_loss:3.7540 train_time:171648ms step_avg:141.62ms step:1223/4578 train_loss:3.6798 train_time:171788ms step_avg:141.62ms step:1224/4578 train_loss:3.5241 train_time:171932ms step_avg:141.62ms step:1225/4578 train_loss:3.5184 train_time:172066ms step_avg:141.62ms step:1226/4578 train_loss:3.6007 train_time:172211ms step_avg:141.62ms step:1227/4578 train_loss:3.5874 train_time:172347ms step_avg:141.62ms step:1228/4578 train_loss:3.5204 train_time:172494ms step_avg:141.62ms step:1229/4578 train_loss:3.6876 train_time:172635ms step_avg:141.62ms step:1230/4578 train_loss:3.6101 train_time:172773ms step_avg:141.62ms step:1231/4578 train_loss:3.6717 train_time:172906ms step_avg:141.61ms step:1232/4578 train_loss:3.8255 train_time:173045ms step_avg:141.61ms step:1233/4578 train_loss:3.7248 train_time:173186ms step_avg:141.61ms step:1234/4578 train_loss:3.6570 train_time:173327ms step_avg:141.61ms step:1235/4578 train_loss:3.8160 train_time:173466ms step_avg:141.61ms step:1236/4578 train_loss:3.5767 train_time:173607ms step_avg:141.60ms step:1237/4578 train_loss:3.5434 train_time:173747ms step_avg:141.60ms step:1238/4578 train_loss:3.4940 train_time:173888ms step_avg:141.60ms step:1239/4578 train_loss:3.5696 train_time:174027ms step_avg:141.60ms step:1240/4578 train_loss:3.5774 train_time:174169ms step_avg:141.60ms step:1241/4578 train_loss:3.6227 train_time:174309ms step_avg:141.60ms step:1242/4578 train_loss:3.6704 train_time:174452ms step_avg:141.60ms step:1243/4578 train_loss:3.5436 train_time:174589ms step_avg:141.60ms step:1244/4578 train_loss:3.6347 train_time:174729ms step_avg:141.60ms step:1245/4578 train_loss:3.6592 train_time:174868ms step_avg:141.59ms step:1246/4578 train_loss:3.6563 train_time:175008ms step_avg:141.59ms step:1247/4578 train_loss:3.4886 train_time:175148ms step_avg:141.59ms step:1248/4578 train_loss:3.6206 train_time:175288ms step_avg:141.59ms step:1249/4578 train_loss:3.6889 train_time:175429ms step_avg:141.59ms step:1250/4578 train_loss:3.6523 train_time:175568ms step_avg:141.59ms step:1250/4578 val_loss:3.6037 train_time:175624ms step_avg:141.63ms step:1251/4578 train_loss:3.5459 train_time:175714ms step_avg:141.59ms step:1252/4578 train_loss:3.7598 train_time:175865ms step_avg:141.60ms step:1253/4578 train_loss:3.6264 train_time:176003ms step_avg:141.60ms step:1254/4578 train_loss:3.5465 train_time:176142ms step_avg:141.59ms step:1255/4578 train_loss:3.6910 train_time:176286ms step_avg:141.60ms step:1256/4578 train_loss:3.7504 train_time:176419ms step_avg:141.59ms step:1257/4578 train_loss:3.5632 train_time:176558ms step_avg:141.59ms step:1258/4578 train_loss:3.5886 train_time:176701ms step_avg:141.59ms step:1259/4578 train_loss:3.6134 train_time:176841ms step_avg:141.59ms step:1260/4578 train_loss:3.5924 train_time:176981ms step_avg:141.58ms step:1261/4578 train_loss:3.4444 train_time:177121ms step_avg:141.58ms step:1262/4578 train_loss:3.5490 train_time:177266ms step_avg:141.59ms step:1263/4578 train_loss:3.6168 train_time:177400ms step_avg:141.58ms step:1264/4578 train_loss:3.4625 train_time:177539ms step_avg:141.58ms step:1265/4578 train_loss:3.6865 train_time:177679ms step_avg:141.58ms step:1266/4578 train_loss:3.6639 train_time:177820ms step_avg:141.58ms step:1267/4578 train_loss:3.6704 train_time:177959ms step_avg:141.57ms step:1268/4578 train_loss:3.6137 train_time:178099ms step_avg:141.57ms step:1269/4578 train_loss:3.6483 train_time:178238ms step_avg:141.57ms step:1270/4578 train_loss:3.5042 train_time:178377ms step_avg:141.57ms step:1271/4578 train_loss:3.3599 train_time:178516ms step_avg:141.57ms step:1272/4578 train_loss:3.6359 train_time:178656ms step_avg:141.57ms step:1273/4578 train_loss:3.5964 train_time:178798ms step_avg:141.57ms step:1274/4578 train_loss:3.6516 train_time:178939ms step_avg:141.57ms step:1275/4578 train_loss:3.5992 train_time:179081ms step_avg:141.57ms step:1276/4578 train_loss:3.6945 train_time:179217ms step_avg:141.56ms step:1277/4578 train_loss:3.7105 train_time:179356ms step_avg:141.56ms step:1278/4578 train_loss:3.6640 train_time:179496ms step_avg:141.56ms step:1279/4578 train_loss:3.6662 train_time:179636ms step_avg:141.56ms step:1280/4578 train_loss:3.4922 train_time:179777ms step_avg:141.56ms step:1281/4578 train_loss:3.6151 train_time:179918ms step_avg:141.56ms step:1282/4578 train_loss:3.6791 train_time:180057ms step_avg:141.55ms step:1283/4578 train_loss:3.7118 train_time:180198ms step_avg:141.55ms step:1284/4578 train_loss:3.5963 train_time:180337ms step_avg:141.55ms step:1285/4578 train_loss:3.6226 train_time:180477ms step_avg:141.55ms step:1286/4578 train_loss:3.6076 train_time:180618ms step_avg:141.55ms step:1287/4578 train_loss:3.5912 train_time:180757ms step_avg:141.55ms step:1288/4578 train_loss:3.7166 train_time:180900ms step_avg:141.55ms step:1289/4578 train_loss:3.5574 train_time:181037ms step_avg:141.55ms step:1290/4578 train_loss:3.6451 train_time:181176ms step_avg:141.54ms step:1291/4578 train_loss:3.7110 train_time:181316ms step_avg:141.54ms step:1292/4578 train_loss:3.6356 train_time:181455ms step_avg:141.54ms step:1293/4578 train_loss:3.7372 train_time:181597ms step_avg:141.54ms step:1294/4578 train_loss:3.7522 train_time:181734ms step_avg:141.54ms step:1295/4578 train_loss:3.7403 train_time:181874ms step_avg:141.54ms step:1296/4578 train_loss:3.5385 train_time:182015ms step_avg:141.54ms step:1297/4578 train_loss:3.6088 train_time:182154ms step_avg:141.53ms step:1298/4578 train_loss:3.5117 train_time:182295ms step_avg:141.53ms step:1299/4578 train_loss:3.5775 train_time:182434ms step_avg:141.53ms step:1300/4578 train_loss:3.6503 train_time:182575ms step_avg:141.53ms step:1301/4578 train_loss:3.6539 train_time:182714ms step_avg:141.53ms step:1302/4578 train_loss:3.6597 train_time:182854ms step_avg:141.53ms step:1303/4578 train_loss:3.8124 train_time:182995ms step_avg:141.53ms step:1304/4578 train_loss:3.5893 train_time:183135ms step_avg:141.53ms step:1305/4578 train_loss:3.7999 train_time:183276ms step_avg:141.53ms step:1306/4578 train_loss:3.5164 train_time:183416ms step_avg:141.53ms step:1307/4578 train_loss:3.7116 train_time:183556ms step_avg:141.52ms step:1308/4578 train_loss:3.7088 train_time:183696ms step_avg:141.52ms step:1309/4578 train_loss:3.5715 train_time:183835ms step_avg:141.52ms step:1310/4578 train_loss:3.5456 train_time:183975ms step_avg:141.52ms step:1311/4578 train_loss:3.5845 train_time:184115ms step_avg:141.52ms step:1312/4578 train_loss:3.5431 train_time:184255ms step_avg:141.52ms step:1313/4578 train_loss:3.6587 train_time:184394ms step_avg:141.52ms step:1314/4578 train_loss:3.6080 train_time:184534ms step_avg:141.51ms step:1315/4578 train_loss:3.3184 train_time:184676ms step_avg:141.51ms step:1316/4578 train_loss:3.5528 train_time:184816ms step_avg:141.51ms step:1317/4578 train_loss:3.6306 train_time:184955ms step_avg:141.51ms step:1318/4578 train_loss:3.6603 train_time:185096ms step_avg:141.51ms step:1319/4578 train_loss:3.5351 train_time:185237ms step_avg:141.51ms step:1320/4578 train_loss:3.6750 train_time:185375ms step_avg:141.51ms step:1321/4578 train_loss:3.7330 train_time:185514ms step_avg:141.51ms step:1322/4578 train_loss:3.6114 train_time:185654ms step_avg:141.50ms step:1323/4578 train_loss:3.5611 train_time:185930ms step_avg:141.61ms step:1324/4578 train_loss:3.5965 train_time:186066ms step_avg:141.60ms step:1325/4578 train_loss:3.6862 train_time:186207ms step_avg:141.60ms step:1326/4578 train_loss:3.7394 train_time:186345ms step_avg:141.60ms step:1327/4578 train_loss:3.5014 train_time:186485ms step_avg:141.60ms step:1328/4578 train_loss:3.4275 train_time:186624ms step_avg:141.60ms step:1329/4578 train_loss:3.7268 train_time:186764ms step_avg:141.60ms step:1330/4578 train_loss:3.5675 train_time:187064ms step_avg:141.72ms step:1331/4578 train_loss:3.7025 train_time:187200ms step_avg:141.71ms step:1332/4578 train_loss:3.6084 train_time:187337ms step_avg:141.71ms step:1333/4578 train_loss:4.0052 train_time:187476ms step_avg:141.71ms step:1334/4578 train_loss:3.7017 train_time:187614ms step_avg:141.70ms step:1335/4578 train_loss:3.6193 train_time:187751ms step_avg:141.70ms step:1336/4578 train_loss:3.5641 train_time:187891ms step_avg:141.70ms step:1337/4578 train_loss:3.5562 train_time:188038ms step_avg:141.70ms step:1338/4578 train_loss:3.8116 train_time:188178ms step_avg:141.70ms step:1339/4578 train_loss:3.7483 train_time:188319ms step_avg:141.70ms step:1340/4578 train_loss:3.5912 train_time:188458ms step_avg:141.70ms step:1341/4578 train_loss:3.5497 train_time:188597ms step_avg:141.70ms step:1342/4578 train_loss:3.8533 train_time:188735ms step_avg:141.69ms step:1343/4578 train_loss:3.6216 train_time:188876ms step_avg:141.69ms step:1344/4578 train_loss:3.6216 train_time:189019ms step_avg:141.69ms step:1345/4578 train_loss:3.6781 train_time:189157ms step_avg:141.69ms step:1346/4578 train_loss:3.6409 train_time:189298ms step_avg:141.69ms step:1347/4578 train_loss:3.5480 train_time:189438ms step_avg:141.69ms step:1348/4578 train_loss:3.4967 train_time:189576ms step_avg:141.69ms step:1349/4578 train_loss:3.5918 train_time:189718ms step_avg:141.69ms step:1350/4578 train_loss:3.5244 train_time:189854ms step_avg:141.68ms step:1351/4578 train_loss:3.6548 train_time:189996ms step_avg:141.68ms step:1352/4578 train_loss:3.5036 train_time:190137ms step_avg:141.68ms step:1353/4578 train_loss:3.5719 train_time:190284ms step_avg:141.69ms step:1354/4578 train_loss:3.6761 train_time:190416ms step_avg:141.68ms step:1355/4578 train_loss:3.5164 train_time:190554ms step_avg:141.68ms step:1356/4578 train_loss:3.4407 train_time:190694ms step_avg:141.67ms step:1357/4578 train_loss:3.7798 train_time:190843ms step_avg:141.68ms step:1358/4578 train_loss:3.7140 train_time:190976ms step_avg:141.67ms step:1359/4578 train_loss:3.4298 train_time:191116ms step_avg:141.67ms step:1360/4578 train_loss:3.7123 train_time:191256ms step_avg:141.67ms step:1361/4578 train_loss:3.6007 train_time:191395ms step_avg:141.67ms step:1362/4578 train_loss:3.4597 train_time:191534ms step_avg:141.67ms step:1363/4578 train_loss:3.6345 train_time:191674ms step_avg:141.67ms step:1364/4578 train_loss:3.5304 train_time:191815ms step_avg:141.67ms step:1365/4578 train_loss:3.5544 train_time:191954ms step_avg:141.66ms step:1366/4578 train_loss:3.5805 train_time:192095ms step_avg:141.66ms step:1367/4578 train_loss:3.6788 train_time:192236ms step_avg:141.66ms step:1368/4578 train_loss:3.6677 train_time:192376ms step_avg:141.66ms step:1369/4578 train_loss:3.6199 train_time:192516ms step_avg:141.66ms step:1370/4578 train_loss:3.5243 train_time:192656ms step_avg:141.66ms step:1371/4578 train_loss:3.8459 train_time:192795ms step_avg:141.66ms step:1372/4578 train_loss:3.5908 train_time:192934ms step_avg:141.65ms step:1373/4578 train_loss:3.6279 train_time:193074ms step_avg:141.65ms step:1374/4578 train_loss:3.6239 train_time:193219ms step_avg:141.66ms step:1375/4578 train_loss:3.4264 train_time:193354ms step_avg:141.65ms step:1375/4578 val_loss:3.5820 train_time:193410ms step_avg:141.69ms step:1376/4578 train_loss:3.8189 train_time:193509ms step_avg:141.66ms step:1377/4578 train_loss:3.5997 train_time:193655ms step_avg:141.66ms step:1378/4578 train_loss:3.7464 train_time:193794ms step_avg:141.66ms step:1379/4578 train_loss:3.7970 train_time:193933ms step_avg:141.66ms step:1380/4578 train_loss:3.4657 train_time:194070ms step_avg:141.66ms step:1381/4578 train_loss:3.5850 train_time:194210ms step_avg:141.66ms step:1382/4578 train_loss:4.0516 train_time:194347ms step_avg:141.65ms step:1383/4578 train_loss:3.5029 train_time:194487ms step_avg:141.65ms step:1384/4578 train_loss:3.6601 train_time:194632ms step_avg:141.65ms step:1385/4578 train_loss:3.7389 train_time:194773ms step_avg:141.65ms step:1386/4578 train_loss:3.6425 train_time:194912ms step_avg:141.65ms step:1387/4578 train_loss:3.6429 train_time:195051ms step_avg:141.65ms step:1388/4578 train_loss:3.4649 train_time:195191ms step_avg:141.65ms step:1389/4578 train_loss:3.6121 train_time:195330ms step_avg:141.65ms step:1390/4578 train_loss:3.5822 train_time:195470ms step_avg:141.64ms step:1391/4578 train_loss:3.8430 train_time:195611ms step_avg:141.64ms step:1392/4578 train_loss:3.5550 train_time:195751ms step_avg:141.64ms step:1393/4578 train_loss:3.5521 train_time:195891ms step_avg:141.64ms step:1394/4578 train_loss:3.5114 train_time:196031ms step_avg:141.64ms step:1395/4578 train_loss:3.7984 train_time:196170ms step_avg:141.64ms step:1396/4578 train_loss:3.6935 train_time:196309ms step_avg:141.64ms step:1397/4578 train_loss:3.6961 train_time:196449ms step_avg:141.64ms step:1398/4578 train_loss:3.5661 train_time:196590ms step_avg:141.64ms step:1399/4578 train_loss:3.5415 train_time:196730ms step_avg:141.63ms step:1400/4578 train_loss:3.6013 train_time:196871ms step_avg:141.63ms step:1401/4578 train_loss:3.5765 train_time:197011ms step_avg:141.63ms step:1402/4578 train_loss:3.5972 train_time:197150ms step_avg:141.63ms step:1403/4578 train_loss:3.5628 train_time:197290ms step_avg:141.63ms step:1404/4578 train_loss:3.7941 train_time:197429ms step_avg:141.63ms step:1405/4578 train_loss:3.5400 train_time:197569ms step_avg:141.63ms step:1406/4578 train_loss:3.5862 train_time:197711ms step_avg:141.63ms step:1407/4578 train_loss:3.5863 train_time:197850ms step_avg:141.63ms step:1408/4578 train_loss:3.4505 train_time:197990ms step_avg:141.62ms step:1409/4578 train_loss:3.5682 train_time:198131ms step_avg:141.62ms step:1410/4578 train_loss:3.5491 train_time:198270ms step_avg:141.62ms step:1411/4578 train_loss:3.5474 train_time:198409ms step_avg:141.62ms step:1412/4578 train_loss:3.6351 train_time:198549ms step_avg:141.62ms step:1413/4578 train_loss:3.5717 train_time:198689ms step_avg:141.62ms step:1414/4578 train_loss:3.6231 train_time:198829ms step_avg:141.62ms step:1415/4578 train_loss:3.6099 train_time:198972ms step_avg:141.62ms step:1416/4578 train_loss:3.6880 train_time:199110ms step_avg:141.61ms step:1417/4578 train_loss:3.4910 train_time:199251ms step_avg:141.61ms step:1418/4578 train_loss:3.5614 train_time:199388ms step_avg:141.61ms step:1419/4578 train_loss:3.6469 train_time:199528ms step_avg:141.61ms step:1420/4578 train_loss:3.6785 train_time:199668ms step_avg:141.61ms step:1421/4578 train_loss:3.6510 train_time:199808ms step_avg:141.61ms step:1422/4578 train_loss:3.6373 train_time:199948ms step_avg:141.61ms step:1423/4578 train_loss:3.6287 train_time:200089ms step_avg:141.61ms step:1424/4578 train_loss:3.6109 train_time:200230ms step_avg:141.61ms step:1425/4578 train_loss:3.6052 train_time:200369ms step_avg:141.60ms step:1426/4578 train_loss:3.4808 train_time:200510ms step_avg:141.60ms step:1427/4578 train_loss:3.5981 train_time:200652ms step_avg:141.60ms step:1428/4578 train_loss:3.5345 train_time:200790ms step_avg:141.60ms step:1429/4578 train_loss:3.6493 train_time:200931ms step_avg:141.60ms step:1430/4578 train_loss:3.6106 train_time:201070ms step_avg:141.60ms step:1431/4578 train_loss:3.5450 train_time:201211ms step_avg:141.60ms step:1432/4578 train_loss:3.5854 train_time:201350ms step_avg:141.60ms step:1433/4578 train_loss:3.6266 train_time:201493ms step_avg:141.60ms step:1434/4578 train_loss:3.5047 train_time:201631ms step_avg:141.59ms step:1435/4578 train_loss:3.6042 train_time:201770ms step_avg:141.59ms step:1436/4578 train_loss:3.4190 train_time:201917ms step_avg:141.60ms step:1437/4578 train_loss:3.4876 train_time:202050ms step_avg:141.59ms step:1438/4578 train_loss:3.6815 train_time:202190ms step_avg:141.59ms step:1439/4578 train_loss:3.6378 train_time:202334ms step_avg:141.59ms step:1440/4578 train_loss:3.5907 train_time:202470ms step_avg:141.59ms step:1441/4578 train_loss:3.4435 train_time:202610ms step_avg:141.59ms step:1442/4578 train_loss:3.6107 train_time:202749ms step_avg:141.58ms step:1443/4578 train_loss:3.6807 train_time:202889ms step_avg:141.58ms step:1444/4578 train_loss:3.7522 train_time:203029ms step_avg:141.58ms step:1445/4578 train_loss:3.7132 train_time:203169ms step_avg:141.58ms step:1446/4578 train_loss:3.6060 train_time:203309ms step_avg:141.58ms step:1447/4578 train_loss:3.4745 train_time:203448ms step_avg:141.58ms step:1448/4578 train_loss:3.5494 train_time:203588ms step_avg:141.58ms step:1449/4578 train_loss:3.5715 train_time:203728ms step_avg:141.58ms step:1450/4578 train_loss:3.6913 train_time:203868ms step_avg:141.57ms step:1451/4578 train_loss:3.6746 train_time:204008ms step_avg:141.57ms step:1452/4578 train_loss:3.4948 train_time:204148ms step_avg:141.57ms step:1453/4578 train_loss:3.6096 train_time:204289ms step_avg:141.57ms step:1454/4578 train_loss:3.5218 train_time:204428ms step_avg:141.57ms step:1455/4578 train_loss:3.5557 train_time:204569ms step_avg:141.57ms step:1456/4578 train_loss:3.6004 train_time:204708ms step_avg:141.57ms step:1457/4578 train_loss:3.5321 train_time:204847ms step_avg:141.57ms step:1458/4578 train_loss:3.4300 train_time:204988ms step_avg:141.57ms step:1459/4578 train_loss:3.6726 train_time:205128ms step_avg:141.56ms step:1460/4578 train_loss:3.5463 train_time:205267ms step_avg:141.56ms step:1461/4578 train_loss:3.5938 train_time:205408ms step_avg:141.56ms step:1462/4578 train_loss:3.7220 train_time:205547ms step_avg:141.56ms step:1463/4578 train_loss:3.5435 train_time:205688ms step_avg:141.56ms step:1464/4578 train_loss:3.7304 train_time:205838ms step_avg:141.57ms step:1465/4578 train_loss:3.6220 train_time:205968ms step_avg:141.56ms step:1466/4578 train_loss:3.6338 train_time:206107ms step_avg:141.56ms step:1467/4578 train_loss:3.5474 train_time:206256ms step_avg:141.56ms step:1468/4578 train_loss:3.7086 train_time:206388ms step_avg:141.56ms step:1469/4578 train_loss:3.5647 train_time:206528ms step_avg:141.55ms step:1470/4578 train_loss:3.5339 train_time:206671ms step_avg:141.56ms step:1471/4578 train_loss:3.5893 train_time:206807ms step_avg:141.55ms step:1472/4578 train_loss:3.5166 train_time:206947ms step_avg:141.55ms step:1473/4578 train_loss:3.6075 train_time:207091ms step_avg:141.55ms step:1474/4578 train_loss:3.7040 train_time:207227ms step_avg:141.55ms step:1475/4578 train_loss:3.5784 train_time:207367ms step_avg:141.55ms step:1476/4578 train_loss:3.4101 train_time:207517ms step_avg:141.55ms step:1477/4578 train_loss:3.5244 train_time:207646ms step_avg:141.54ms step:1478/4578 train_loss:3.5057 train_time:207787ms step_avg:141.54ms step:1479/4578 train_loss:3.5834 train_time:207927ms step_avg:141.54ms step:1480/4578 train_loss:3.6709 train_time:208066ms step_avg:141.54ms step:1481/4578 train_loss:3.5389 train_time:208207ms step_avg:141.54ms step:1482/4578 train_loss:3.7151 train_time:208347ms step_avg:141.54ms step:1483/4578 train_loss:3.6385 train_time:208487ms step_avg:141.54ms step:1484/4578 train_loss:3.5467 train_time:208627ms step_avg:141.54ms step:1485/4578 train_loss:3.5352 train_time:208767ms step_avg:141.54ms step:1486/4578 train_loss:3.5314 train_time:208910ms step_avg:141.54ms step:1487/4578 train_loss:3.5126 train_time:209047ms step_avg:141.53ms step:1488/4578 train_loss:3.5973 train_time:209187ms step_avg:141.53ms step:1489/4578 train_loss:3.5057 train_time:209340ms step_avg:141.54ms step:1490/4578 train_loss:3.5904 train_time:209466ms step_avg:141.53ms step:1491/4578 train_loss:3.5331 train_time:209609ms step_avg:141.53ms step:1492/4578 train_loss:3.4600 train_time:209747ms step_avg:141.53ms step:1493/4578 train_loss:3.5308 train_time:209888ms step_avg:141.53ms step:1494/4578 train_loss:3.7029 train_time:210027ms step_avg:141.53ms step:1495/4578 train_loss:3.5585 train_time:210169ms step_avg:141.53ms step:1496/4578 train_loss:3.3108 train_time:210307ms step_avg:141.53ms step:1497/4578 train_loss:3.6156 train_time:210446ms step_avg:141.52ms step:1498/4578 train_loss:3.5835 train_time:210589ms step_avg:141.52ms step:1499/4578 train_loss:3.6267 train_time:210727ms step_avg:141.52ms step:1500/4578 train_loss:3.5828 train_time:210866ms step_avg:141.52ms step:1500/4578 val_loss:3.5598 train_time:210922ms step_avg:141.56ms step:1501/4578 train_loss:3.5624 train_time:211020ms step_avg:141.53ms step:1502/4578 train_loss:3.3626 train_time:211163ms step_avg:141.53ms step:1503/4578 train_loss:3.6295 train_time:211303ms step_avg:141.53ms step:1504/4578 train_loss:3.5101 train_time:211443ms step_avg:141.53ms step:1505/4578 train_loss:3.5185 train_time:211587ms step_avg:141.53ms step:1506/4578 train_loss:3.4738 train_time:211721ms step_avg:141.52ms step:1507/4578 train_loss:3.5650 train_time:211859ms step_avg:141.52ms step:1508/4578 train_loss:3.4805 train_time:212001ms step_avg:141.52ms step:1509/4578 train_loss:3.7943 train_time:212142ms step_avg:141.52ms step:1510/4578 train_loss:3.5316 train_time:212283ms step_avg:141.52ms step:1511/4578 train_loss:3.5336 train_time:212437ms step_avg:141.53ms step:1512/4578 train_loss:3.6640 train_time:212718ms step_avg:141.62ms step:1513/4578 train_loss:3.6955 train_time:212856ms step_avg:141.62ms step:1514/4578 train_loss:3.5564 train_time:212993ms step_avg:141.62ms step:1515/4578 train_loss:3.3881 train_time:213132ms step_avg:141.62ms step:1516/4578 train_loss:3.5136 train_time:213271ms step_avg:141.61ms step:1517/4578 train_loss:3.5151 train_time:213409ms step_avg:141.61ms step:1518/4578 train_loss:3.5995 train_time:213548ms step_avg:141.61ms step:1519/4578 train_loss:3.4838 train_time:213695ms step_avg:141.61ms step:1520/4578 train_loss:3.7760 train_time:213996ms step_avg:141.72ms step:1521/4578 train_loss:3.4396 train_time:214136ms step_avg:141.72ms step:1522/4578 train_loss:3.4985 train_time:214274ms step_avg:141.72ms step:1523/4578 train_loss:3.6478 train_time:214412ms step_avg:141.71ms step:1524/4578 train_loss:3.5030 train_time:214551ms step_avg:141.71ms step:1525/4578 train_loss:3.5954 train_time:214689ms step_avg:141.71ms step:1526/4578 train_loss:3.5950 train_time:214828ms step_avg:141.71ms step:1527/4578 train_loss:3.5597 train_time:214974ms step_avg:141.71ms step:1528/4578 train_loss:3.5571 train_time:215116ms step_avg:141.71ms step:1529/4578 train_loss:3.7031 train_time:215255ms step_avg:141.71ms step:1530/4578 train_loss:3.6738 train_time:215394ms step_avg:141.71ms step:1531/4578 train_loss:3.5097 train_time:215533ms step_avg:141.70ms step:1532/4578 train_loss:3.4749 train_time:215672ms step_avg:141.70ms step:1533/4578 train_loss:3.6234 train_time:215813ms step_avg:141.70ms step:1534/4578 train_loss:3.5758 train_time:215956ms step_avg:141.70ms step:1535/4578 train_loss:3.5672 train_time:216098ms step_avg:141.70ms step:1536/4578 train_loss:3.5583 train_time:216238ms step_avg:141.70ms step:1537/4578 train_loss:3.4986 train_time:216377ms step_avg:141.70ms step:1538/4578 train_loss:3.5591 train_time:216517ms step_avg:141.70ms step:1539/4578 train_loss:3.7283 train_time:216656ms step_avg:141.70ms step:1540/4578 train_loss:3.6636 train_time:216796ms step_avg:141.70ms step:1541/4578 train_loss:3.5705 train_time:216937ms step_avg:141.70ms step:1542/4578 train_loss:3.5250 train_time:217079ms step_avg:141.70ms step:1543/4578 train_loss:3.5188 train_time:217219ms step_avg:141.70ms step:1544/4578 train_loss:3.4902 train_time:217359ms step_avg:141.69ms step:1545/4578 train_loss:3.5710 train_time:217506ms step_avg:141.70ms step:1546/4578 train_loss:3.5426 train_time:217637ms step_avg:141.69ms step:1547/4578 train_loss:3.5236 train_time:217777ms step_avg:141.69ms step:1548/4578 train_loss:3.4825 train_time:217918ms step_avg:141.69ms step:1549/4578 train_loss:3.5118 train_time:218058ms step_avg:141.69ms step:1550/4578 train_loss:3.6367 train_time:218199ms step_avg:141.69ms step:1551/4578 train_loss:3.5574 train_time:218338ms step_avg:141.69ms step:1552/4578 train_loss:3.4994 train_time:218477ms step_avg:141.68ms step:1553/4578 train_loss:3.4963 train_time:218618ms step_avg:141.68ms step:1554/4578 train_loss:3.4783 train_time:218756ms step_avg:141.68ms step:1555/4578 train_loss:3.6106 train_time:218897ms step_avg:141.68ms step:1556/4578 train_loss:3.6166 train_time:219037ms step_avg:141.68ms step:1557/4578 train_loss:3.5527 train_time:219177ms step_avg:141.68ms step:1558/4578 train_loss:3.6047 train_time:219319ms step_avg:141.68ms step:1559/4578 train_loss:3.5297 train_time:219455ms step_avg:141.68ms step:1560/4578 train_loss:3.4406 train_time:219618ms step_avg:141.69ms step:1561/4578 train_loss:3.6855 train_time:219743ms step_avg:141.68ms step:1562/4578 train_loss:3.5015 train_time:219883ms step_avg:141.68ms step:1563/4578 train_loss:3.4851 train_time:220029ms step_avg:141.68ms step:1564/4578 train_loss:3.6119 train_time:220157ms step_avg:141.67ms step:1565/4578 train_loss:3.4438 train_time:220297ms step_avg:141.67ms step:1566/4578 train_loss:3.4949 train_time:220437ms step_avg:141.67ms step:1567/4578 train_loss:3.6444 train_time:220577ms step_avg:141.67ms step:1568/4578 train_loss:3.5248 train_time:220717ms step_avg:141.67ms step:1569/4578 train_loss:3.5058 train_time:220857ms step_avg:141.67ms step:1570/4578 train_loss:3.6039 train_time:220997ms step_avg:141.66ms step:1571/4578 train_loss:3.6125 train_time:221137ms step_avg:141.66ms step:1572/4578 train_loss:3.4409 train_time:221278ms step_avg:141.66ms step:1573/4578 train_loss:3.4701 train_time:221418ms step_avg:141.66ms step:1574/4578 train_loss:3.5912 train_time:221560ms step_avg:141.66ms step:1575/4578 train_loss:3.4531 train_time:221700ms step_avg:141.66ms step:1576/4578 train_loss:3.6056 train_time:221841ms step_avg:141.66ms step:1577/4578 train_loss:3.5048 train_time:221977ms step_avg:141.66ms step:1578/4578 train_loss:3.5635 train_time:222117ms step_avg:141.66ms step:1579/4578 train_loss:3.5375 train_time:222256ms step_avg:141.65ms step:1580/4578 train_loss:3.5020 train_time:222397ms step_avg:141.65ms step:1581/4578 train_loss:3.4777 train_time:222538ms step_avg:141.65ms step:1582/4578 train_loss:3.7253 train_time:222677ms step_avg:141.65ms step:1583/4578 train_loss:3.4909 train_time:222818ms step_avg:141.65ms step:1584/4578 train_loss:3.6431 train_time:222957ms step_avg:141.65ms step:1585/4578 train_loss:3.4824 train_time:223096ms step_avg:141.65ms step:1586/4578 train_loss:3.6395 train_time:223236ms step_avg:141.65ms step:1587/4578 train_loss:3.4205 train_time:223376ms step_avg:141.65ms step:1588/4578 train_loss:3.6207 train_time:223518ms step_avg:141.65ms step:1589/4578 train_loss:3.5375 train_time:223656ms step_avg:141.64ms step:1590/4578 train_loss:3.6927 train_time:223797ms step_avg:141.64ms step:1591/4578 train_loss:3.5016 train_time:223937ms step_avg:141.64ms step:1592/4578 train_loss:3.5213 train_time:224077ms step_avg:141.64ms step:1593/4578 train_loss:3.5875 train_time:224222ms step_avg:141.64ms step:1594/4578 train_loss:3.5648 train_time:224356ms step_avg:141.64ms step:1595/4578 train_loss:3.5399 train_time:224496ms step_avg:141.64ms step:1596/4578 train_loss:3.6793 train_time:224643ms step_avg:141.64ms step:1597/4578 train_loss:3.4142 train_time:224775ms step_avg:141.64ms step:1598/4578 train_loss:3.5786 train_time:224916ms step_avg:141.63ms step:1599/4578 train_loss:3.6147 train_time:225055ms step_avg:141.63ms step:1600/4578 train_loss:3.6675 train_time:225196ms step_avg:141.63ms step:1601/4578 train_loss:3.5128 train_time:225342ms step_avg:141.64ms step:1602/4578 train_loss:3.8113 train_time:225475ms step_avg:141.63ms step:1603/4578 train_loss:3.6962 train_time:225616ms step_avg:141.63ms step:1604/4578 train_loss:3.4732 train_time:225757ms step_avg:141.63ms step:1605/4578 train_loss:3.5129 train_time:225895ms step_avg:141.63ms step:1606/4578 train_loss:3.4026 train_time:226035ms step_avg:141.63ms step:1607/4578 train_loss:3.7199 train_time:226174ms step_avg:141.62ms step:1608/4578 train_loss:3.5219 train_time:226315ms step_avg:141.62ms step:1609/4578 train_loss:3.5489 train_time:226454ms step_avg:141.62ms step:1610/4578 train_loss:3.4993 train_time:226600ms step_avg:141.63ms step:1611/4578 train_loss:4.1038 train_time:226735ms step_avg:141.62ms step:1612/4578 train_loss:3.7344 train_time:226878ms step_avg:141.62ms step:1613/4578 train_loss:3.6457 train_time:227015ms step_avg:141.62ms step:1614/4578 train_loss:3.5139 train_time:227155ms step_avg:141.62ms step:1615/4578 train_loss:3.5579 train_time:227296ms step_avg:141.62ms step:1616/4578 train_loss:3.5425 train_time:227436ms step_avg:141.62ms step:1617/4578 train_loss:3.5050 train_time:227589ms step_avg:141.62ms step:1618/4578 train_loss:3.5856 train_time:227715ms step_avg:141.61ms step:1619/4578 train_loss:3.5376 train_time:227855ms step_avg:141.61ms step:1620/4578 train_loss:3.4330 train_time:228003ms step_avg:141.62ms step:1621/4578 train_loss:3.6991 train_time:228137ms step_avg:141.61ms step:1622/4578 train_loss:3.6053 train_time:228276ms step_avg:141.61ms step:1623/4578 train_loss:3.3902 train_time:228415ms step_avg:141.61ms step:1624/4578 train_loss:3.5188 train_time:228554ms step_avg:141.61ms step:1625/4578 train_loss:3.4777 train_time:228694ms step_avg:141.61ms step:1625/4578 val_loss:3.5452 train_time:228750ms step_avg:141.64ms step:1626/4578 train_loss:3.5535 train_time:228840ms step_avg:141.61ms step:1627/4578 train_loss:3.5184 train_time:228991ms step_avg:141.61ms step:1628/4578 train_loss:3.4758 train_time:229131ms step_avg:141.61ms step:1629/4578 train_loss:3.5852 train_time:229271ms step_avg:141.61ms step:1630/4578 train_loss:3.4838 train_time:229409ms step_avg:141.61ms step:1631/4578 train_loss:3.5437 train_time:229548ms step_avg:141.61ms step:1632/4578 train_loss:3.4205 train_time:229688ms step_avg:141.61ms step:1633/4578 train_loss:3.3965 train_time:229832ms step_avg:141.61ms step:1634/4578 train_loss:3.5540 train_time:229975ms step_avg:141.61ms step:1635/4578 train_loss:3.5385 train_time:230115ms step_avg:141.61ms step:1636/4578 train_loss:3.4792 train_time:230254ms step_avg:141.61ms step:1637/4578 train_loss:3.5697 train_time:230398ms step_avg:141.61ms step:1638/4578 train_loss:3.6175 train_time:230534ms step_avg:141.61ms step:1639/4578 train_loss:3.6482 train_time:230673ms step_avg:141.60ms step:1640/4578 train_loss:3.8089 train_time:230816ms step_avg:141.61ms step:1641/4578 train_loss:3.6310 train_time:230959ms step_avg:141.61ms step:1642/4578 train_loss:3.5467 train_time:231099ms step_avg:141.60ms step:1643/4578 train_loss:3.6319 train_time:231238ms step_avg:141.60ms step:1644/4578 train_loss:3.5337 train_time:231378ms step_avg:141.60ms step:1645/4578 train_loss:3.5466 train_time:231518ms step_avg:141.60ms step:1646/4578 train_loss:3.5454 train_time:231658ms step_avg:141.60ms step:1647/4578 train_loss:3.3241 train_time:231800ms step_avg:141.60ms step:1648/4578 train_loss:3.5777 train_time:231939ms step_avg:141.60ms step:1649/4578 train_loss:3.4495 train_time:232079ms step_avg:141.60ms step:1650/4578 train_loss:3.5222 train_time:232219ms step_avg:141.60ms step:1651/4578 train_loss:3.5062 train_time:232358ms step_avg:141.60ms step:1652/4578 train_loss:3.5694 train_time:232499ms step_avg:141.59ms step:1653/4578 train_loss:3.5022 train_time:232638ms step_avg:141.59ms step:1654/4578 train_loss:3.6260 train_time:232780ms step_avg:141.59ms step:1655/4578 train_loss:3.6194 train_time:232919ms step_avg:141.59ms step:1656/4578 train_loss:3.4383 train_time:233070ms step_avg:141.60ms step:1657/4578 train_loss:3.6009 train_time:233199ms step_avg:141.59ms step:1658/4578 train_loss:3.4920 train_time:233338ms step_avg:141.59ms step:1659/4578 train_loss:3.4735 train_time:233479ms step_avg:141.59ms step:1660/4578 train_loss:3.5591 train_time:233622ms step_avg:141.59ms step:1661/4578 train_loss:3.5759 train_time:233759ms step_avg:141.59ms step:1662/4578 train_loss:3.4920 train_time:233899ms step_avg:141.59ms step:1663/4578 train_loss:3.5913 train_time:234039ms step_avg:141.58ms step:1664/4578 train_loss:3.6005 train_time:234179ms step_avg:141.58ms step:1665/4578 train_loss:3.6224 train_time:234318ms step_avg:141.58ms step:1666/4578 train_loss:3.5977 train_time:234462ms step_avg:141.58ms step:1667/4578 train_loss:3.7342 train_time:234599ms step_avg:141.58ms step:1668/4578 train_loss:3.4476 train_time:234738ms step_avg:141.58ms step:1669/4578 train_loss:3.5336 train_time:234879ms step_avg:141.58ms step:1670/4578 train_loss:3.4548 train_time:235019ms step_avg:141.58ms step:1671/4578 train_loss:3.4565 train_time:235158ms step_avg:141.58ms step:1672/4578 train_loss:3.6172 train_time:235299ms step_avg:141.58ms step:1673/4578 train_loss:3.7979 train_time:235440ms step_avg:141.58ms step:1674/4578 train_loss:3.5161 train_time:235579ms step_avg:141.57ms step:1675/4578 train_loss:3.5041 train_time:235718ms step_avg:141.57ms step:1676/4578 train_loss:3.3902 train_time:235858ms step_avg:141.57ms step:1677/4578 train_loss:3.5982 train_time:235999ms step_avg:141.57ms step:1678/4578 train_loss:3.5044 train_time:236138ms step_avg:141.57ms step:1679/4578 train_loss:3.5345 train_time:236278ms step_avg:141.57ms step:1680/4578 train_loss:3.5188 train_time:236419ms step_avg:141.57ms step:1681/4578 train_loss:3.3607 train_time:236559ms step_avg:141.57ms step:1682/4578 train_loss:3.5236 train_time:236698ms step_avg:141.57ms step:1683/4578 train_loss:3.5427 train_time:236838ms step_avg:141.56ms step:1684/4578 train_loss:3.5917 train_time:236980ms step_avg:141.57ms step:1685/4578 train_loss:3.5829 train_time:237119ms step_avg:141.56ms step:1686/4578 train_loss:3.4943 train_time:237259ms step_avg:141.56ms step:1687/4578 train_loss:3.5927 train_time:237400ms step_avg:141.56ms step:1688/4578 train_loss:3.4787 train_time:237539ms step_avg:141.56ms step:1689/4578 train_loss:3.5625 train_time:237679ms step_avg:141.56ms step:1690/4578 train_loss:3.4761 train_time:237820ms step_avg:141.56ms step:1691/4578 train_loss:3.3822 train_time:237960ms step_avg:141.56ms step:1692/4578 train_loss:3.5279 train_time:238100ms step_avg:141.56ms step:1693/4578 train_loss:3.5198 train_time:238239ms step_avg:141.56ms step:1694/4578 train_loss:3.4396 train_time:238379ms step_avg:141.56ms step:1695/4578 train_loss:3.8858 train_time:238519ms step_avg:141.55ms step:1696/4578 train_loss:3.6010 train_time:238661ms step_avg:141.55ms step:1697/4578 train_loss:3.5789 train_time:238799ms step_avg:141.55ms step:1698/4578 train_loss:3.4828 train_time:238939ms step_avg:141.55ms step:1699/4578 train_loss:3.4017 train_time:239079ms step_avg:141.55ms step:1700/4578 train_loss:3.4896 train_time:239220ms step_avg:141.55ms step:1701/4578 train_loss:3.4836 train_time:239504ms step_avg:141.63ms step:1702/4578 train_loss:3.5549 train_time:239641ms step_avg:141.63ms step:1703/4578 train_loss:3.4869 train_time:239779ms step_avg:141.63ms step:1704/4578 train_loss:3.6870 train_time:239917ms step_avg:141.63ms step:1705/4578 train_loss:3.4510 train_time:240054ms step_avg:141.62ms step:1706/4578 train_loss:3.6809 train_time:240194ms step_avg:141.62ms step:1707/4578 train_loss:3.5190 train_time:240332ms step_avg:141.62ms step:1708/4578 train_loss:3.3001 train_time:240480ms step_avg:141.63ms step:1709/4578 train_loss:3.6323 train_time:240623ms step_avg:141.63ms step:1710/4578 train_loss:3.5444 train_time:240930ms step_avg:141.72ms step:1711/4578 train_loss:3.5312 train_time:241065ms step_avg:141.72ms step:1712/4578 train_loss:3.5247 train_time:241206ms step_avg:141.72ms step:1713/4578 train_loss:3.5609 train_time:241344ms step_avg:141.72ms step:1714/4578 train_loss:3.5839 train_time:241482ms step_avg:141.71ms step:1715/4578 train_loss:3.5092 train_time:241621ms step_avg:141.71ms step:1716/4578 train_loss:3.5133 train_time:241758ms step_avg:141.71ms step:1717/4578 train_loss:3.3514 train_time:241905ms step_avg:141.71ms step:1718/4578 train_loss:3.4887 train_time:242045ms step_avg:141.71ms step:1719/4578 train_loss:3.5080 train_time:242185ms step_avg:141.71ms step:1720/4578 train_loss:3.4571 train_time:242326ms step_avg:141.71ms step:1721/4578 train_loss:3.6114 train_time:242468ms step_avg:141.71ms step:1722/4578 train_loss:3.4185 train_time:242605ms step_avg:141.71ms step:1723/4578 train_loss:3.5606 train_time:242745ms step_avg:141.71ms step:1724/4578 train_loss:3.6415 train_time:242885ms step_avg:141.71ms step:1725/4578 train_loss:3.4942 train_time:243027ms step_avg:141.71ms step:1726/4578 train_loss:3.7176 train_time:243168ms step_avg:141.71ms step:1727/4578 train_loss:3.5145 train_time:243314ms step_avg:141.71ms step:1728/4578 train_loss:3.5701 train_time:243473ms step_avg:141.72ms step:1729/4578 train_loss:3.5415 train_time:243603ms step_avg:141.71ms step:1730/4578 train_loss:3.5484 train_time:243743ms step_avg:141.71ms step:1731/4578 train_loss:3.9158 train_time:243873ms step_avg:141.70ms step:1732/4578 train_loss:3.5352 train_time:244016ms step_avg:141.71ms step:1733/4578 train_loss:3.6564 train_time:244158ms step_avg:141.71ms step:1734/4578 train_loss:3.4443 train_time:244296ms step_avg:141.70ms step:1735/4578 train_loss:3.4880 train_time:244438ms step_avg:141.70ms step:1736/4578 train_loss:3.5023 train_time:244587ms step_avg:141.71ms step:1737/4578 train_loss:3.4884 train_time:244717ms step_avg:141.70ms step:1738/4578 train_loss:3.6317 train_time:244856ms step_avg:141.70ms step:1739/4578 train_loss:3.5007 train_time:245013ms step_avg:141.71ms step:1740/4578 train_loss:3.5537 train_time:245139ms step_avg:141.70ms step:1741/4578 train_loss:3.6043 train_time:245279ms step_avg:141.70ms step:1742/4578 train_loss:3.4151 train_time:245421ms step_avg:141.70ms step:1743/4578 train_loss:3.3050 train_time:245560ms step_avg:141.70ms step:1744/4578 train_loss:3.2565 train_time:245702ms step_avg:141.70ms step:1745/4578 train_loss:3.5309 train_time:245842ms step_avg:141.70ms step:1746/4578 train_loss:3.5408 train_time:245983ms step_avg:141.70ms step:1747/4578 train_loss:3.5127 train_time:246124ms step_avg:141.69ms step:1748/4578 train_loss:3.5194 train_time:246263ms step_avg:141.69ms step:1749/4578 train_loss:3.7660 train_time:246407ms step_avg:141.69ms step:1750/4578 train_loss:3.4714 train_time:246543ms step_avg:141.69ms step:1750/4578 val_loss:3.5278 train_time:246598ms step_avg:141.72ms step:1751/4578 train_loss:3.5559 train_time:246690ms step_avg:141.69ms step:1752/4578 train_loss:3.5314 train_time:246837ms step_avg:141.70ms step:1753/4578 train_loss:3.1732 train_time:246977ms step_avg:141.70ms step:1754/4578 train_loss:3.2937 train_time:247116ms step_avg:141.70ms step:1755/4578 train_loss:3.3887 train_time:247258ms step_avg:141.70ms step:1756/4578 train_loss:3.3437 train_time:247394ms step_avg:141.69ms step:1757/4578 train_loss:3.5024 train_time:247533ms step_avg:141.69ms step:1758/4578 train_loss:3.3840 train_time:247677ms step_avg:141.69ms step:1759/4578 train_loss:3.3759 train_time:247818ms step_avg:141.69ms step:1760/4578 train_loss:4.4330 train_time:247962ms step_avg:141.69ms step:1761/4578 train_loss:3.5083 train_time:248100ms step_avg:141.69ms step:1762/4578 train_loss:3.5461 train_time:248240ms step_avg:141.69ms step:1763/4578 train_loss:3.5407 train_time:248389ms step_avg:141.69ms step:1764/4578 train_loss:3.5619 train_time:248521ms step_avg:141.69ms step:1765/4578 train_loss:3.4760 train_time:248662ms step_avg:141.69ms step:1766/4578 train_loss:3.5180 train_time:248804ms step_avg:141.69ms step:1767/4578 train_loss:3.5318 train_time:248944ms step_avg:141.69ms step:1768/4578 train_loss:3.7897 train_time:249086ms step_avg:141.69ms step:1769/4578 train_loss:3.5108 train_time:249227ms step_avg:141.69ms step:1770/4578 train_loss:3.5764 train_time:249364ms step_avg:141.68ms step:1771/4578 train_loss:3.9511 train_time:249504ms step_avg:141.68ms step:1772/4578 train_loss:3.5154 train_time:249645ms step_avg:141.68ms step:1773/4578 train_loss:3.4188 train_time:249786ms step_avg:141.68ms step:1774/4578 train_loss:3.6758 train_time:249939ms step_avg:141.69ms step:1775/4578 train_loss:3.4100 train_time:250068ms step_avg:141.68ms step:1776/4578 train_loss:3.5750 train_time:250207ms step_avg:141.68ms step:1777/4578 train_loss:3.6261 train_time:250348ms step_avg:141.68ms step:1778/4578 train_loss:3.7198 train_time:250491ms step_avg:141.68ms step:1779/4578 train_loss:3.5245 train_time:250633ms step_avg:141.68ms step:1780/4578 train_loss:3.8268 train_time:250767ms step_avg:141.68ms step:1781/4578 train_loss:3.5909 train_time:250908ms step_avg:141.68ms step:1782/4578 train_loss:3.6085 train_time:251048ms step_avg:141.68ms step:1783/4578 train_loss:3.3949 train_time:251188ms step_avg:141.67ms step:1784/4578 train_loss:3.4686 train_time:251327ms step_avg:141.67ms step:1785/4578 train_loss:3.6189 train_time:251468ms step_avg:141.67ms step:1786/4578 train_loss:3.5114 train_time:251608ms step_avg:141.67ms step:1787/4578 train_loss:3.6789 train_time:251749ms step_avg:141.67ms step:1788/4578 train_loss:3.4993 train_time:251889ms step_avg:141.67ms step:1789/4578 train_loss:3.4687 train_time:252029ms step_avg:141.67ms step:1790/4578 train_loss:3.6118 train_time:252169ms step_avg:141.67ms step:1791/4578 train_loss:3.5132 train_time:252308ms step_avg:141.67ms step:1792/4578 train_loss:3.4544 train_time:252448ms step_avg:141.67ms step:1793/4578 train_loss:3.5938 train_time:252592ms step_avg:141.67ms step:1794/4578 train_loss:3.4746 train_time:252728ms step_avg:141.66ms step:1795/4578 train_loss:3.4591 train_time:252868ms step_avg:141.66ms step:1796/4578 train_loss:3.5133 train_time:253015ms step_avg:141.67ms step:1797/4578 train_loss:3.4975 train_time:253155ms step_avg:141.66ms step:1798/4578 train_loss:3.6197 train_time:253290ms step_avg:141.66ms step:1799/4578 train_loss:3.5004 train_time:253432ms step_avg:141.66ms step:1800/4578 train_loss:3.5805 train_time:253570ms step_avg:141.66ms step:1801/4578 train_loss:3.5105 train_time:253709ms step_avg:141.66ms step:1802/4578 train_loss:3.5460 train_time:253850ms step_avg:141.66ms step:1803/4578 train_loss:3.4616 train_time:253993ms step_avg:141.66ms step:1804/4578 train_loss:3.3923 train_time:254133ms step_avg:141.66ms step:1805/4578 train_loss:3.6326 train_time:254269ms step_avg:141.65ms step:1806/4578 train_loss:3.5552 train_time:254409ms step_avg:141.65ms step:1807/4578 train_loss:3.5697 train_time:254555ms step_avg:141.66ms step:1808/4578 train_loss:3.6757 train_time:254690ms step_avg:141.65ms step:1809/4578 train_loss:3.4743 train_time:254833ms step_avg:141.65ms step:1810/4578 train_loss:3.5781 train_time:254973ms step_avg:141.65ms step:1811/4578 train_loss:3.7034 train_time:255122ms step_avg:141.66ms step:1812/4578 train_loss:3.5653 train_time:255247ms step_avg:141.65ms step:1813/4578 train_loss:3.6015 train_time:255387ms step_avg:141.65ms step:1814/4578 train_loss:3.6317 train_time:255533ms step_avg:141.65ms step:1815/4578 train_loss:3.5760 train_time:255668ms step_avg:141.64ms step:1816/4578 train_loss:3.6105 train_time:255808ms step_avg:141.64ms step:1817/4578 train_loss:3.5607 train_time:255957ms step_avg:141.65ms step:1818/4578 train_loss:3.6190 train_time:256087ms step_avg:141.64ms step:1819/4578 train_loss:3.5440 train_time:256227ms step_avg:141.64ms step:1820/4578 train_loss:3.5342 train_time:256368ms step_avg:141.64ms step:1821/4578 train_loss:3.4858 train_time:256507ms step_avg:141.64ms step:1822/4578 train_loss:3.4621 train_time:256648ms step_avg:141.64ms step:1823/4578 train_loss:3.3855 train_time:256788ms step_avg:141.64ms step:1824/4578 train_loss:3.5416 train_time:256927ms step_avg:141.64ms step:1825/4578 train_loss:3.6702 train_time:257071ms step_avg:141.64ms step:1826/4578 train_loss:3.6170 train_time:257208ms step_avg:141.63ms step:1827/4578 train_loss:3.6007 train_time:257349ms step_avg:141.63ms step:1828/4578 train_loss:3.4674 train_time:257492ms step_avg:141.63ms step:1829/4578 train_loss:3.4952 train_time:257637ms step_avg:141.64ms step:1830/4578 train_loss:3.6231 train_time:257769ms step_avg:141.63ms step:1831/4578 train_loss:3.4014 train_time:257909ms step_avg:141.63ms step:1832/4578 train_loss:3.5542 train_time:258050ms step_avg:141.63ms step:1833/4578 train_loss:3.4376 train_time:258193ms step_avg:141.63ms step:1834/4578 train_loss:3.7577 train_time:258331ms step_avg:141.63ms step:1835/4578 train_loss:3.5917 train_time:258472ms step_avg:141.63ms step:1836/4578 train_loss:3.5703 train_time:258611ms step_avg:141.63ms step:1837/4578 train_loss:3.6952 train_time:258751ms step_avg:141.63ms step:1838/4578 train_loss:3.5507 train_time:258891ms step_avg:141.63ms step:1839/4578 train_loss:3.4270 train_time:259031ms step_avg:141.62ms step:1840/4578 train_loss:3.5534 train_time:259172ms step_avg:141.62ms step:1841/4578 train_loss:3.4362 train_time:259313ms step_avg:141.62ms step:1842/4578 train_loss:3.5467 train_time:259452ms step_avg:141.62ms step:1843/4578 train_loss:3.6047 train_time:259592ms step_avg:141.62ms step:1844/4578 train_loss:3.3615 train_time:259731ms step_avg:141.62ms step:1845/4578 train_loss:3.4798 train_time:259871ms step_avg:141.62ms step:1846/4578 train_loss:3.5395 train_time:260011ms step_avg:141.62ms step:1847/4578 train_loss:3.4777 train_time:260155ms step_avg:141.62ms step:1848/4578 train_loss:3.3751 train_time:260300ms step_avg:141.62ms step:1849/4578 train_loss:3.6471 train_time:260435ms step_avg:141.62ms step:1850/4578 train_loss:3.4123 train_time:260573ms step_avg:141.62ms step:1851/4578 train_loss:3.4945 train_time:260710ms step_avg:141.61ms step:1852/4578 train_loss:3.4538 train_time:260849ms step_avg:141.61ms step:1853/4578 train_loss:3.6495 train_time:260990ms step_avg:141.61ms step:1854/4578 train_loss:3.6266 train_time:261130ms step_avg:141.61ms step:1855/4578 train_loss:3.5065 train_time:261270ms step_avg:141.61ms step:1856/4578 train_loss:3.4583 train_time:261409ms step_avg:141.61ms step:1857/4578 train_loss:3.4887 train_time:261564ms step_avg:141.62ms step:1858/4578 train_loss:3.7294 train_time:261694ms step_avg:141.61ms step:1859/4578 train_loss:3.5845 train_time:261829ms step_avg:141.61ms step:1860/4578 train_loss:3.5168 train_time:261971ms step_avg:141.61ms step:1861/4578 train_loss:3.5611 train_time:262109ms step_avg:141.60ms step:1862/4578 train_loss:3.4542 train_time:262256ms step_avg:141.61ms step:1863/4578 train_loss:3.4470 train_time:262391ms step_avg:141.60ms step:1864/4578 train_loss:3.5185 train_time:262529ms step_avg:141.60ms step:1865/4578 train_loss:3.5503 train_time:262671ms step_avg:141.60ms step:1866/4578 train_loss:3.3184 train_time:262810ms step_avg:141.60ms step:1867/4578 train_loss:3.4494 train_time:262950ms step_avg:141.60ms step:1868/4578 train_loss:3.4069 train_time:263090ms step_avg:141.60ms step:1869/4578 train_loss:3.4063 train_time:263237ms step_avg:141.60ms step:1870/4578 train_loss:3.5627 train_time:263379ms step_avg:141.60ms step:1871/4578 train_loss:3.5460 train_time:263512ms step_avg:141.60ms step:1872/4578 train_loss:3.4974 train_time:263652ms step_avg:141.60ms step:1873/4578 train_loss:3.5086 train_time:263798ms step_avg:141.60ms step:1874/4578 train_loss:3.4400 train_time:263933ms step_avg:141.59ms step:1875/4578 train_loss:3.5376 train_time:264073ms step_avg:141.59ms step:1875/4578 val_loss:3.5127 train_time:264125ms step_avg:141.62ms step:1876/4578 train_loss:3.5402 train_time:264223ms step_avg:141.60ms step:1877/4578 train_loss:3.4651 train_time:264368ms step_avg:141.60ms step:1878/4578 train_loss:3.5050 train_time:264517ms step_avg:141.60ms step:1879/4578 train_loss:3.6257 train_time:264646ms step_avg:141.60ms step:1880/4578 train_loss:3.4985 train_time:264790ms step_avg:141.60ms step:1881/4578 train_loss:3.5513 train_time:264924ms step_avg:141.59ms step:1882/4578 train_loss:3.4677 train_time:265062ms step_avg:141.59ms step:1883/4578 train_loss:3.5413 train_time:265205ms step_avg:141.59ms step:1884/4578 train_loss:3.5359 train_time:265350ms step_avg:141.60ms step:1885/4578 train_loss:3.2872 train_time:265491ms step_avg:141.60ms step:1886/4578 train_loss:3.6905 train_time:265631ms step_avg:141.59ms step:1887/4578 train_loss:3.4184 train_time:265770ms step_avg:141.59ms step:1888/4578 train_loss:3.4388 train_time:265909ms step_avg:141.59ms step:1889/4578 train_loss:3.5099 train_time:266055ms step_avg:141.59ms step:1890/4578 train_loss:3.5593 train_time:266332ms step_avg:141.67ms step:1891/4578 train_loss:3.3862 train_time:266467ms step_avg:141.66ms step:1892/4578 train_loss:3.6576 train_time:266604ms step_avg:141.66ms step:1893/4578 train_loss:3.4090 train_time:266743ms step_avg:141.66ms step:1894/4578 train_loss:3.5395 train_time:266882ms step_avg:141.66ms step:1895/4578 train_loss:3.5794 train_time:267021ms step_avg:141.66ms step:1896/4578 train_loss:3.3839 train_time:267161ms step_avg:141.65ms step:1897/4578 train_loss:3.5409 train_time:267304ms step_avg:141.66ms step:1898/4578 train_loss:3.5079 train_time:267451ms step_avg:141.66ms step:1899/4578 train_loss:3.5857 train_time:267590ms step_avg:141.66ms step:1900/4578 train_loss:3.3622 train_time:267953ms step_avg:141.77ms step:1901/4578 train_loss:3.6071 train_time:268088ms step_avg:141.77ms step:1902/4578 train_loss:3.4926 train_time:268224ms step_avg:141.77ms step:1903/4578 train_loss:3.6486 train_time:268363ms step_avg:141.77ms step:1904/4578 train_loss:3.4545 train_time:268507ms step_avg:141.77ms step:1905/4578 train_loss:3.7355 train_time:268641ms step_avg:141.76ms step:1906/4578 train_loss:3.4680 train_time:268781ms step_avg:141.76ms step:1907/4578 train_loss:3.4586 train_time:268926ms step_avg:141.76ms step:1908/4578 train_loss:3.5376 train_time:269069ms step_avg:141.76ms step:1909/4578 train_loss:3.4187 train_time:269212ms step_avg:141.77ms step:1910/4578 train_loss:3.4874 train_time:269347ms step_avg:141.76ms step:1911/4578 train_loss:3.5777 train_time:269486ms step_avg:141.76ms step:1912/4578 train_loss:3.5062 train_time:269626ms step_avg:141.76ms step:1913/4578 train_loss:3.3799 train_time:269765ms step_avg:141.76ms step:1914/4578 train_loss:3.2618 train_time:269909ms step_avg:141.76ms step:1915/4578 train_loss:3.4543 train_time:270055ms step_avg:141.76ms step:1916/4578 train_loss:3.6787 train_time:270190ms step_avg:141.76ms step:1917/4578 train_loss:3.6709 train_time:270330ms step_avg:141.76ms step:1918/4578 train_loss:3.6195 train_time:270477ms step_avg:141.76ms step:1919/4578 train_loss:3.4541 train_time:270608ms step_avg:141.75ms step:1920/4578 train_loss:3.6931 train_time:270749ms step_avg:141.75ms step:1921/4578 train_loss:3.5129 train_time:270889ms step_avg:141.75ms step:1922/4578 train_loss:3.4462 train_time:271030ms step_avg:141.75ms step:1923/4578 train_loss:3.6259 train_time:271170ms step_avg:141.75ms step:1924/4578 train_loss:3.5886 train_time:271310ms step_avg:141.75ms step:1925/4578 train_loss:3.4312 train_time:271449ms step_avg:141.75ms step:1926/4578 train_loss:3.4600 train_time:271589ms step_avg:141.75ms step:1927/4578 train_loss:3.3699 train_time:271730ms step_avg:141.75ms step:1928/4578 train_loss:3.4828 train_time:271870ms step_avg:141.75ms step:1929/4578 train_loss:3.3461 train_time:272019ms step_avg:141.75ms step:1930/4578 train_loss:3.4540 train_time:272150ms step_avg:141.75ms step:1931/4578 train_loss:3.5938 train_time:272290ms step_avg:141.74ms step:1932/4578 train_loss:3.4561 train_time:272439ms step_avg:141.75ms step:1933/4578 train_loss:3.5953 train_time:272569ms step_avg:141.74ms step:1934/4578 train_loss:3.4721 train_time:272709ms step_avg:141.74ms step:1935/4578 train_loss:3.5189 train_time:272853ms step_avg:141.74ms step:1936/4578 train_loss:3.5481 train_time:272990ms step_avg:141.74ms step:1937/4578 train_loss:3.5046 train_time:273130ms step_avg:141.74ms step:1938/4578 train_loss:3.5346 train_time:273273ms step_avg:141.74ms step:1939/4578 train_loss:3.4583 train_time:273410ms step_avg:141.74ms step:1940/4578 train_loss:3.5505 train_time:273550ms step_avg:141.74ms step:1941/4578 train_loss:3.5911 train_time:273690ms step_avg:141.73ms step:1942/4578 train_loss:3.4284 train_time:273830ms step_avg:141.73ms step:1943/4578 train_loss:3.4662 train_time:273970ms step_avg:141.73ms step:1944/4578 train_loss:3.5320 train_time:274112ms step_avg:141.73ms step:1945/4578 train_loss:3.3762 train_time:274255ms step_avg:141.73ms step:1946/4578 train_loss:3.6480 train_time:274398ms step_avg:141.73ms step:1947/4578 train_loss:3.5183 train_time:274544ms step_avg:141.74ms step:1948/4578 train_loss:3.4958 train_time:274677ms step_avg:141.73ms step:1949/4578 train_loss:3.4971 train_time:274811ms step_avg:141.73ms step:1950/4578 train_loss:3.3777 train_time:274954ms step_avg:141.73ms step:1951/4578 train_loss:3.4981 train_time:275094ms step_avg:141.73ms step:1952/4578 train_loss:3.3425 train_time:275233ms step_avg:141.73ms step:1953/4578 train_loss:3.5565 train_time:275372ms step_avg:141.73ms step:1954/4578 train_loss:3.5525 train_time:275514ms step_avg:141.73ms step:1955/4578 train_loss:3.5025 train_time:275652ms step_avg:141.72ms step:1956/4578 train_loss:3.3892 train_time:275791ms step_avg:141.72ms step:1957/4578 train_loss:3.4815 train_time:275932ms step_avg:141.72ms step:1958/4578 train_loss:3.6640 train_time:276072ms step_avg:141.72ms step:1959/4578 train_loss:3.5834 train_time:276211ms step_avg:141.72ms step:1960/4578 train_loss:3.6073 train_time:276352ms step_avg:141.72ms step:1961/4578 train_loss:3.3994 train_time:276491ms step_avg:141.72ms step:1962/4578 train_loss:3.5282 train_time:276631ms step_avg:141.72ms step:1963/4578 train_loss:3.5770 train_time:276771ms step_avg:141.72ms step:1964/4578 train_loss:3.5146 train_time:276911ms step_avg:141.71ms step:1965/4578 train_loss:3.4276 train_time:277056ms step_avg:141.72ms step:1966/4578 train_loss:3.8310 train_time:277191ms step_avg:141.71ms step:1967/4578 train_loss:3.4498 train_time:277332ms step_avg:141.71ms step:1968/4578 train_loss:3.4892 train_time:277473ms step_avg:141.71ms step:1969/4578 train_loss:3.5426 train_time:277610ms step_avg:141.71ms step:1970/4578 train_loss:3.5063 train_time:277751ms step_avg:141.71ms step:1971/4578 train_loss:3.3890 train_time:277897ms step_avg:141.71ms step:1972/4578 train_loss:3.3649 train_time:278033ms step_avg:141.71ms step:1973/4578 train_loss:3.4913 train_time:278171ms step_avg:141.71ms step:1974/4578 train_loss:3.4572 train_time:278319ms step_avg:141.71ms step:1975/4578 train_loss:3.4295 train_time:278451ms step_avg:141.71ms step:1976/4578 train_loss:3.5924 train_time:278590ms step_avg:141.70ms step:1977/4578 train_loss:3.4627 train_time:278732ms step_avg:141.70ms step:1978/4578 train_loss:3.8253 train_time:278871ms step_avg:141.70ms step:1979/4578 train_loss:3.5050 train_time:279025ms step_avg:141.71ms step:1980/4578 train_loss:3.5097 train_time:279152ms step_avg:141.70ms step:1981/4578 train_loss:3.5125 train_time:279290ms step_avg:141.70ms step:1982/4578 train_loss:3.5380 train_time:279431ms step_avg:141.70ms step:1983/4578 train_loss:3.4702 train_time:279571ms step_avg:141.70ms step:1984/4578 train_loss:3.4341 train_time:279711ms step_avg:141.70ms step:1985/4578 train_loss:3.4837 train_time:279852ms step_avg:141.70ms step:1986/4578 train_loss:3.5490 train_time:279991ms step_avg:141.70ms step:1987/4578 train_loss:3.5219 train_time:280132ms step_avg:141.70ms step:1988/4578 train_loss:3.4939 train_time:280274ms step_avg:141.70ms step:1989/4578 train_loss:3.5786 train_time:280410ms step_avg:141.69ms step:1990/4578 train_loss:3.6102 train_time:280561ms step_avg:141.70ms step:1991/4578 train_loss:3.3842 train_time:280692ms step_avg:141.69ms step:1992/4578 train_loss:3.3887 train_time:280830ms step_avg:141.69ms step:1993/4578 train_loss:3.5720 train_time:280971ms step_avg:141.69ms step:1994/4578 train_loss:3.3921 train_time:281109ms step_avg:141.69ms step:1995/4578 train_loss:3.4773 train_time:281250ms step_avg:141.69ms step:1996/4578 train_loss:3.5549 train_time:281389ms step_avg:141.69ms step:1997/4578 train_loss:3.4182 train_time:281530ms step_avg:141.69ms step:1998/4578 train_loss:3.5212 train_time:281670ms step_avg:141.69ms step:1999/4578 train_loss:3.5216 train_time:281810ms step_avg:141.68ms step:2000/4578 train_loss:3.4449 train_time:281952ms step_avg:141.68ms step:2000/4578 val_loss:3.4991 train_time:282005ms step_avg:141.71ms step:2001/4578 train_loss:3.5885 train_time:282098ms step_avg:141.69ms step:2002/4578 train_loss:3.5326 train_time:282241ms step_avg:141.69ms step:2003/4578 train_loss:3.6205 train_time:282381ms step_avg:141.69ms step:2004/4578 train_loss:3.5378 train_time:282533ms step_avg:141.69ms step:2005/4578 train_loss:3.5540 train_time:282659ms step_avg:141.68ms step:2006/4578 train_loss:3.4347 train_time:282799ms step_avg:141.68ms step:2007/4578 train_loss:3.4655 train_time:282939ms step_avg:141.68ms step:2008/4578 train_loss:3.5121 train_time:283090ms step_avg:141.69ms step:2009/4578 train_loss:3.5550 train_time:283227ms step_avg:141.68ms step:2010/4578 train_loss:3.4517 train_time:283370ms step_avg:141.68ms step:2011/4578 train_loss:3.5308 train_time:283507ms step_avg:141.68ms step:2012/4578 train_loss:3.5054 train_time:283647ms step_avg:141.68ms step:2013/4578 train_loss:3.5130 train_time:283785ms step_avg:141.68ms step:2014/4578 train_loss:3.4373 train_time:283926ms step_avg:141.68ms step:2015/4578 train_loss:3.4735 train_time:284068ms step_avg:141.68ms step:2016/4578 train_loss:3.4947 train_time:284209ms step_avg:141.68ms step:2017/4578 train_loss:3.6250 train_time:284349ms step_avg:141.68ms step:2018/4578 train_loss:3.4724 train_time:284490ms step_avg:141.68ms step:2019/4578 train_loss:3.6281 train_time:284628ms step_avg:141.68ms step:2020/4578 train_loss:3.6381 train_time:284778ms step_avg:141.68ms step:2021/4578 train_loss:3.3441 train_time:284908ms step_avg:141.67ms step:2022/4578 train_loss:3.5773 train_time:285050ms step_avg:141.68ms step:2023/4578 train_loss:3.5016 train_time:285195ms step_avg:141.68ms step:2024/4578 train_loss:3.5971 train_time:285331ms step_avg:141.67ms step:2025/4578 train_loss:3.6406 train_time:285469ms step_avg:141.67ms step:2026/4578 train_loss:3.4269 train_time:285608ms step_avg:141.67ms step:2027/4578 train_loss:3.4624 train_time:285748ms step_avg:141.67ms step:2028/4578 train_loss:3.3710 train_time:285888ms step_avg:141.67ms step:2029/4578 train_loss:3.4804 train_time:286029ms step_avg:141.67ms step:2030/4578 train_loss:3.4013 train_time:286168ms step_avg:141.67ms step:2031/4578 train_loss:3.4902 train_time:286309ms step_avg:141.67ms step:2032/4578 train_loss:3.4870 train_time:286448ms step_avg:141.67ms step:2033/4578 train_loss:3.4971 train_time:286587ms step_avg:141.66ms step:2034/4578 train_loss:3.3969 train_time:286727ms step_avg:141.66ms step:2035/4578 train_loss:3.5570 train_time:286868ms step_avg:141.66ms step:2036/4578 train_loss:3.5571 train_time:287015ms step_avg:141.67ms step:2037/4578 train_loss:3.5432 train_time:287148ms step_avg:141.66ms step:2038/4578 train_loss:3.4136 train_time:287287ms step_avg:141.66ms step:2039/4578 train_loss:3.6723 train_time:287428ms step_avg:141.66ms step:2040/4578 train_loss:3.5097 train_time:287568ms step_avg:141.66ms step:2041/4578 train_loss:3.5318 train_time:287708ms step_avg:141.66ms step:2042/4578 train_loss:3.4799 train_time:287848ms step_avg:141.66ms step:2043/4578 train_loss:3.3751 train_time:287987ms step_avg:141.66ms step:2044/4578 train_loss:3.5028 train_time:288128ms step_avg:141.66ms step:2045/4578 train_loss:3.4985 train_time:288267ms step_avg:141.65ms step:2046/4578 train_loss:3.3632 train_time:288409ms step_avg:141.65ms step:2047/4578 train_loss:3.4443 train_time:288548ms step_avg:141.65ms step:2048/4578 train_loss:3.5164 train_time:288688ms step_avg:141.65ms step:2049/4578 train_loss:3.4630 train_time:288828ms step_avg:141.65ms step:2050/4578 train_loss:3.5150 train_time:288968ms step_avg:141.65ms step:2051/4578 train_loss:3.6583 train_time:289108ms step_avg:141.65ms step:2052/4578 train_loss:3.5251 train_time:289249ms step_avg:141.65ms step:2053/4578 train_loss:3.4757 train_time:289387ms step_avg:141.65ms step:2054/4578 train_loss:3.4563 train_time:289531ms step_avg:141.65ms step:2055/4578 train_loss:3.3251 train_time:289674ms step_avg:141.65ms step:2056/4578 train_loss:3.4378 train_time:289808ms step_avg:141.65ms step:2057/4578 train_loss:3.6145 train_time:289953ms step_avg:141.65ms step:2058/4578 train_loss:3.6323 train_time:290097ms step_avg:141.65ms step:2059/4578 train_loss:3.4950 train_time:290228ms step_avg:141.64ms step:2060/4578 train_loss:3.5394 train_time:290367ms step_avg:141.64ms step:2061/4578 train_loss:3.5252 train_time:290521ms step_avg:141.65ms step:2062/4578 train_loss:3.4704 train_time:290648ms step_avg:141.64ms step:2063/4578 train_loss:3.3933 train_time:290788ms step_avg:141.64ms step:2064/4578 train_loss:3.6995 train_time:290931ms step_avg:141.64ms step:2065/4578 train_loss:3.5556 train_time:291070ms step_avg:141.64ms step:2066/4578 train_loss:3.5051 train_time:291210ms step_avg:141.64ms step:2067/4578 train_loss:3.5518 train_time:291347ms step_avg:141.64ms step:2068/4578 train_loss:3.4537 train_time:291487ms step_avg:141.64ms step:2069/4578 train_loss:3.5076 train_time:291628ms step_avg:141.64ms step:2070/4578 train_loss:3.6357 train_time:291767ms step_avg:141.63ms step:2071/4578 train_loss:3.6423 train_time:291908ms step_avg:141.63ms step:2072/4578 train_loss:3.4934 train_time:292048ms step_avg:141.63ms step:2073/4578 train_loss:3.5313 train_time:292188ms step_avg:141.63ms step:2074/4578 train_loss:3.4185 train_time:292329ms step_avg:141.63ms step:2075/4578 train_loss:3.9483 train_time:292468ms step_avg:141.63ms step:2076/4578 train_loss:3.3770 train_time:292608ms step_avg:141.63ms step:2077/4578 train_loss:3.5349 train_time:292749ms step_avg:141.63ms step:2078/4578 train_loss:3.4238 train_time:292889ms step_avg:141.63ms step:2079/4578 train_loss:3.4031 train_time:293173ms step_avg:141.70ms step:2080/4578 train_loss:3.4970 train_time:293307ms step_avg:141.69ms step:2081/4578 train_loss:3.7716 train_time:293446ms step_avg:141.69ms step:2082/4578 train_loss:3.3812 train_time:293585ms step_avg:141.69ms step:2083/4578 train_loss:3.7125 train_time:293725ms step_avg:141.69ms step:2084/4578 train_loss:3.4280 train_time:293868ms step_avg:141.69ms step:2085/4578 train_loss:3.4007 train_time:294001ms step_avg:141.69ms step:2086/4578 train_loss:3.6511 train_time:294146ms step_avg:141.69ms step:2087/4578 train_loss:3.5756 train_time:294289ms step_avg:141.69ms step:2088/4578 train_loss:3.5595 train_time:294429ms step_avg:141.69ms step:2089/4578 train_loss:3.6239 train_time:294568ms step_avg:141.69ms step:2090/4578 train_loss:3.5395 train_time:294898ms step_avg:141.78ms step:2091/4578 train_loss:3.5319 train_time:295039ms step_avg:141.78ms step:2092/4578 train_loss:3.4842 train_time:295178ms step_avg:141.78ms step:2093/4578 train_loss:3.5570 train_time:295317ms step_avg:141.77ms step:2094/4578 train_loss:3.4556 train_time:295457ms step_avg:141.77ms step:2095/4578 train_loss:3.2528 train_time:295595ms step_avg:141.77ms step:2096/4578 train_loss:3.4769 train_time:295733ms step_avg:141.77ms step:2097/4578 train_loss:3.6516 train_time:295877ms step_avg:141.77ms step:2098/4578 train_loss:3.4776 train_time:296018ms step_avg:141.77ms step:2099/4578 train_loss:3.3673 train_time:296162ms step_avg:141.77ms step:2100/4578 train_loss:3.4695 train_time:296299ms step_avg:141.77ms step:2101/4578 train_loss:3.4325 train_time:296440ms step_avg:141.77ms step:2102/4578 train_loss:3.5698 train_time:296580ms step_avg:141.77ms step:2103/4578 train_loss:3.3998 train_time:296722ms step_avg:141.77ms step:2104/4578 train_loss:3.3727 train_time:296862ms step_avg:141.77ms step:2105/4578 train_loss:3.6303 train_time:297004ms step_avg:141.77ms step:2106/4578 train_loss:3.3613 train_time:297147ms step_avg:141.77ms step:2107/4578 train_loss:3.7593 train_time:297285ms step_avg:141.77ms step:2108/4578 train_loss:3.5948 train_time:297426ms step_avg:141.77ms step:2109/4578 train_loss:3.5034 train_time:297566ms step_avg:141.77ms step:2110/4578 train_loss:3.5188 train_time:297706ms step_avg:141.76ms step:2111/4578 train_loss:3.3397 train_time:297846ms step_avg:141.76ms step:2112/4578 train_loss:3.8259 train_time:297986ms step_avg:141.76ms step:2113/4578 train_loss:3.5196 train_time:298128ms step_avg:141.76ms step:2114/4578 train_loss:3.4418 train_time:298268ms step_avg:141.76ms step:2115/4578 train_loss:3.5594 train_time:298408ms step_avg:141.76ms step:2116/4578 train_loss:3.5152 train_time:298550ms step_avg:141.76ms step:2117/4578 train_loss:3.5046 train_time:298686ms step_avg:141.76ms step:2118/4578 train_loss:3.5575 train_time:298829ms step_avg:141.76ms step:2119/4578 train_loss:3.4105 train_time:298968ms step_avg:141.76ms step:2120/4578 train_loss:3.4765 train_time:299110ms step_avg:141.76ms step:2121/4578 train_loss:3.1802 train_time:299250ms step_avg:141.76ms step:2122/4578 train_loss:3.3818 train_time:299389ms step_avg:141.76ms step:2123/4578 train_loss:3.5341 train_time:299529ms step_avg:141.76ms step:2124/4578 train_loss:3.4572 train_time:299669ms step_avg:141.75ms step:2125/4578 train_loss:3.6146 train_time:299809ms step_avg:141.75ms step:2125/4578 val_loss:3.4888 train_time:299865ms step_avg:141.78ms step:2126/4578 train_loss:3.4737 train_time:299961ms step_avg:141.76ms step:2127/4578 train_loss:3.5900 train_time:300108ms step_avg:141.76ms step:2128/4578 train_loss:3.5722 train_time:300247ms step_avg:141.76ms step:2129/4578 train_loss:3.4278 train_time:300385ms step_avg:141.76ms step:2130/4578 train_loss:3.4146 train_time:300523ms step_avg:141.76ms step:2131/4578 train_loss:3.4443 train_time:300662ms step_avg:141.75ms step:2132/4578 train_loss:3.5934 train_time:300801ms step_avg:141.75ms step:2133/4578 train_loss:3.4743 train_time:300943ms step_avg:141.75ms step:2134/4578 train_loss:3.3842 train_time:301089ms step_avg:141.76ms step:2135/4578 train_loss:3.4475 train_time:301231ms step_avg:141.76ms step:2136/4578 train_loss:3.5696 train_time:301372ms step_avg:141.76ms step:2137/4578 train_loss:3.5817 train_time:301511ms step_avg:141.75ms step:2138/4578 train_loss:3.5239 train_time:301650ms step_avg:141.75ms step:2139/4578 train_loss:3.5104 train_time:301790ms step_avg:141.75ms step:2140/4578 train_loss:3.5011 train_time:301936ms step_avg:141.75ms step:2141/4578 train_loss:3.5838 train_time:302072ms step_avg:141.75ms step:2142/4578 train_loss:3.8930 train_time:302212ms step_avg:141.75ms step:2143/4578 train_loss:3.4155 train_time:302353ms step_avg:141.75ms step:2144/4578 train_loss:3.4454 train_time:302492ms step_avg:141.75ms step:2145/4578 train_loss:3.4901 train_time:302631ms step_avg:141.75ms step:2146/4578 train_loss:3.6102 train_time:302770ms step_avg:141.75ms step:2147/4578 train_loss:3.5453 train_time:302910ms step_avg:141.75ms step:2148/4578 train_loss:3.9658 train_time:303051ms step_avg:141.75ms step:2149/4578 train_loss:3.4746 train_time:303191ms step_avg:141.74ms step:2150/4578 train_loss:3.4457 train_time:303330ms step_avg:141.74ms step:2151/4578 train_loss:3.5131 train_time:303472ms step_avg:141.74ms step:2152/4578 train_loss:3.5419 train_time:303611ms step_avg:141.74ms step:2153/4578 train_loss:3.4974 train_time:303753ms step_avg:141.74ms step:2154/4578 train_loss:3.4301 train_time:303891ms step_avg:141.74ms step:2155/4578 train_loss:3.6419 train_time:304031ms step_avg:141.74ms step:2156/4578 train_loss:3.2656 train_time:304172ms step_avg:141.74ms step:2157/4578 train_loss:3.4230 train_time:304312ms step_avg:141.74ms step:2158/4578 train_loss:3.5630 train_time:304456ms step_avg:141.74ms step:2159/4578 train_loss:3.4983 train_time:304591ms step_avg:141.74ms step:2160/4578 train_loss:3.6558 train_time:304730ms step_avg:141.73ms step:2161/4578 train_loss:3.5615 train_time:304871ms step_avg:141.73ms step:2162/4578 train_loss:3.4933 train_time:305010ms step_avg:141.73ms step:2163/4578 train_loss:3.4659 train_time:305151ms step_avg:141.73ms step:2164/4578 train_loss:3.4616 train_time:305293ms step_avg:141.73ms step:2165/4578 train_loss:3.5469 train_time:305431ms step_avg:141.73ms step:2166/4578 train_loss:3.5706 train_time:305572ms step_avg:141.73ms step:2167/4578 train_loss:3.5020 train_time:305716ms step_avg:141.73ms step:2168/4578 train_loss:3.4004 train_time:305851ms step_avg:141.73ms step:2169/4578 train_loss:3.4794 train_time:305991ms step_avg:141.73ms step:2170/4578 train_loss:3.5185 train_time:306130ms step_avg:141.73ms step:2171/4578 train_loss:3.6439 train_time:306270ms step_avg:141.73ms step:2172/4578 train_loss:3.4399 train_time:306410ms step_avg:141.73ms step:2173/4578 train_loss:3.4254 train_time:306551ms step_avg:141.72ms step:2174/4578 train_loss:3.4356 train_time:306690ms step_avg:141.72ms step:2175/4578 train_loss:3.4895 train_time:306829ms step_avg:141.72ms step:2176/4578 train_loss:3.4537 train_time:306969ms step_avg:141.72ms step:2177/4578 train_loss:3.4237 train_time:307108ms step_avg:141.72ms step:2178/4578 train_loss:3.6459 train_time:307248ms step_avg:141.72ms step:2179/4578 train_loss:3.4645 train_time:307388ms step_avg:141.72ms step:2180/4578 train_loss:3.4766 train_time:307529ms step_avg:141.72ms step:2181/4578 train_loss:3.5398 train_time:307668ms step_avg:141.72ms step:2182/4578 train_loss:3.5126 train_time:307807ms step_avg:141.72ms step:2183/4578 train_loss:3.4831 train_time:307949ms step_avg:141.72ms step:2184/4578 train_loss:3.3797 train_time:308088ms step_avg:141.71ms step:2185/4578 train_loss:3.5480 train_time:308228ms step_avg:141.71ms step:2186/4578 train_loss:3.7181 train_time:308367ms step_avg:141.71ms step:2187/4578 train_loss:3.3641 train_time:308507ms step_avg:141.71ms step:2188/4578 train_loss:3.4071 train_time:308647ms step_avg:141.71ms step:2189/4578 train_loss:3.2595 train_time:308787ms step_avg:141.71ms step:2190/4578 train_loss:3.4102 train_time:308927ms step_avg:141.71ms step:2191/4578 train_loss:3.5571 train_time:309066ms step_avg:141.71ms step:2192/4578 train_loss:3.4859 train_time:309206ms step_avg:141.71ms step:2193/4578 train_loss:3.7180 train_time:309346ms step_avg:141.71ms step:2194/4578 train_loss:3.4850 train_time:309486ms step_avg:141.71ms step:2195/4578 train_loss:3.5541 train_time:309627ms step_avg:141.71ms step:2196/4578 train_loss:3.4896 train_time:309767ms step_avg:141.70ms step:2197/4578 train_loss:3.4116 train_time:309913ms step_avg:141.71ms step:2198/4578 train_loss:3.4938 train_time:310048ms step_avg:141.70ms step:2199/4578 train_loss:3.4393 train_time:310187ms step_avg:141.70ms step:2200/4578 train_loss:3.4440 train_time:310327ms step_avg:141.70ms step:2201/4578 train_loss:3.4919 train_time:310466ms step_avg:141.70ms step:2202/4578 train_loss:3.4748 train_time:310607ms step_avg:141.70ms step:2203/4578 train_loss:3.4517 train_time:310748ms step_avg:141.70ms step:2204/4578 train_loss:3.9532 train_time:310886ms step_avg:141.70ms step:2205/4578 train_loss:3.3664 train_time:311026ms step_avg:141.70ms step:2206/4578 train_loss:3.4933 train_time:311168ms step_avg:141.70ms step:2207/4578 train_loss:3.5042 train_time:311306ms step_avg:141.70ms step:2208/4578 train_loss:3.5245 train_time:311446ms step_avg:141.70ms step:2209/4578 train_loss:3.4176 train_time:311586ms step_avg:141.69ms step:2210/4578 train_loss:3.4943 train_time:311728ms step_avg:141.69ms step:2211/4578 train_loss:3.5027 train_time:311868ms step_avg:141.69ms step:2212/4578 train_loss:3.5019 train_time:312007ms step_avg:141.69ms step:2213/4578 train_loss:3.5291 train_time:312147ms step_avg:141.69ms step:2214/4578 train_loss:3.3820 train_time:312287ms step_avg:141.69ms step:2215/4578 train_loss:3.4512 train_time:312429ms step_avg:141.69ms step:2216/4578 train_loss:3.5856 train_time:312567ms step_avg:141.69ms step:2217/4578 train_loss:3.5403 train_time:312707ms step_avg:141.69ms step:2218/4578 train_loss:3.5012 train_time:312848ms step_avg:141.69ms step:2219/4578 train_loss:3.5134 train_time:312987ms step_avg:141.69ms step:2220/4578 train_loss:3.4163 train_time:313127ms step_avg:141.69ms step:2221/4578 train_loss:3.6766 train_time:313267ms step_avg:141.69ms step:2222/4578 train_loss:3.5593 train_time:313407ms step_avg:141.68ms step:2223/4578 train_loss:3.5913 train_time:313548ms step_avg:141.68ms step:2224/4578 train_loss:3.4772 train_time:313686ms step_avg:141.68ms step:2225/4578 train_loss:3.5933 train_time:313827ms step_avg:141.68ms step:2226/4578 train_loss:3.3505 train_time:313967ms step_avg:141.68ms step:2227/4578 train_loss:3.6180 train_time:314109ms step_avg:141.68ms step:2228/4578 train_loss:3.5505 train_time:314247ms step_avg:141.68ms step:2229/4578 train_loss:3.3540 train_time:314386ms step_avg:141.68ms step:2230/4578 train_loss:3.6983 train_time:314527ms step_avg:141.68ms step:2231/4578 train_loss:3.3901 train_time:314666ms step_avg:141.68ms step:2232/4578 train_loss:3.8616 train_time:314806ms step_avg:141.68ms step:2233/4578 train_loss:3.5439 train_time:314946ms step_avg:141.68ms step:2234/4578 train_loss:3.4927 train_time:315086ms step_avg:141.68ms step:2235/4578 train_loss:3.5368 train_time:315227ms step_avg:141.68ms step:2236/4578 train_loss:3.3052 train_time:315367ms step_avg:141.67ms step:2237/4578 train_loss:3.3085 train_time:315506ms step_avg:141.67ms step:2238/4578 train_loss:3.5333 train_time:315646ms step_avg:141.67ms step:2239/4578 train_loss:3.6287 train_time:315786ms step_avg:141.67ms step:2240/4578 train_loss:3.3481 train_time:315926ms step_avg:141.67ms step:2241/4578 train_loss:3.4196 train_time:316066ms step_avg:141.67ms step:2242/4578 train_loss:3.5997 train_time:316206ms step_avg:141.67ms step:2243/4578 train_loss:3.5667 train_time:316346ms step_avg:141.67ms step:2244/4578 train_loss:3.4216 train_time:316486ms step_avg:141.67ms step:2245/4578 train_loss:3.4917 train_time:316626ms step_avg:141.67ms step:2246/4578 train_loss:3.5156 train_time:316766ms step_avg:141.67ms step:2247/4578 train_loss:3.3472 train_time:316906ms step_avg:141.67ms step:2248/4578 train_loss:3.3673 train_time:317047ms step_avg:141.67ms step:2249/4578 train_loss:3.6307 train_time:317187ms step_avg:141.66ms step:2250/4578 train_loss:3.3603 train_time:317327ms step_avg:141.66ms step:2250/4578 val_loss:3.4784 train_time:317382ms step_avg:141.69ms step:2251/4578 train_loss:3.3553 train_time:317477ms step_avg:141.67ms step:2252/4578 train_loss:3.4207 train_time:317623ms step_avg:141.67ms step:2253/4578 train_loss:3.3982 train_time:317764ms step_avg:141.67ms step:2254/4578 train_loss:3.4540 train_time:317902ms step_avg:141.67ms step:2255/4578 train_loss:3.5057 train_time:318041ms step_avg:141.67ms step:2256/4578 train_loss:3.3830 train_time:318179ms step_avg:141.66ms step:2257/4578 train_loss:3.6667 train_time:318316ms step_avg:141.66ms step:2258/4578 train_loss:3.5444 train_time:318459ms step_avg:141.66ms step:2259/4578 train_loss:3.8586 train_time:318603ms step_avg:141.66ms step:2260/4578 train_loss:3.5451 train_time:318744ms step_avg:141.66ms step:2261/4578 train_loss:3.5922 train_time:318883ms step_avg:141.66ms step:2262/4578 train_loss:3.5092 train_time:319022ms step_avg:141.66ms step:2263/4578 train_loss:3.5103 train_time:319162ms step_avg:141.66ms step:2264/4578 train_loss:3.2658 train_time:319303ms step_avg:141.66ms step:2265/4578 train_loss:3.3967 train_time:319442ms step_avg:141.66ms step:2266/4578 train_loss:3.6068 train_time:319584ms step_avg:141.66ms step:2267/4578 train_loss:3.3420 train_time:319725ms step_avg:141.66ms step:2268/4578 train_loss:3.4119 train_time:320056ms step_avg:141.74ms step:2269/4578 train_loss:3.3920 train_time:320190ms step_avg:141.74ms step:2270/4578 train_loss:3.3563 train_time:320329ms step_avg:141.74ms step:2271/4578 train_loss:3.7535 train_time:320469ms step_avg:141.74ms step:2272/4578 train_loss:3.4132 train_time:320608ms step_avg:141.74ms step:2273/4578 train_loss:3.4190 train_time:320746ms step_avg:141.74ms step:2274/4578 train_loss:3.4949 train_time:320886ms step_avg:141.73ms step:2275/4578 train_loss:3.4472 train_time:321029ms step_avg:141.73ms step:2276/4578 train_loss:3.4646 train_time:321172ms step_avg:141.74ms step:2277/4578 train_loss:3.3463 train_time:321312ms step_avg:141.73ms step:2278/4578 train_loss:3.4523 train_time:321453ms step_avg:141.73ms step:2279/4578 train_loss:3.5814 train_time:321592ms step_avg:141.73ms step:2280/4578 train_loss:3.3709 train_time:321893ms step_avg:141.80ms step:2281/4578 train_loss:3.4365 train_time:322028ms step_avg:141.80ms step:2282/4578 train_loss:3.4498 train_time:322167ms step_avg:141.80ms step:2283/4578 train_loss:3.5965 train_time:322307ms step_avg:141.80ms step:2284/4578 train_loss:3.4687 train_time:322448ms step_avg:141.80ms step:2285/4578 train_loss:3.4914 train_time:322585ms step_avg:141.80ms step:2286/4578 train_loss:3.4937 train_time:322733ms step_avg:141.80ms step:2287/4578 train_loss:3.4860 train_time:322868ms step_avg:141.80ms step:2288/4578 train_loss:3.4438 train_time:323016ms step_avg:141.80ms step:2289/4578 train_loss:3.5796 train_time:323158ms step_avg:141.80ms step:2290/4578 train_loss:3.5456 train_time:323299ms step_avg:141.80ms step:2291/4578 train_loss:3.4296 train_time:323431ms step_avg:141.79ms step:2292/4578 train_loss:3.7692 train_time:323580ms step_avg:141.80ms step:2293/4578 train_loss:3.4341 train_time:323711ms step_avg:141.79ms step:2294/4578 train_loss:3.3787 train_time:323854ms step_avg:141.79ms step:2295/4578 train_loss:3.5581 train_time:324000ms step_avg:141.79ms step:2296/4578 train_loss:3.5063 train_time:324137ms step_avg:141.79ms step:2297/4578 train_loss:3.4842 train_time:324276ms step_avg:141.79ms step:2298/4578 train_loss:3.8590 train_time:324416ms step_avg:141.79ms step:2299/4578 train_loss:3.3798 train_time:324557ms step_avg:141.79ms step:2300/4578 train_loss:3.3804 train_time:324696ms step_avg:141.79ms step:2301/4578 train_loss:3.7239 train_time:324837ms step_avg:141.79ms step:2302/4578 train_loss:3.4429 train_time:324978ms step_avg:141.79ms step:2303/4578 train_loss:3.4576 train_time:325118ms step_avg:141.79ms step:2304/4578 train_loss:3.4463 train_time:325258ms step_avg:141.79ms step:2305/4578 train_loss:3.3803 train_time:325398ms step_avg:141.79ms step:2306/4578 train_loss:3.5420 train_time:325549ms step_avg:141.79ms step:2307/4578 train_loss:3.4018 train_time:325680ms step_avg:141.78ms step:2308/4578 train_loss:3.4186 train_time:325817ms step_avg:141.78ms step:2309/4578 train_loss:3.5557 train_time:325959ms step_avg:141.78ms step:2310/4578 train_loss:3.5060 train_time:326098ms step_avg:141.78ms step:2311/4578 train_loss:3.3799 train_time:326238ms step_avg:141.78ms step:2312/4578 train_loss:3.4950 train_time:326378ms step_avg:141.78ms step:2313/4578 train_loss:3.6146 train_time:326517ms step_avg:141.78ms step:2314/4578 train_loss:3.4386 train_time:326665ms step_avg:141.78ms step:2315/4578 train_loss:3.3620 train_time:326798ms step_avg:141.78ms step:2316/4578 train_loss:3.4515 train_time:326938ms step_avg:141.78ms step:2317/4578 train_loss:3.3370 train_time:327082ms step_avg:141.78ms step:2318/4578 train_loss:3.4412 train_time:327218ms step_avg:141.78ms step:2319/4578 train_loss:3.4616 train_time:327358ms step_avg:141.77ms step:2320/4578 train_loss:3.3045 train_time:327498ms step_avg:141.77ms step:2321/4578 train_loss:3.4447 train_time:327641ms step_avg:141.77ms step:2322/4578 train_loss:3.4938 train_time:327783ms step_avg:141.77ms step:2323/4578 train_loss:3.4154 train_time:327918ms step_avg:141.77ms step:2324/4578 train_loss:3.4597 train_time:328069ms step_avg:141.78ms step:2325/4578 train_loss:3.3748 train_time:328198ms step_avg:141.77ms step:2326/4578 train_loss:3.5164 train_time:328338ms step_avg:141.77ms step:2327/4578 train_loss:3.5243 train_time:328480ms step_avg:141.77ms step:2328/4578 train_loss:3.3022 train_time:328617ms step_avg:141.77ms step:2329/4578 train_loss:3.4107 train_time:328758ms step_avg:141.77ms step:2330/4578 train_loss:3.4382 train_time:328905ms step_avg:141.77ms step:2331/4578 train_loss:3.4086 train_time:329037ms step_avg:141.77ms step:2332/4578 train_loss:3.5958 train_time:329193ms step_avg:141.77ms step:2333/4578 train_loss:3.4733 train_time:329319ms step_avg:141.76ms step:2334/4578 train_loss:3.4447 train_time:329458ms step_avg:141.76ms step:2335/4578 train_loss:3.5337 train_time:329599ms step_avg:141.76ms step:2336/4578 train_loss:3.3724 train_time:329740ms step_avg:141.76ms step:2337/4578 train_loss:3.5261 train_time:329881ms step_avg:141.76ms step:2338/4578 train_loss:3.4807 train_time:330017ms step_avg:141.76ms step:2339/4578 train_loss:3.4308 train_time:330158ms step_avg:141.76ms step:2340/4578 train_loss:3.5153 train_time:330308ms step_avg:141.76ms step:2341/4578 train_loss:3.5562 train_time:330437ms step_avg:141.76ms step:2342/4578 train_loss:3.4238 train_time:330577ms step_avg:141.76ms step:2343/4578 train_loss:3.4382 train_time:330718ms step_avg:141.76ms step:2344/4578 train_loss:3.5024 train_time:330863ms step_avg:141.76ms step:2345/4578 train_loss:3.4445 train_time:330996ms step_avg:141.75ms step:2346/4578 train_loss:3.5633 train_time:331137ms step_avg:141.75ms step:2347/4578 train_loss:3.4753 train_time:331277ms step_avg:141.75ms step:2348/4578 train_loss:3.5769 train_time:331419ms step_avg:141.75ms step:2349/4578 train_loss:3.5366 train_time:331560ms step_avg:141.75ms step:2350/4578 train_loss:3.5742 train_time:331697ms step_avg:141.75ms step:2351/4578 train_loss:3.2736 train_time:331837ms step_avg:141.75ms step:2352/4578 train_loss:3.3940 train_time:331977ms step_avg:141.75ms step:2353/4578 train_loss:3.3865 train_time:332117ms step_avg:141.75ms step:2354/4578 train_loss:3.6078 train_time:332263ms step_avg:141.75ms step:2355/4578 train_loss:3.4036 train_time:332397ms step_avg:141.75ms step:2356/4578 train_loss:3.3921 train_time:332543ms step_avg:141.75ms step:2357/4578 train_loss:3.5520 train_time:332680ms step_avg:141.75ms step:2358/4578 train_loss:3.3999 train_time:332817ms step_avg:141.74ms step:2359/4578 train_loss:3.5053 train_time:332963ms step_avg:141.75ms step:2360/4578 train_loss:3.4076 train_time:333100ms step_avg:141.74ms step:2361/4578 train_loss:3.4182 train_time:333244ms step_avg:141.75ms step:2362/4578 train_loss:3.4558 train_time:333382ms step_avg:141.74ms step:2363/4578 train_loss:3.5067 train_time:333524ms step_avg:141.74ms step:2364/4578 train_loss:3.4562 train_time:333658ms step_avg:141.74ms step:2365/4578 train_loss:3.8872 train_time:333806ms step_avg:141.74ms step:2366/4578 train_loss:3.5201 train_time:333937ms step_avg:141.74ms step:2367/4578 train_loss:3.6616 train_time:334078ms step_avg:141.74ms step:2368/4578 train_loss:3.4881 train_time:334220ms step_avg:141.74ms step:2369/4578 train_loss:3.4890 train_time:334358ms step_avg:141.74ms step:2370/4578 train_loss:3.5233 train_time:334497ms step_avg:141.74ms step:2371/4578 train_loss:3.4044 train_time:334638ms step_avg:141.74ms step:2372/4578 train_loss:3.6327 train_time:334778ms step_avg:141.73ms step:2373/4578 train_loss:3.4800 train_time:334917ms step_avg:141.73ms step:2374/4578 train_loss:4.0336 train_time:335057ms step_avg:141.73ms step:2375/4578 train_loss:3.4604 train_time:335197ms step_avg:141.73ms step:2375/4578 val_loss:3.4676 train_time:335253ms step_avg:141.76ms step:2376/4578 train_loss:3.3659 train_time:335349ms step_avg:141.74ms step:2377/4578 train_loss:3.5242 train_time:335497ms step_avg:141.74ms step:2378/4578 train_loss:3.4987 train_time:335635ms step_avg:141.74ms step:2379/4578 train_loss:3.5122 train_time:335776ms step_avg:141.74ms step:2380/4578 train_loss:3.4918 train_time:335912ms step_avg:141.73ms step:2381/4578 train_loss:3.3976 train_time:336049ms step_avg:141.73ms step:2382/4578 train_loss:3.4921 train_time:336189ms step_avg:141.73ms step:2383/4578 train_loss:3.5048 train_time:336329ms step_avg:141.73ms step:2384/4578 train_loss:3.4582 train_time:336484ms step_avg:141.74ms step:2385/4578 train_loss:3.3865 train_time:336615ms step_avg:141.73ms step:2386/4578 train_loss:3.5000 train_time:336759ms step_avg:141.73ms step:2387/4578 train_loss:3.4600 train_time:336894ms step_avg:141.73ms step:2388/4578 train_loss:3.4597 train_time:337033ms step_avg:141.73ms step:2389/4578 train_loss:3.4880 train_time:337173ms step_avg:141.73ms step:2390/4578 train_loss:3.4710 train_time:337315ms step_avg:141.73ms step:2391/4578 train_loss:3.4699 train_time:337460ms step_avg:141.73ms step:2392/4578 train_loss:3.3504 train_time:337596ms step_avg:141.73ms step:2393/4578 train_loss:3.5732 train_time:337743ms step_avg:141.73ms step:2394/4578 train_loss:3.4111 train_time:337882ms step_avg:141.73ms step:2395/4578 train_loss:3.5056 train_time:338017ms step_avg:141.73ms step:2396/4578 train_loss:3.6211 train_time:338156ms step_avg:141.73ms step:2397/4578 train_loss:3.6368 train_time:338302ms step_avg:141.73ms step:2398/4578 train_loss:3.5876 train_time:338437ms step_avg:141.72ms step:2399/4578 train_loss:3.5587 train_time:338587ms step_avg:141.73ms step:2400/4578 train_loss:3.4308 train_time:338719ms step_avg:141.72ms step:2401/4578 train_loss:3.4310 train_time:338858ms step_avg:141.72ms step:2402/4578 train_loss:3.5395 train_time:339003ms step_avg:141.72ms step:2403/4578 train_loss:3.3699 train_time:339137ms step_avg:141.72ms step:2404/4578 train_loss:3.5038 train_time:339280ms step_avg:141.72ms step:2405/4578 train_loss:3.7169 train_time:339422ms step_avg:141.72ms step:2406/4578 train_loss:3.4463 train_time:339558ms step_avg:141.72ms step:2407/4578 train_loss:3.5889 train_time:339698ms step_avg:141.72ms step:2408/4578 train_loss:3.4567 train_time:339839ms step_avg:141.72ms step:2409/4578 train_loss:3.3904 train_time:339980ms step_avg:141.72ms step:2410/4578 train_loss:3.5285 train_time:340119ms step_avg:141.72ms step:2411/4578 train_loss:3.3080 train_time:340260ms step_avg:141.72ms step:2412/4578 train_loss:3.7461 train_time:340399ms step_avg:141.72ms step:2413/4578 train_loss:3.4292 train_time:340539ms step_avg:141.71ms step:2414/4578 train_loss:3.5104 train_time:340680ms step_avg:141.71ms step:2415/4578 train_loss:3.4251 train_time:340821ms step_avg:141.71ms step:2416/4578 train_loss:3.4971 train_time:340961ms step_avg:141.71ms step:2417/4578 train_loss:3.3263 train_time:341101ms step_avg:141.71ms step:2418/4578 train_loss:3.2485 train_time:341242ms step_avg:141.71ms step:2419/4578 train_loss:3.5426 train_time:341384ms step_avg:141.71ms step:2420/4578 train_loss:3.4186 train_time:341523ms step_avg:141.71ms step:2421/4578 train_loss:3.4535 train_time:341664ms step_avg:141.71ms step:2422/4578 train_loss:3.5532 train_time:341805ms step_avg:141.71ms step:2423/4578 train_loss:3.5960 train_time:341946ms step_avg:141.71ms step:2424/4578 train_loss:3.4197 train_time:342087ms step_avg:141.71ms step:2425/4578 train_loss:3.5132 train_time:342229ms step_avg:141.71ms step:2426/4578 train_loss:3.5170 train_time:342369ms step_avg:141.71ms step:2427/4578 train_loss:3.4338 train_time:342509ms step_avg:141.71ms step:2428/4578 train_loss:3.3877 train_time:342649ms step_avg:141.71ms step:2429/4578 train_loss:3.5133 train_time:342790ms step_avg:141.71ms step:2430/4578 train_loss:3.4081 train_time:342930ms step_avg:141.71ms step:2431/4578 train_loss:3.4640 train_time:343070ms step_avg:141.71ms step:2432/4578 train_loss:3.5239 train_time:343211ms step_avg:141.71ms step:2433/4578 train_loss:3.4899 train_time:343352ms step_avg:141.71ms step:2434/4578 train_loss:3.3630 train_time:343491ms step_avg:141.70ms step:2435/4578 train_loss:3.3267 train_time:343631ms step_avg:141.70ms step:2436/4578 train_loss:3.4904 train_time:343772ms step_avg:141.70ms step:2437/4578 train_loss:3.3511 train_time:343911ms step_avg:141.70ms step:2438/4578 train_loss:3.4256 train_time:344052ms step_avg:141.70ms step:2439/4578 train_loss:3.5168 train_time:344192ms step_avg:141.70ms step:2440/4578 train_loss:3.4351 train_time:344330ms step_avg:141.70ms step:2441/4578 train_loss:3.5251 train_time:344472ms step_avg:141.70ms step:2442/4578 train_loss:3.4160 train_time:344612ms step_avg:141.70ms step:2443/4578 train_loss:3.4644 train_time:344753ms step_avg:141.70ms step:2444/4578 train_loss:3.3516 train_time:344892ms step_avg:141.70ms step:2445/4578 train_loss:3.3600 train_time:345033ms step_avg:141.70ms step:2446/4578 train_loss:3.5242 train_time:345172ms step_avg:141.70ms step:2447/4578 train_loss:3.3916 train_time:345312ms step_avg:141.70ms step:2448/4578 train_loss:3.4572 train_time:345453ms step_avg:141.70ms step:2449/4578 train_loss:3.6173 train_time:345593ms step_avg:141.69ms step:2450/4578 train_loss:3.4480 train_time:345732ms step_avg:141.69ms step:2451/4578 train_loss:3.5212 train_time:345878ms step_avg:141.70ms step:2452/4578 train_loss:3.4249 train_time:346020ms step_avg:141.70ms step:2453/4578 train_loss:3.5317 train_time:346156ms step_avg:141.69ms step:2454/4578 train_loss:3.4201 train_time:346298ms step_avg:141.69ms step:2455/4578 train_loss:3.5512 train_time:346437ms step_avg:141.69ms step:2456/4578 train_loss:3.4782 train_time:346573ms step_avg:141.69ms step:2457/4578 train_loss:3.4016 train_time:346851ms step_avg:141.75ms step:2458/4578 train_loss:3.3391 train_time:346988ms step_avg:141.74ms step:2459/4578 train_loss:3.4645 train_time:347125ms step_avg:141.74ms step:2460/4578 train_loss:4.0569 train_time:347264ms step_avg:141.74ms step:2461/4578 train_loss:3.5232 train_time:347405ms step_avg:141.74ms step:2462/4578 train_loss:3.3428 train_time:347543ms step_avg:141.74ms step:2463/4578 train_loss:3.5397 train_time:347682ms step_avg:141.74ms step:2464/4578 train_loss:3.4552 train_time:347827ms step_avg:141.74ms step:2465/4578 train_loss:3.6501 train_time:347970ms step_avg:141.74ms step:2466/4578 train_loss:3.8313 train_time:348109ms step_avg:141.74ms step:2467/4578 train_loss:3.5641 train_time:348248ms step_avg:141.74ms step:2468/4578 train_loss:3.4419 train_time:348388ms step_avg:141.74ms step:2469/4578 train_loss:3.5521 train_time:348526ms step_avg:141.73ms step:2470/4578 train_loss:3.5662 train_time:348825ms step_avg:141.80ms step:2471/4578 train_loss:3.3682 train_time:348962ms step_avg:141.80ms step:2472/4578 train_loss:3.4610 train_time:349100ms step_avg:141.80ms step:2473/4578 train_loss:3.4584 train_time:349239ms step_avg:141.79ms step:2474/4578 train_loss:3.6063 train_time:349378ms step_avg:141.79ms step:2475/4578 train_loss:3.7353 train_time:349517ms step_avg:141.79ms step:2476/4578 train_loss:3.3220 train_time:349655ms step_avg:141.79ms step:2477/4578 train_loss:3.5321 train_time:349799ms step_avg:141.79ms step:2478/4578 train_loss:3.4904 train_time:349940ms step_avg:141.79ms step:2479/4578 train_loss:3.3390 train_time:350082ms step_avg:141.79ms step:2480/4578 train_loss:3.3277 train_time:350221ms step_avg:141.79ms step:2481/4578 train_loss:3.4783 train_time:350361ms step_avg:141.79ms step:2482/4578 train_loss:3.4901 train_time:350502ms step_avg:141.79ms step:2483/4578 train_loss:3.5018 train_time:350640ms step_avg:141.79ms step:2484/4578 train_loss:3.4600 train_time:350781ms step_avg:141.79ms step:2485/4578 train_loss:3.4677 train_time:350922ms step_avg:141.79ms step:2486/4578 train_loss:3.3607 train_time:351066ms step_avg:141.79ms step:2487/4578 train_loss:3.5537 train_time:351205ms step_avg:141.79ms step:2488/4578 train_loss:3.5139 train_time:351345ms step_avg:141.79ms step:2489/4578 train_loss:3.4121 train_time:351486ms step_avg:141.79ms step:2490/4578 train_loss:3.5240 train_time:351625ms step_avg:141.78ms step:2491/4578 train_loss:3.5736 train_time:351767ms step_avg:141.78ms step:2492/4578 train_loss:3.6537 train_time:351909ms step_avg:141.78ms step:2493/4578 train_loss:3.5059 train_time:352051ms step_avg:141.78ms step:2494/4578 train_loss:3.4302 train_time:352193ms step_avg:141.78ms step:2495/4578 train_loss:3.5527 train_time:352333ms step_avg:141.78ms step:2496/4578 train_loss:3.5070 train_time:352470ms step_avg:141.78ms step:2497/4578 train_loss:3.4167 train_time:352610ms step_avg:141.78ms step:2498/4578 train_loss:3.5116 train_time:352752ms step_avg:141.78ms step:2499/4578 train_loss:3.5695 train_time:352895ms step_avg:141.78ms step:2500/4578 train_loss:3.5836 train_time:353038ms step_avg:141.78ms step:2500/4578 val_loss:3.4604 train_time:353087ms step_avg:141.80ms step:2501/4578 train_loss:3.5313 train_time:353196ms step_avg:141.79ms step:2502/4578 train_loss:3.4837 train_time:353333ms step_avg:141.79ms step:2503/4578 train_loss:3.5002 train_time:353466ms step_avg:141.78ms step:2504/4578 train_loss:3.3680 train_time:353613ms step_avg:141.79ms step:2505/4578 train_loss:3.5623 train_time:353745ms step_avg:141.78ms step:2506/4578 train_loss:3.5134 train_time:353881ms step_avg:141.78ms step:2507/4578 train_loss:3.4556 train_time:354020ms step_avg:141.78ms step:2508/4578 train_loss:3.4630 train_time:354169ms step_avg:141.78ms step:2509/4578 train_loss:3.4223 train_time:354309ms step_avg:141.78ms step:2510/4578 train_loss:3.5965 train_time:354450ms step_avg:141.78ms step:2511/4578 train_loss:3.4249 train_time:354588ms step_avg:141.78ms step:2512/4578 train_loss:3.4116 train_time:354731ms step_avg:141.78ms step:2513/4578 train_loss:3.4863 train_time:354868ms step_avg:141.78ms step:2514/4578 train_loss:3.5202 train_time:355008ms step_avg:141.78ms step:2515/4578 train_loss:3.4176 train_time:355155ms step_avg:141.78ms step:2516/4578 train_loss:3.5065 train_time:355289ms step_avg:141.78ms step:2517/4578 train_loss:3.5105 train_time:355430ms step_avg:141.77ms step:2518/4578 train_loss:3.3803 train_time:355573ms step_avg:141.78ms step:2519/4578 train_loss:3.4095 train_time:355710ms step_avg:141.77ms step:2520/4578 train_loss:3.5288 train_time:355851ms step_avg:141.77ms step:2521/4578 train_loss:3.5207 train_time:355990ms step_avg:141.77ms step:2522/4578 train_loss:3.4024 train_time:356129ms step_avg:141.77ms step:2523/4578 train_loss:3.3841 train_time:356271ms step_avg:141.77ms step:2524/4578 train_loss:3.4836 train_time:356417ms step_avg:141.77ms step:2525/4578 train_loss:3.3243 train_time:356553ms step_avg:141.77ms step:2526/4578 train_loss:3.5499 train_time:356692ms step_avg:141.77ms step:2527/4578 train_loss:3.4528 train_time:356834ms step_avg:141.77ms step:2528/4578 train_loss:3.4596 train_time:356970ms step_avg:141.77ms step:2529/4578 train_loss:3.4489 train_time:357110ms step_avg:141.77ms step:2530/4578 train_loss:3.4662 train_time:357252ms step_avg:141.77ms step:2531/4578 train_loss:3.5005 train_time:357393ms step_avg:141.77ms step:2532/4578 train_loss:3.3302 train_time:357530ms step_avg:141.76ms step:2533/4578 train_loss:3.4863 train_time:357670ms step_avg:141.76ms step:2534/4578 train_loss:3.3872 train_time:357811ms step_avg:141.76ms step:2535/4578 train_loss:3.4193 train_time:357949ms step_avg:141.76ms step:2536/4578 train_loss:3.4773 train_time:358089ms step_avg:141.76ms step:2537/4578 train_loss:3.4847 train_time:358230ms step_avg:141.76ms step:2538/4578 train_loss:3.3093 train_time:358370ms step_avg:141.76ms step:2539/4578 train_loss:3.6180 train_time:358509ms step_avg:141.76ms step:2540/4578 train_loss:3.3058 train_time:358658ms step_avg:141.76ms step:2541/4578 train_loss:3.4815 train_time:358790ms step_avg:141.76ms step:2542/4578 train_loss:3.2587 train_time:358930ms step_avg:141.76ms step:2543/4578 train_loss:3.6892 train_time:359069ms step_avg:141.76ms step:2544/4578 train_loss:3.4592 train_time:359210ms step_avg:141.76ms step:2545/4578 train_loss:3.6105 train_time:359350ms step_avg:141.76ms step:2546/4578 train_loss:3.4433 train_time:359498ms step_avg:141.76ms step:2547/4578 train_loss:3.4296 train_time:359631ms step_avg:141.75ms step:2548/4578 train_loss:3.4316 train_time:359777ms step_avg:141.76ms step:2549/4578 train_loss:3.5973 train_time:359911ms step_avg:141.75ms step:2550/4578 train_loss:3.4501 train_time:360050ms step_avg:141.75ms step:2551/4578 train_loss:3.4510 train_time:360190ms step_avg:141.75ms step:2552/4578 train_loss:3.4762 train_time:360335ms step_avg:141.75ms step:2553/4578 train_loss:3.5017 train_time:360472ms step_avg:141.75ms step:2554/4578 train_loss:3.4168 train_time:360610ms step_avg:141.75ms step:2555/4578 train_loss:3.5177 train_time:360750ms step_avg:141.75ms step:2556/4578 train_loss:3.5725 train_time:360892ms step_avg:141.75ms step:2557/4578 train_loss:3.5626 train_time:361029ms step_avg:141.75ms step:2558/4578 train_loss:3.4014 train_time:361174ms step_avg:141.75ms step:2559/4578 train_loss:3.4033 train_time:361318ms step_avg:141.75ms step:2560/4578 train_loss:3.4144 train_time:361454ms step_avg:141.75ms step:2561/4578 train_loss:3.5305 train_time:361589ms step_avg:141.74ms step:2562/4578 train_loss:3.5696 train_time:361728ms step_avg:141.74ms step:2563/4578 train_loss:3.4514 train_time:361880ms step_avg:141.75ms step:2564/4578 train_loss:3.4857 train_time:362007ms step_avg:141.74ms step:2565/4578 train_loss:3.4008 train_time:362147ms step_avg:141.74ms step:2566/4578 train_loss:3.4124 train_time:362289ms step_avg:141.74ms step:2567/4578 train_loss:3.4011 train_time:362439ms step_avg:141.74ms step:2568/4578 train_loss:3.4552 train_time:362567ms step_avg:141.74ms step:2569/4578 train_loss:3.6040 train_time:362716ms step_avg:141.74ms step:2570/4578 train_loss:3.5036 train_time:362852ms step_avg:141.74ms step:2571/4578 train_loss:3.5844 train_time:362986ms step_avg:141.74ms step:2572/4578 train_loss:3.3343 train_time:363131ms step_avg:141.74ms step:2573/4578 train_loss:3.4480 train_time:363272ms step_avg:141.74ms step:2574/4578 train_loss:3.1156 train_time:363410ms step_avg:141.74ms step:2575/4578 train_loss:3.3577 train_time:363547ms step_avg:141.73ms step:2576/4578 train_loss:3.2999 train_time:363686ms step_avg:141.73ms step:2577/4578 train_loss:3.4113 train_time:363830ms step_avg:141.73ms step:2578/4578 train_loss:3.4597 train_time:363968ms step_avg:141.73ms step:2579/4578 train_loss:3.3747 train_time:364106ms step_avg:141.73ms step:2580/4578 train_loss:3.4291 train_time:364246ms step_avg:141.73ms step:2581/4578 train_loss:3.3766 train_time:364387ms step_avg:141.73ms step:2582/4578 train_loss:3.4828 train_time:364526ms step_avg:141.73ms step:2583/4578 train_loss:3.3628 train_time:364666ms step_avg:141.73ms step:2584/4578 train_loss:3.5560 train_time:364806ms step_avg:141.73ms step:2585/4578 train_loss:3.4626 train_time:364950ms step_avg:141.73ms step:2586/4578 train_loss:3.4747 train_time:365087ms step_avg:141.73ms step:2587/4578 train_loss:3.5983 train_time:365229ms step_avg:141.73ms step:2588/4578 train_loss:3.4864 train_time:365368ms step_avg:141.73ms step:2589/4578 train_loss:3.3487 train_time:365509ms step_avg:141.72ms step:2590/4578 train_loss:3.5134 train_time:365650ms step_avg:141.72ms step:2591/4578 train_loss:3.4205 train_time:365789ms step_avg:141.72ms step:2592/4578 train_loss:3.6236 train_time:365929ms step_avg:141.72ms step:2593/4578 train_loss:3.5014 train_time:366070ms step_avg:141.72ms step:2594/4578 train_loss:3.3187 train_time:366209ms step_avg:141.72ms step:2595/4578 train_loss:3.3902 train_time:366350ms step_avg:141.72ms step:2596/4578 train_loss:3.8380 train_time:366489ms step_avg:141.72ms step:2597/4578 train_loss:3.4720 train_time:366632ms step_avg:141.72ms step:2598/4578 train_loss:3.4838 train_time:366776ms step_avg:141.72ms step:2599/4578 train_loss:3.3378 train_time:366910ms step_avg:141.72ms step:2600/4578 train_loss:3.5693 train_time:367051ms step_avg:141.72ms step:2601/4578 train_loss:3.7349 train_time:367190ms step_avg:141.72ms step:2602/4578 train_loss:3.3098 train_time:367334ms step_avg:141.72ms step:2603/4578 train_loss:3.4592 train_time:367470ms step_avg:141.72ms step:2604/4578 train_loss:3.2924 train_time:367611ms step_avg:141.72ms step:2605/4578 train_loss:3.5800 train_time:367755ms step_avg:141.72ms step:2606/4578 train_loss:3.4548 train_time:367890ms step_avg:141.71ms step:2607/4578 train_loss:3.3508 train_time:368030ms step_avg:141.71ms step:2608/4578 train_loss:3.3032 train_time:368180ms step_avg:141.72ms step:2609/4578 train_loss:3.4225 train_time:368310ms step_avg:141.71ms step:2610/4578 train_loss:3.5954 train_time:368455ms step_avg:141.71ms step:2611/4578 train_loss:3.4742 train_time:368590ms step_avg:141.71ms step:2612/4578 train_loss:3.3001 train_time:368730ms step_avg:141.71ms step:2613/4578 train_loss:3.4064 train_time:368870ms step_avg:141.71ms step:2614/4578 train_loss:3.5116 train_time:369011ms step_avg:141.71ms step:2615/4578 train_loss:3.4501 train_time:369160ms step_avg:141.71ms step:2616/4578 train_loss:3.4405 train_time:369302ms step_avg:141.71ms step:2617/4578 train_loss:3.4786 train_time:369439ms step_avg:141.71ms step:2618/4578 train_loss:3.5199 train_time:369585ms step_avg:141.71ms step:2619/4578 train_loss:3.3666 train_time:369715ms step_avg:141.71ms step:2620/4578 train_loss:3.5406 train_time:369856ms step_avg:141.71ms step:2621/4578 train_loss:3.5014 train_time:369997ms step_avg:141.71ms step:2622/4578 train_loss:3.6322 train_time:370138ms step_avg:141.71ms step:2623/4578 train_loss:3.5466 train_time:370279ms step_avg:141.71ms step:2624/4578 train_loss:3.4636 train_time:370422ms step_avg:141.71ms step:2625/4578 train_loss:3.4193 train_time:370561ms step_avg:141.71ms step:2625/4578 val_loss:3.4501 train_time:370617ms step_avg:141.73ms step:2626/4578 train_loss:3.4404 train_time:370717ms step_avg:141.71ms step:2627/4578 train_loss:3.5075 train_time:370861ms step_avg:141.71ms step:2628/4578 train_loss:3.3344 train_time:371000ms step_avg:141.71ms step:2629/4578 train_loss:3.5876 train_time:371138ms step_avg:141.71ms step:2630/4578 train_loss:3.4701 train_time:371281ms step_avg:141.71ms step:2631/4578 train_loss:3.5286 train_time:371414ms step_avg:141.71ms step:2632/4578 train_loss:3.7537 train_time:371554ms step_avg:141.71ms step:2633/4578 train_loss:3.4957 train_time:371696ms step_avg:141.71ms step:2634/4578 train_loss:3.4140 train_time:371840ms step_avg:141.71ms step:2635/4578 train_loss:3.3899 train_time:371981ms step_avg:141.71ms step:2636/4578 train_loss:3.4320 train_time:372121ms step_avg:141.71ms step:2637/4578 train_loss:3.2225 train_time:372260ms step_avg:141.71ms step:2638/4578 train_loss:3.5239 train_time:372403ms step_avg:141.71ms step:2639/4578 train_loss:3.4984 train_time:372539ms step_avg:141.70ms step:2640/4578 train_loss:3.3996 train_time:372679ms step_avg:141.70ms step:2641/4578 train_loss:3.4766 train_time:372827ms step_avg:141.71ms step:2642/4578 train_loss:3.5075 train_time:372966ms step_avg:141.70ms step:2643/4578 train_loss:3.3022 train_time:373103ms step_avg:141.70ms step:2644/4578 train_loss:3.4221 train_time:373241ms step_avg:141.70ms step:2645/4578 train_loss:3.4914 train_time:373382ms step_avg:141.70ms step:2646/4578 train_loss:3.4616 train_time:373662ms step_avg:141.75ms step:2647/4578 train_loss:3.3516 train_time:373799ms step_avg:141.75ms step:2648/4578 train_loss:3.5757 train_time:373938ms step_avg:141.75ms step:2649/4578 train_loss:3.8298 train_time:374076ms step_avg:141.75ms step:2650/4578 train_loss:3.4683 train_time:374213ms step_avg:141.75ms step:2651/4578 train_loss:3.4337 train_time:374352ms step_avg:141.75ms step:2652/4578 train_loss:3.5659 train_time:374490ms step_avg:141.75ms step:2653/4578 train_loss:3.4056 train_time:374637ms step_avg:141.75ms step:2654/4578 train_loss:3.3889 train_time:374780ms step_avg:141.75ms step:2655/4578 train_loss:3.4595 train_time:374920ms step_avg:141.75ms step:2656/4578 train_loss:3.3810 train_time:375059ms step_avg:141.75ms step:2657/4578 train_loss:3.4142 train_time:375197ms step_avg:141.74ms step:2658/4578 train_loss:3.3874 train_time:375337ms step_avg:141.74ms step:2659/4578 train_loss:3.4686 train_time:375477ms step_avg:141.74ms step:2660/4578 train_loss:3.6106 train_time:375813ms step_avg:141.82ms step:2661/4578 train_loss:3.4079 train_time:375953ms step_avg:141.82ms step:2662/4578 train_loss:3.5599 train_time:376091ms step_avg:141.81ms step:2663/4578 train_loss:3.4231 train_time:376231ms step_avg:141.81ms step:2664/4578 train_loss:3.4211 train_time:376368ms step_avg:141.81ms step:2665/4578 train_loss:3.3438 train_time:376510ms step_avg:141.81ms step:2666/4578 train_loss:3.4008 train_time:376647ms step_avg:141.81ms step:2667/4578 train_loss:3.4336 train_time:376794ms step_avg:141.81ms step:2668/4578 train_loss:3.4809 train_time:376937ms step_avg:141.81ms step:2669/4578 train_loss:3.3944 train_time:377077ms step_avg:141.81ms step:2670/4578 train_loss:3.4559 train_time:377216ms step_avg:141.81ms step:2671/4578 train_loss:3.3358 train_time:377355ms step_avg:141.81ms step:2672/4578 train_loss:3.4175 train_time:377497ms step_avg:141.81ms step:2673/4578 train_loss:3.3952 train_time:377634ms step_avg:141.81ms step:2674/4578 train_loss:3.4521 train_time:377779ms step_avg:141.81ms step:2675/4578 train_loss:3.4755 train_time:377921ms step_avg:141.81ms step:2676/4578 train_loss:3.4503 train_time:378060ms step_avg:141.81ms step:2677/4578 train_loss:3.4372 train_time:378199ms step_avg:141.81ms step:2678/4578 train_loss:3.4743 train_time:378347ms step_avg:141.81ms step:2679/4578 train_loss:3.5146 train_time:378478ms step_avg:141.81ms step:2680/4578 train_loss:3.4280 train_time:378626ms step_avg:141.81ms step:2681/4578 train_loss:3.3495 train_time:378759ms step_avg:141.80ms step:2682/4578 train_loss:3.3908 train_time:378900ms step_avg:141.80ms step:2683/4578 train_loss:3.8668 train_time:379047ms step_avg:141.81ms step:2684/4578 train_loss:3.4513 train_time:379181ms step_avg:141.80ms step:2685/4578 train_loss:3.4786 train_time:379322ms step_avg:141.80ms step:2686/4578 train_loss:3.5280 train_time:379461ms step_avg:141.80ms step:2687/4578 train_loss:3.4457 train_time:379599ms step_avg:141.80ms step:2688/4578 train_loss:3.5286 train_time:379743ms step_avg:141.80ms step:2689/4578 train_loss:3.4558 train_time:379880ms step_avg:141.80ms step:2690/4578 train_loss:3.4479 train_time:380019ms step_avg:141.80ms step:2691/4578 train_loss:3.4775 train_time:380165ms step_avg:141.80ms step:2692/4578 train_loss:3.5428 train_time:380299ms step_avg:141.80ms step:2693/4578 train_loss:3.3441 train_time:380443ms step_avg:141.80ms step:2694/4578 train_loss:3.7187 train_time:380584ms step_avg:141.80ms step:2695/4578 train_loss:3.5246 train_time:380717ms step_avg:141.79ms step:2696/4578 train_loss:3.3260 train_time:380858ms step_avg:141.79ms step:2697/4578 train_loss:3.5067 train_time:380997ms step_avg:141.79ms step:2698/4578 train_loss:3.4676 train_time:381138ms step_avg:141.79ms step:2699/4578 train_loss:3.4210 train_time:381278ms step_avg:141.79ms step:2700/4578 train_loss:3.5241 train_time:381425ms step_avg:141.79ms step:2701/4578 train_loss:3.4952 train_time:381564ms step_avg:141.79ms step:2702/4578 train_loss:3.3996 train_time:381697ms step_avg:141.79ms step:2703/4578 train_loss:3.4231 train_time:381838ms step_avg:141.79ms step:2704/4578 train_loss:3.4431 train_time:381980ms step_avg:141.79ms step:2705/4578 train_loss:3.4010 train_time:382122ms step_avg:141.79ms step:2706/4578 train_loss:3.5792 train_time:382258ms step_avg:141.79ms step:2707/4578 train_loss:3.5414 train_time:382406ms step_avg:141.79ms step:2708/4578 train_loss:3.4490 train_time:382537ms step_avg:141.79ms step:2709/4578 train_loss:3.4442 train_time:382677ms step_avg:141.78ms step:2710/4578 train_loss:3.5459 train_time:382817ms step_avg:141.78ms step:2711/4578 train_loss:3.4202 train_time:382958ms step_avg:141.78ms step:2712/4578 train_loss:3.5320 train_time:383102ms step_avg:141.78ms step:2713/4578 train_loss:3.2755 train_time:383238ms step_avg:141.78ms step:2714/4578 train_loss:3.4721 train_time:383377ms step_avg:141.78ms step:2715/4578 train_loss:3.3646 train_time:383520ms step_avg:141.78ms step:2716/4578 train_loss:3.3688 train_time:383670ms step_avg:141.78ms step:2717/4578 train_loss:3.5595 train_time:383802ms step_avg:141.78ms step:2718/4578 train_loss:3.4588 train_time:383940ms step_avg:141.78ms step:2719/4578 train_loss:3.6962 train_time:384077ms step_avg:141.78ms step:2720/4578 train_loss:3.4440 train_time:384217ms step_avg:141.78ms step:2721/4578 train_loss:3.4288 train_time:384360ms step_avg:141.78ms step:2722/4578 train_loss:3.6608 train_time:384504ms step_avg:141.78ms step:2723/4578 train_loss:3.4248 train_time:384644ms step_avg:141.78ms step:2724/4578 train_loss:3.6013 train_time:384781ms step_avg:141.78ms step:2725/4578 train_loss:3.4779 train_time:384917ms step_avg:141.77ms step:2726/4578 train_loss:3.4382 train_time:385063ms step_avg:141.78ms step:2727/4578 train_loss:3.4468 train_time:385197ms step_avg:141.77ms step:2728/4578 train_loss:3.7866 train_time:385337ms step_avg:141.77ms step:2729/4578 train_loss:3.5158 train_time:385478ms step_avg:141.77ms step:2730/4578 train_loss:3.3828 train_time:385615ms step_avg:141.77ms step:2731/4578 train_loss:3.4922 train_time:385761ms step_avg:141.77ms step:2732/4578 train_loss:3.3931 train_time:385897ms step_avg:141.77ms step:2733/4578 train_loss:3.2849 train_time:386037ms step_avg:141.77ms step:2734/4578 train_loss:3.3916 train_time:386178ms step_avg:141.77ms step:2735/4578 train_loss:3.4747 train_time:386319ms step_avg:141.77ms step:2736/4578 train_loss:3.3678 train_time:386461ms step_avg:141.77ms step:2737/4578 train_loss:3.7683 train_time:386598ms step_avg:141.77ms step:2738/4578 train_loss:3.5060 train_time:386740ms step_avg:141.77ms step:2739/4578 train_loss:3.7072 train_time:386879ms step_avg:141.77ms step:2740/4578 train_loss:3.4532 train_time:387021ms step_avg:141.77ms step:2741/4578 train_loss:3.4515 train_time:387161ms step_avg:141.77ms step:2742/4578 train_loss:3.3898 train_time:387298ms step_avg:141.76ms step:2743/4578 train_loss:3.4596 train_time:387439ms step_avg:141.76ms step:2744/4578 train_loss:3.4733 train_time:387578ms step_avg:141.76ms step:2745/4578 train_loss:3.5697 train_time:387731ms step_avg:141.77ms step:2746/4578 train_loss:3.3440 train_time:387863ms step_avg:141.76ms step:2747/4578 train_loss:3.4372 train_time:387999ms step_avg:141.76ms step:2748/4578 train_loss:3.4770 train_time:388138ms step_avg:141.76ms step:2749/4578 train_loss:3.5859 train_time:388278ms step_avg:141.76ms step:2750/4578 train_loss:3.4299 train_time:388418ms step_avg:141.76ms step:2750/4578 val_loss:3.4432 train_time:388473ms step_avg:141.78ms step:2751/4578 train_loss:3.4961 train_time:388568ms step_avg:141.76ms step:2752/4578 train_loss:3.5553 train_time:388718ms step_avg:141.76ms step:2753/4578 train_loss:3.4676 train_time:388855ms step_avg:141.76ms step:2754/4578 train_loss:3.3910 train_time:388994ms step_avg:141.76ms step:2755/4578 train_loss:3.3902 train_time:389136ms step_avg:141.76ms step:2756/4578 train_loss:3.4778 train_time:389271ms step_avg:141.76ms step:2757/4578 train_loss:3.4165 train_time:389410ms step_avg:141.76ms step:2758/4578 train_loss:3.2920 train_time:389552ms step_avg:141.76ms step:2759/4578 train_loss:3.6839 train_time:389702ms step_avg:141.76ms step:2760/4578 train_loss:3.4959 train_time:389843ms step_avg:141.76ms step:2761/4578 train_loss:3.4565 train_time:389988ms step_avg:141.76ms step:2762/4578 train_loss:3.4282 train_time:390118ms step_avg:141.76ms step:2763/4578 train_loss:3.3430 train_time:390255ms step_avg:141.76ms step:2764/4578 train_loss:3.5175 train_time:390395ms step_avg:141.76ms step:2765/4578 train_loss:3.4308 train_time:390537ms step_avg:141.76ms step:2766/4578 train_loss:3.3293 train_time:390678ms step_avg:141.76ms step:2767/4578 train_loss:3.4152 train_time:390828ms step_avg:141.76ms step:2768/4578 train_loss:3.5019 train_time:390958ms step_avg:141.75ms step:2769/4578 train_loss:3.3792 train_time:391101ms step_avg:141.75ms step:2770/4578 train_loss:3.4562 train_time:391239ms step_avg:141.75ms step:2771/4578 train_loss:3.4333 train_time:391380ms step_avg:141.75ms step:2772/4578 train_loss:3.8747 train_time:391518ms step_avg:141.75ms step:2773/4578 train_loss:3.3415 train_time:391660ms step_avg:141.75ms step:2774/4578 train_loss:3.4735 train_time:391799ms step_avg:141.75ms step:2775/4578 train_loss:3.5334 train_time:391940ms step_avg:141.75ms step:2776/4578 train_loss:3.5042 train_time:392080ms step_avg:141.75ms step:2777/4578 train_loss:3.5804 train_time:392220ms step_avg:141.75ms step:2778/4578 train_loss:3.5820 train_time:392370ms step_avg:141.75ms step:2779/4578 train_loss:3.4472 train_time:392499ms step_avg:141.75ms step:2780/4578 train_loss:3.3239 train_time:392639ms step_avg:141.75ms step:2781/4578 train_loss:3.4682 train_time:392785ms step_avg:141.75ms step:2782/4578 train_loss:3.4859 train_time:392921ms step_avg:141.75ms step:2783/4578 train_loss:3.3499 train_time:393062ms step_avg:141.75ms step:2784/4578 train_loss:3.4595 train_time:393201ms step_avg:141.74ms step:2785/4578 train_loss:3.5169 train_time:393345ms step_avg:141.75ms step:2786/4578 train_loss:3.3860 train_time:393482ms step_avg:141.74ms step:2787/4578 train_loss:3.5154 train_time:393623ms step_avg:141.74ms step:2788/4578 train_loss:3.4745 train_time:393764ms step_avg:141.74ms step:2789/4578 train_loss:3.4139 train_time:393905ms step_avg:141.74ms step:2790/4578 train_loss:3.4970 train_time:394046ms step_avg:141.74ms step:2791/4578 train_loss:3.4169 train_time:394188ms step_avg:141.74ms step:2792/4578 train_loss:3.3238 train_time:394327ms step_avg:141.74ms step:2793/4578 train_loss:3.4246 train_time:394470ms step_avg:141.74ms step:2794/4578 train_loss:3.4636 train_time:394614ms step_avg:141.74ms step:2795/4578 train_loss:3.3781 train_time:394750ms step_avg:141.74ms step:2796/4578 train_loss:3.4218 train_time:394891ms step_avg:141.74ms step:2797/4578 train_loss:3.3467 train_time:395034ms step_avg:141.74ms step:2798/4578 train_loss:3.4461 train_time:395173ms step_avg:141.74ms step:2799/4578 train_loss:3.4031 train_time:395312ms step_avg:141.74ms step:2800/4578 train_loss:3.5677 train_time:395452ms step_avg:141.74ms step:2801/4578 train_loss:3.5209 train_time:395592ms step_avg:141.74ms step:2802/4578 train_loss:3.4859 train_time:395739ms step_avg:141.74ms step:2803/4578 train_loss:3.4348 train_time:395872ms step_avg:141.74ms step:2804/4578 train_loss:3.6202 train_time:396012ms step_avg:141.74ms step:2805/4578 train_loss:3.5775 train_time:396153ms step_avg:141.74ms step:2806/4578 train_loss:3.3050 train_time:396293ms step_avg:141.74ms step:2807/4578 train_loss:3.7102 train_time:396440ms step_avg:141.74ms step:2808/4578 train_loss:3.4548 train_time:396578ms step_avg:141.74ms step:2809/4578 train_loss:3.3779 train_time:396726ms step_avg:141.74ms step:2810/4578 train_loss:3.4087 train_time:396852ms step_avg:141.73ms step:2811/4578 train_loss:3.5609 train_time:396992ms step_avg:141.73ms step:2812/4578 train_loss:3.5478 train_time:397131ms step_avg:141.73ms step:2813/4578 train_loss:3.2968 train_time:397275ms step_avg:141.73ms step:2814/4578 train_loss:3.5244 train_time:397415ms step_avg:141.73ms step:2815/4578 train_loss:3.5967 train_time:397559ms step_avg:141.73ms step:2816/4578 train_loss:3.4039 train_time:397692ms step_avg:141.73ms step:2817/4578 train_loss:2.9860 train_time:397832ms step_avg:141.73ms step:2818/4578 train_loss:3.4283 train_time:397973ms step_avg:141.73ms step:2819/4578 train_loss:3.3978 train_time:398112ms step_avg:141.73ms step:2820/4578 train_loss:3.5874 train_time:398261ms step_avg:141.73ms step:2821/4578 train_loss:3.4829 train_time:398395ms step_avg:141.73ms step:2822/4578 train_loss:3.5175 train_time:398533ms step_avg:141.73ms step:2823/4578 train_loss:3.4508 train_time:398678ms step_avg:141.73ms step:2824/4578 train_loss:3.4130 train_time:398817ms step_avg:141.73ms step:2825/4578 train_loss:3.3098 train_time:398963ms step_avg:141.73ms step:2826/4578 train_loss:3.5752 train_time:399096ms step_avg:141.72ms step:2827/4578 train_loss:3.4702 train_time:399233ms step_avg:141.72ms step:2828/4578 train_loss:3.3541 train_time:399380ms step_avg:141.72ms step:2829/4578 train_loss:3.4834 train_time:399513ms step_avg:141.72ms step:2830/4578 train_loss:3.4848 train_time:399653ms step_avg:141.72ms step:2831/4578 train_loss:3.4173 train_time:399793ms step_avg:141.72ms step:2832/4578 train_loss:3.5637 train_time:399935ms step_avg:141.72ms step:2833/4578 train_loss:3.4816 train_time:400073ms step_avg:141.72ms step:2834/4578 train_loss:3.4694 train_time:400218ms step_avg:141.72ms step:2835/4578 train_loss:3.2825 train_time:400496ms step_avg:141.77ms step:2836/4578 train_loss:3.5020 train_time:400632ms step_avg:141.77ms step:2837/4578 train_loss:3.4341 train_time:400771ms step_avg:141.77ms step:2838/4578 train_loss:3.7456 train_time:400909ms step_avg:141.76ms step:2839/4578 train_loss:3.3881 train_time:401047ms step_avg:141.76ms step:2840/4578 train_loss:3.3948 train_time:401187ms step_avg:141.76ms step:2841/4578 train_loss:3.4605 train_time:401325ms step_avg:141.76ms step:2842/4578 train_loss:3.3811 train_time:401470ms step_avg:141.76ms step:2843/4578 train_loss:3.3827 train_time:401613ms step_avg:141.76ms step:2844/4578 train_loss:3.5585 train_time:401754ms step_avg:141.76ms step:2845/4578 train_loss:3.4508 train_time:401892ms step_avg:141.76ms step:2846/4578 train_loss:3.4695 train_time:402031ms step_avg:141.76ms step:2847/4578 train_loss:3.4304 train_time:402172ms step_avg:141.76ms step:2848/4578 train_loss:3.7009 train_time:402311ms step_avg:141.76ms step:2849/4578 train_loss:3.3678 train_time:402452ms step_avg:141.76ms step:2850/4578 train_loss:3.3931 train_time:402761ms step_avg:141.82ms step:2851/4578 train_loss:3.4957 train_time:402897ms step_avg:141.82ms step:2852/4578 train_loss:3.4670 train_time:403034ms step_avg:141.81ms step:2853/4578 train_loss:3.4302 train_time:403172ms step_avg:141.81ms step:2854/4578 train_loss:3.5095 train_time:403311ms step_avg:141.81ms step:2855/4578 train_loss:3.3273 train_time:403450ms step_avg:141.81ms step:2856/4578 train_loss:3.3435 train_time:403589ms step_avg:141.81ms step:2857/4578 train_loss:3.4473 train_time:403740ms step_avg:141.81ms step:2858/4578 train_loss:3.4466 train_time:403876ms step_avg:141.81ms step:2859/4578 train_loss:3.3273 train_time:404016ms step_avg:141.81ms step:2860/4578 train_loss:3.4315 train_time:404168ms step_avg:141.81ms step:2861/4578 train_loss:3.3915 train_time:404295ms step_avg:141.81ms step:2862/4578 train_loss:3.4358 train_time:404434ms step_avg:141.81ms step:2863/4578 train_loss:3.4755 train_time:404574ms step_avg:141.81ms step:2864/4578 train_loss:3.7402 train_time:404716ms step_avg:141.81ms step:2865/4578 train_loss:3.5570 train_time:404857ms step_avg:141.81ms step:2866/4578 train_loss:3.4423 train_time:404999ms step_avg:141.81ms step:2867/4578 train_loss:3.3296 train_time:405139ms step_avg:141.81ms step:2868/4578 train_loss:3.5269 train_time:405279ms step_avg:141.80ms step:2869/4578 train_loss:3.4753 train_time:405418ms step_avg:141.80ms step:2870/4578 train_loss:3.4387 train_time:405557ms step_avg:141.80ms step:2871/4578 train_loss:3.5772 train_time:405698ms step_avg:141.80ms step:2872/4578 train_loss:3.3537 train_time:405838ms step_avg:141.80ms step:2873/4578 train_loss:3.4105 train_time:405985ms step_avg:141.80ms step:2874/4578 train_loss:3.2787 train_time:406119ms step_avg:141.80ms step:2875/4578 train_loss:3.4392 train_time:406269ms step_avg:141.80ms step:2875/4578 val_loss:3.4372 train_time:406315ms step_avg:141.82ms step:2876/4578 train_loss:3.3576 train_time:406410ms step_avg:141.80ms step:2877/4578 train_loss:3.3384 train_time:406556ms step_avg:141.81ms step:2878/4578 train_loss:3.4289 train_time:406696ms step_avg:141.80ms step:2879/4578 train_loss:3.5399 train_time:406835ms step_avg:141.80ms step:2880/4578 train_loss:3.4969 train_time:406981ms step_avg:141.81ms step:2881/4578 train_loss:3.4350 train_time:407112ms step_avg:141.80ms step:2882/4578 train_loss:3.4260 train_time:407255ms step_avg:141.80ms step:2883/4578 train_loss:3.5467 train_time:407404ms step_avg:141.80ms step:2884/4578 train_loss:3.3362 train_time:407537ms step_avg:141.80ms step:2885/4578 train_loss:3.3553 train_time:407678ms step_avg:141.80ms step:2886/4578 train_loss:3.3999 train_time:407826ms step_avg:141.80ms step:2887/4578 train_loss:3.3965 train_time:407956ms step_avg:141.80ms step:2888/4578 train_loss:3.4096 train_time:408097ms step_avg:141.80ms step:2889/4578 train_loss:3.4280 train_time:408236ms step_avg:141.80ms step:2890/4578 train_loss:3.6112 train_time:408382ms step_avg:141.80ms step:2891/4578 train_loss:3.4621 train_time:408519ms step_avg:141.80ms step:2892/4578 train_loss:3.2973 train_time:408660ms step_avg:141.80ms step:2893/4578 train_loss:3.2353 train_time:408801ms step_avg:141.80ms step:2894/4578 train_loss:3.3680 train_time:408939ms step_avg:141.80ms step:2895/4578 train_loss:3.2497 train_time:409082ms step_avg:141.80ms step:2896/4578 train_loss:3.4290 train_time:409231ms step_avg:141.80ms step:2897/4578 train_loss:3.5558 train_time:409359ms step_avg:141.79ms step:2898/4578 train_loss:3.3788 train_time:409500ms step_avg:141.79ms step:2899/4578 train_loss:3.4735 train_time:409652ms step_avg:141.80ms step:2900/4578 train_loss:3.3579 train_time:409790ms step_avg:141.80ms step:2901/4578 train_loss:3.5489 train_time:409924ms step_avg:141.79ms step:2902/4578 train_loss:3.5348 train_time:410063ms step_avg:141.79ms step:2903/4578 train_loss:3.5797 train_time:410200ms step_avg:141.79ms step:2904/4578 train_loss:3.2878 train_time:410338ms step_avg:141.79ms step:2905/4578 train_loss:3.4307 train_time:410479ms step_avg:141.79ms step:2906/4578 train_loss:3.4074 train_time:410619ms step_avg:141.79ms step:2907/4578 train_loss:3.4903 train_time:410763ms step_avg:141.79ms step:2908/4578 train_loss:3.4272 train_time:410916ms step_avg:141.79ms step:2909/4578 train_loss:3.3886 train_time:411042ms step_avg:141.79ms step:2910/4578 train_loss:3.7261 train_time:411181ms step_avg:141.79ms step:2911/4578 train_loss:3.4398 train_time:411332ms step_avg:141.79ms step:2912/4578 train_loss:3.3414 train_time:411460ms step_avg:141.78ms step:2913/4578 train_loss:3.3289 train_time:411600ms step_avg:141.78ms step:2914/4578 train_loss:3.8198 train_time:411754ms step_avg:141.79ms step:2915/4578 train_loss:3.4094 train_time:411881ms step_avg:141.78ms step:2916/4578 train_loss:3.3524 train_time:412021ms step_avg:141.78ms step:2917/4578 train_loss:3.3411 train_time:412159ms step_avg:141.78ms step:2918/4578 train_loss:3.6228 train_time:412298ms step_avg:141.78ms step:2919/4578 train_loss:3.1365 train_time:412438ms step_avg:141.78ms step:2920/4578 train_loss:3.3272 train_time:412578ms step_avg:141.78ms step:2921/4578 train_loss:3.3533 train_time:412718ms step_avg:141.78ms step:2922/4578 train_loss:3.4423 train_time:412859ms step_avg:141.78ms step:2923/4578 train_loss:3.4838 train_time:412999ms step_avg:141.78ms step:2924/4578 train_loss:3.5219 train_time:413137ms step_avg:141.78ms step:2925/4578 train_loss:3.5253 train_time:413280ms step_avg:141.78ms step:2926/4578 train_loss:3.4038 train_time:413417ms step_avg:141.78ms step:2927/4578 train_loss:3.4109 train_time:413558ms step_avg:141.78ms step:2928/4578 train_loss:3.4077 train_time:413698ms step_avg:141.77ms step:2929/4578 train_loss:3.4049 train_time:413839ms step_avg:141.77ms step:2930/4578 train_loss:3.3755 train_time:413980ms step_avg:141.77ms step:2931/4578 train_loss:3.3989 train_time:414119ms step_avg:141.77ms step:2932/4578 train_loss:3.5319 train_time:414260ms step_avg:141.77ms step:2933/4578 train_loss:3.5707 train_time:414399ms step_avg:141.77ms step:2934/4578 train_loss:3.5455 train_time:414539ms step_avg:141.77ms step:2935/4578 train_loss:3.3854 train_time:414678ms step_avg:141.77ms step:2936/4578 train_loss:3.4407 train_time:414817ms step_avg:141.77ms step:2937/4578 train_loss:3.3737 train_time:414959ms step_avg:141.77ms step:2938/4578 train_loss:3.4037 train_time:415099ms step_avg:141.77ms step:2939/4578 train_loss:3.4372 train_time:415239ms step_avg:141.77ms step:2940/4578 train_loss:3.4682 train_time:415379ms step_avg:141.77ms step:2941/4578 train_loss:3.5211 train_time:415520ms step_avg:141.77ms step:2942/4578 train_loss:3.5024 train_time:415666ms step_avg:141.77ms step:2943/4578 train_loss:3.4341 train_time:415812ms step_avg:141.77ms step:2944/4578 train_loss:3.3168 train_time:415938ms step_avg:141.76ms step:2945/4578 train_loss:3.2508 train_time:416083ms step_avg:141.77ms step:2946/4578 train_loss:3.4545 train_time:416226ms step_avg:141.77ms step:2947/4578 train_loss:3.5235 train_time:416357ms step_avg:141.76ms step:2948/4578 train_loss:3.4534 train_time:416500ms step_avg:141.76ms step:2949/4578 train_loss:3.6339 train_time:416636ms step_avg:141.76ms step:2950/4578 train_loss:3.4565 train_time:416777ms step_avg:141.76ms step:2951/4578 train_loss:3.4611 train_time:416917ms step_avg:141.76ms step:2952/4578 train_loss:3.8747 train_time:417058ms step_avg:141.76ms step:2953/4578 train_loss:3.5440 train_time:417199ms step_avg:141.76ms step:2954/4578 train_loss:3.4825 train_time:417342ms step_avg:141.76ms step:2955/4578 train_loss:3.4933 train_time:417480ms step_avg:141.76ms step:2956/4578 train_loss:3.4253 train_time:417627ms step_avg:141.76ms step:2957/4578 train_loss:3.4522 train_time:417759ms step_avg:141.76ms step:2958/4578 train_loss:3.3220 train_time:417899ms step_avg:141.76ms step:2959/4578 train_loss:3.4047 train_time:418039ms step_avg:141.76ms step:2960/4578 train_loss:3.5448 train_time:418179ms step_avg:141.76ms step:2961/4578 train_loss:3.3646 train_time:418319ms step_avg:141.76ms step:2962/4578 train_loss:3.4889 train_time:418461ms step_avg:141.76ms step:2963/4578 train_loss:3.3447 train_time:418601ms step_avg:141.75ms step:2964/4578 train_loss:3.4024 train_time:418740ms step_avg:141.75ms step:2965/4578 train_loss:3.3878 train_time:418889ms step_avg:141.76ms step:2966/4578 train_loss:3.5045 train_time:419023ms step_avg:141.75ms step:2967/4578 train_loss:3.3725 train_time:419160ms step_avg:141.75ms step:2968/4578 train_loss:3.6168 train_time:419303ms step_avg:141.75ms step:2969/4578 train_loss:3.4729 train_time:419440ms step_avg:141.75ms step:2970/4578 train_loss:3.4794 train_time:419580ms step_avg:141.75ms step:2971/4578 train_loss:3.4670 train_time:419721ms step_avg:141.75ms step:2972/4578 train_loss:3.5431 train_time:419861ms step_avg:141.75ms step:2973/4578 train_loss:3.3714 train_time:420004ms step_avg:141.75ms step:2974/4578 train_loss:3.3879 train_time:420141ms step_avg:141.75ms step:2975/4578 train_loss:3.2971 train_time:420281ms step_avg:141.75ms step:2976/4578 train_loss:3.3763 train_time:420421ms step_avg:141.75ms step:2977/4578 train_loss:3.3653 train_time:420560ms step_avg:141.75ms step:2978/4578 train_loss:3.3819 train_time:420700ms step_avg:141.75ms step:2979/4578 train_loss:3.6627 train_time:420840ms step_avg:141.74ms step:2980/4578 train_loss:3.4682 train_time:420981ms step_avg:141.74ms step:2981/4578 train_loss:3.5128 train_time:421122ms step_avg:141.74ms step:2982/4578 train_loss:3.5217 train_time:421260ms step_avg:141.74ms step:2983/4578 train_loss:3.6078 train_time:421408ms step_avg:141.75ms step:2984/4578 train_loss:3.4095 train_time:421540ms step_avg:141.74ms step:2985/4578 train_loss:3.5017 train_time:421681ms step_avg:141.74ms step:2986/4578 train_loss:3.5036 train_time:421820ms step_avg:141.74ms step:2987/4578 train_loss:3.4560 train_time:421960ms step_avg:141.74ms step:2988/4578 train_loss:3.5825 train_time:422100ms step_avg:141.74ms step:2989/4578 train_loss:3.1805 train_time:422240ms step_avg:141.74ms step:2990/4578 train_loss:3.5200 train_time:422383ms step_avg:141.74ms step:2991/4578 train_loss:3.4755 train_time:422520ms step_avg:141.74ms step:2992/4578 train_loss:3.4479 train_time:422661ms step_avg:141.74ms step:2993/4578 train_loss:3.3720 train_time:422809ms step_avg:141.74ms step:2994/4578 train_loss:3.5074 train_time:422941ms step_avg:141.74ms step:2995/4578 train_loss:3.3251 train_time:423086ms step_avg:141.74ms step:2996/4578 train_loss:3.3599 train_time:423220ms step_avg:141.73ms step:2997/4578 train_loss:3.4238 train_time:423360ms step_avg:141.73ms step:2998/4578 train_loss:3.3665 train_time:423500ms step_avg:141.73ms step:2999/4578 train_loss:3.4872 train_time:423647ms step_avg:141.74ms step:3000/4578 train_loss:3.3909 train_time:423780ms step_avg:141.73ms step:3000/4578 val_loss:3.4295 train_time:423835ms step_avg:141.75ms step:3001/4578 train_loss:3.3845 train_time:423931ms step_avg:141.74ms step:3002/4578 train_loss:3.3268 train_time:424077ms step_avg:141.74ms step:3003/4578 train_loss:3.3728 train_time:424218ms step_avg:141.74ms step:3004/4578 train_loss:3.5011 train_time:424356ms step_avg:141.74ms step:3005/4578 train_loss:3.8395 train_time:424495ms step_avg:141.73ms step:3006/4578 train_loss:3.4152 train_time:424638ms step_avg:141.73ms step:3007/4578 train_loss:3.4833 train_time:424773ms step_avg:141.73ms step:3008/4578 train_loss:3.2941 train_time:424913ms step_avg:141.73ms step:3009/4578 train_loss:3.5030 train_time:425059ms step_avg:141.73ms step:3010/4578 train_loss:3.3989 train_time:425198ms step_avg:141.73ms step:3011/4578 train_loss:3.4660 train_time:425337ms step_avg:141.73ms step:3012/4578 train_loss:3.4668 train_time:425481ms step_avg:141.73ms step:3013/4578 train_loss:3.3503 train_time:425615ms step_avg:141.73ms step:3014/4578 train_loss:3.5473 train_time:425755ms step_avg:141.73ms step:3015/4578 train_loss:3.5135 train_time:425895ms step_avg:141.73ms step:3016/4578 train_loss:3.3748 train_time:426035ms step_avg:141.73ms step:3017/4578 train_loss:3.4123 train_time:426176ms step_avg:141.73ms step:3018/4578 train_loss:3.4598 train_time:426317ms step_avg:141.73ms step:3019/4578 train_loss:3.4977 train_time:426456ms step_avg:141.73ms step:3020/4578 train_loss:3.2864 train_time:426596ms step_avg:141.73ms step:3021/4578 train_loss:3.5737 train_time:426736ms step_avg:141.73ms step:3022/4578 train_loss:3.4104 train_time:426875ms step_avg:141.72ms step:3023/4578 train_loss:3.3294 train_time:427016ms step_avg:141.72ms step:3024/4578 train_loss:3.4255 train_time:427292ms step_avg:141.77ms step:3025/4578 train_loss:3.4114 train_time:427427ms step_avg:141.77ms step:3026/4578 train_loss:3.4668 train_time:427566ms step_avg:141.77ms step:3027/4578 train_loss:3.4866 train_time:427705ms step_avg:141.77ms step:3028/4578 train_loss:3.3939 train_time:427848ms step_avg:141.77ms step:3029/4578 train_loss:3.1954 train_time:427982ms step_avg:141.76ms step:3030/4578 train_loss:3.5469 train_time:428122ms step_avg:141.76ms step:3031/4578 train_loss:3.3020 train_time:428268ms step_avg:141.76ms step:3032/4578 train_loss:3.2928 train_time:428412ms step_avg:141.76ms step:3033/4578 train_loss:3.6341 train_time:428551ms step_avg:141.76ms step:3034/4578 train_loss:3.6267 train_time:428690ms step_avg:141.76ms step:3035/4578 train_loss:3.3933 train_time:428829ms step_avg:141.76ms step:3036/4578 train_loss:3.4729 train_time:428969ms step_avg:141.76ms step:3037/4578 train_loss:3.4242 train_time:429109ms step_avg:141.76ms step:3038/4578 train_loss:3.3225 train_time:429250ms step_avg:141.76ms step:3039/4578 train_loss:3.3812 train_time:429394ms step_avg:141.76ms step:3040/4578 train_loss:3.4736 train_time:429727ms step_avg:141.82ms step:3041/4578 train_loss:3.4692 train_time:429862ms step_avg:141.82ms step:3042/4578 train_loss:3.2732 train_time:430001ms step_avg:141.82ms step:3043/4578 train_loss:3.4164 train_time:430141ms step_avg:141.82ms step:3044/4578 train_loss:3.4510 train_time:430281ms step_avg:141.82ms step:3045/4578 train_loss:3.4562 train_time:430420ms step_avg:141.82ms step:3046/4578 train_loss:3.5358 train_time:430561ms step_avg:141.82ms step:3047/4578 train_loss:3.3526 train_time:430708ms step_avg:141.82ms step:3048/4578 train_loss:3.4720 train_time:430848ms step_avg:141.82ms step:3049/4578 train_loss:3.4220 train_time:430988ms step_avg:141.82ms step:3050/4578 train_loss:3.3547 train_time:431127ms step_avg:141.82ms step:3051/4578 train_loss:3.4830 train_time:431266ms step_avg:141.82ms step:3052/4578 train_loss:3.3235 train_time:431416ms step_avg:141.82ms step:3053/4578 train_loss:3.5674 train_time:431546ms step_avg:141.82ms step:3054/4578 train_loss:3.5111 train_time:431690ms step_avg:141.82ms step:3055/4578 train_loss:3.4884 train_time:431830ms step_avg:141.82ms step:3056/4578 train_loss:3.4890 train_time:431970ms step_avg:141.82ms step:3057/4578 train_loss:3.3703 train_time:432112ms step_avg:141.82ms step:3058/4578 train_loss:3.3972 train_time:432255ms step_avg:141.82ms step:3059/4578 train_loss:3.4813 train_time:432389ms step_avg:141.81ms step:3060/4578 train_loss:3.3827 train_time:432528ms step_avg:141.81ms step:3061/4578 train_loss:3.4341 train_time:432670ms step_avg:141.81ms step:3062/4578 train_loss:3.4379 train_time:432812ms step_avg:141.81ms step:3063/4578 train_loss:3.3807 train_time:432952ms step_avg:141.81ms step:3064/4578 train_loss:3.3517 train_time:433092ms step_avg:141.81ms step:3065/4578 train_loss:3.3715 train_time:433231ms step_avg:141.81ms step:3066/4578 train_loss:3.3482 train_time:433371ms step_avg:141.81ms step:3067/4578 train_loss:3.3418 train_time:433510ms step_avg:141.81ms step:3068/4578 train_loss:3.3019 train_time:433650ms step_avg:141.81ms step:3069/4578 train_loss:3.3427 train_time:433791ms step_avg:141.81ms step:3070/4578 train_loss:3.3319 train_time:433932ms step_avg:141.81ms step:3071/4578 train_loss:3.5118 train_time:434072ms step_avg:141.81ms step:3072/4578 train_loss:3.4466 train_time:434212ms step_avg:141.81ms step:3073/4578 train_loss:3.4911 train_time:434354ms step_avg:141.81ms step:3074/4578 train_loss:3.4741 train_time:434492ms step_avg:141.81ms step:3075/4578 train_loss:3.4229 train_time:434632ms step_avg:141.80ms step:3076/4578 train_loss:3.4764 train_time:434773ms step_avg:141.80ms step:3077/4578 train_loss:3.5303 train_time:434915ms step_avg:141.80ms step:3078/4578 train_loss:3.3375 train_time:435055ms step_avg:141.80ms step:3079/4578 train_loss:3.8692 train_time:435195ms step_avg:141.80ms step:3080/4578 train_loss:3.4255 train_time:435335ms step_avg:141.80ms step:3081/4578 train_loss:3.3936 train_time:435474ms step_avg:141.80ms step:3082/4578 train_loss:3.5333 train_time:435615ms step_avg:141.80ms step:3083/4578 train_loss:3.3483 train_time:435755ms step_avg:141.80ms step:3084/4578 train_loss:3.3764 train_time:435895ms step_avg:141.80ms step:3085/4578 train_loss:3.4243 train_time:436035ms step_avg:141.80ms step:3086/4578 train_loss:3.5184 train_time:436174ms step_avg:141.80ms step:3087/4578 train_loss:3.4295 train_time:436314ms step_avg:141.80ms step:3088/4578 train_loss:3.3409 train_time:436453ms step_avg:141.80ms step:3089/4578 train_loss:3.4916 train_time:436593ms step_avg:141.80ms step:3090/4578 train_loss:3.3553 train_time:436732ms step_avg:141.80ms step:3091/4578 train_loss:3.6186 train_time:436873ms step_avg:141.80ms step:3092/4578 train_loss:4.1808 train_time:437015ms step_avg:141.80ms step:3093/4578 train_loss:3.4504 train_time:437159ms step_avg:141.80ms step:3094/4578 train_loss:3.3400 train_time:437294ms step_avg:141.79ms step:3095/4578 train_loss:3.3020 train_time:437433ms step_avg:141.79ms step:3096/4578 train_loss:3.4742 train_time:437575ms step_avg:141.79ms step:3097/4578 train_loss:3.5932 train_time:437714ms step_avg:141.79ms step:3098/4578 train_loss:3.3741 train_time:437857ms step_avg:141.79ms step:3099/4578 train_loss:3.4060 train_time:437994ms step_avg:141.79ms step:3100/4578 train_loss:3.5860 train_time:438134ms step_avg:141.79ms step:3101/4578 train_loss:3.4812 train_time:438273ms step_avg:141.79ms step:3102/4578 train_loss:3.4888 train_time:438414ms step_avg:141.79ms step:3103/4578 train_loss:3.3868 train_time:438553ms step_avg:141.79ms step:3104/4578 train_loss:3.6408 train_time:438694ms step_avg:141.79ms step:3105/4578 train_loss:3.4598 train_time:438834ms step_avg:141.79ms step:3106/4578 train_loss:3.3181 train_time:438973ms step_avg:141.79ms step:3107/4578 train_loss:3.3510 train_time:439113ms step_avg:141.79ms step:3108/4578 train_loss:3.3058 train_time:439254ms step_avg:141.79ms step:3109/4578 train_loss:3.5253 train_time:439394ms step_avg:141.79ms step:3110/4578 train_loss:3.4109 train_time:439534ms step_avg:141.79ms step:3111/4578 train_loss:3.4537 train_time:439672ms step_avg:141.78ms step:3112/4578 train_loss:3.4325 train_time:439813ms step_avg:141.78ms step:3113/4578 train_loss:3.4816 train_time:439952ms step_avg:141.78ms step:3114/4578 train_loss:3.4291 train_time:440093ms step_avg:141.78ms step:3115/4578 train_loss:3.4454 train_time:440232ms step_avg:141.78ms step:3116/4578 train_loss:3.4729 train_time:440372ms step_avg:141.78ms step:3117/4578 train_loss:3.3362 train_time:440518ms step_avg:141.78ms step:3118/4578 train_loss:3.3516 train_time:440653ms step_avg:141.78ms step:3119/4578 train_loss:3.5318 train_time:440794ms step_avg:141.78ms step:3120/4578 train_loss:3.5070 train_time:440933ms step_avg:141.78ms step:3121/4578 train_loss:3.3029 train_time:441075ms step_avg:141.78ms step:3122/4578 train_loss:3.4927 train_time:441213ms step_avg:141.78ms step:3123/4578 train_loss:3.5467 train_time:441352ms step_avg:141.78ms step:3124/4578 train_loss:3.5178 train_time:441492ms step_avg:141.78ms step:3125/4578 train_loss:3.3114 train_time:441631ms step_avg:141.78ms step:3125/4578 val_loss:3.4210 train_time:441687ms step_avg:141.79ms step:3126/4578 train_loss:3.3937 train_time:441784ms step_avg:141.78ms step:3127/4578 train_loss:3.4339 train_time:441930ms step_avg:141.78ms step:3128/4578 train_loss:3.5180 train_time:442071ms step_avg:141.78ms step:3129/4578 train_loss:3.5938 train_time:442211ms step_avg:141.78ms step:3130/4578 train_loss:3.3014 train_time:442354ms step_avg:141.78ms step:3131/4578 train_loss:3.4598 train_time:442489ms step_avg:141.78ms step:3132/4578 train_loss:3.4597 train_time:442629ms step_avg:141.78ms step:3133/4578 train_loss:3.4728 train_time:442771ms step_avg:141.78ms step:3134/4578 train_loss:3.3800 train_time:442915ms step_avg:141.78ms step:3135/4578 train_loss:3.4960 train_time:443057ms step_avg:141.78ms step:3136/4578 train_loss:3.4115 train_time:443198ms step_avg:141.78ms step:3137/4578 train_loss:3.4720 train_time:443343ms step_avg:141.78ms step:3138/4578 train_loss:3.6646 train_time:443475ms step_avg:141.78ms step:3139/4578 train_loss:3.6312 train_time:443627ms step_avg:141.78ms step:3140/4578 train_loss:3.4012 train_time:443755ms step_avg:141.77ms step:3141/4578 train_loss:3.4218 train_time:443898ms step_avg:141.78ms step:3142/4578 train_loss:3.3400 train_time:444039ms step_avg:141.77ms step:3143/4578 train_loss:3.4293 train_time:444178ms step_avg:141.77ms step:3144/4578 train_loss:3.2333 train_time:444317ms step_avg:141.77ms step:3145/4578 train_loss:3.4726 train_time:444456ms step_avg:141.77ms step:3146/4578 train_loss:3.3865 train_time:444598ms step_avg:141.77ms step:3147/4578 train_loss:3.4081 train_time:444738ms step_avg:141.77ms step:3148/4578 train_loss:3.5851 train_time:444879ms step_avg:141.77ms step:3149/4578 train_loss:3.6656 train_time:445020ms step_avg:141.77ms step:3150/4578 train_loss:3.5329 train_time:445160ms step_avg:141.77ms step:3151/4578 train_loss:3.3401 train_time:445299ms step_avg:141.77ms step:3152/4578 train_loss:3.3880 train_time:445439ms step_avg:141.77ms step:3153/4578 train_loss:3.3698 train_time:445579ms step_avg:141.77ms step:3154/4578 train_loss:3.4856 train_time:445720ms step_avg:141.77ms step:3155/4578 train_loss:3.3059 train_time:445860ms step_avg:141.77ms step:3156/4578 train_loss:3.4472 train_time:446000ms step_avg:141.77ms step:3157/4578 train_loss:3.3954 train_time:446145ms step_avg:141.77ms step:3158/4578 train_loss:3.5053 train_time:446279ms step_avg:141.77ms step:3159/4578 train_loss:3.5719 train_time:446421ms step_avg:141.77ms step:3160/4578 train_loss:3.4203 train_time:446557ms step_avg:141.76ms step:3161/4578 train_loss:3.4802 train_time:446699ms step_avg:141.76ms step:3162/4578 train_loss:3.5568 train_time:446839ms step_avg:141.76ms step:3163/4578 train_loss:3.4578 train_time:446979ms step_avg:141.76ms step:3164/4578 train_loss:3.5185 train_time:447120ms step_avg:141.76ms step:3165/4578 train_loss:3.3332 train_time:447259ms step_avg:141.76ms step:3166/4578 train_loss:3.3283 train_time:447399ms step_avg:141.76ms step:3167/4578 train_loss:3.3568 train_time:447538ms step_avg:141.76ms step:3168/4578 train_loss:3.1945 train_time:447679ms step_avg:141.76ms step:3169/4578 train_loss:3.3533 train_time:447820ms step_avg:141.76ms step:3170/4578 train_loss:3.5017 train_time:447959ms step_avg:141.76ms step:3171/4578 train_loss:3.5321 train_time:448099ms step_avg:141.76ms step:3172/4578 train_loss:3.4853 train_time:448239ms step_avg:141.76ms step:3173/4578 train_loss:3.4560 train_time:448380ms step_avg:141.76ms step:3174/4578 train_loss:3.4255 train_time:448519ms step_avg:141.76ms step:3175/4578 train_loss:3.4274 train_time:448660ms step_avg:141.76ms step:3176/4578 train_loss:3.4372 train_time:448800ms step_avg:141.76ms step:3177/4578 train_loss:3.3583 train_time:448940ms step_avg:141.76ms step:3178/4578 train_loss:3.4845 train_time:449080ms step_avg:141.76ms step:3179/4578 train_loss:3.5649 train_time:449222ms step_avg:141.75ms step:3180/4578 train_loss:3.4058 train_time:449360ms step_avg:141.75ms step:3181/4578 train_loss:3.3942 train_time:449502ms step_avg:141.75ms step:3182/4578 train_loss:3.4371 train_time:449651ms step_avg:141.76ms step:3183/4578 train_loss:3.5331 train_time:449780ms step_avg:141.75ms step:3184/4578 train_loss:3.5513 train_time:449920ms step_avg:141.75ms step:3185/4578 train_loss:3.4510 train_time:450060ms step_avg:141.75ms step:3186/4578 train_loss:3.5195 train_time:450200ms step_avg:141.75ms step:3187/4578 train_loss:3.5037 train_time:450340ms step_avg:141.75ms step:3188/4578 train_loss:3.2999 train_time:450480ms step_avg:141.75ms step:3189/4578 train_loss:3.4616 train_time:450620ms step_avg:141.75ms step:3190/4578 train_loss:3.4103 train_time:450760ms step_avg:141.75ms step:3191/4578 train_loss:3.4348 train_time:450901ms step_avg:141.75ms step:3192/4578 train_loss:3.4021 train_time:451067ms step_avg:141.76ms step:3193/4578 train_loss:3.3287 train_time:451181ms step_avg:141.75ms step:3194/4578 train_loss:4.3467 train_time:451321ms step_avg:141.75ms step:3195/4578 train_loss:3.4465 train_time:451461ms step_avg:141.75ms step:3196/4578 train_loss:3.2425 train_time:451602ms step_avg:141.75ms step:3197/4578 train_loss:3.3905 train_time:451742ms step_avg:141.75ms step:3198/4578 train_loss:3.2786 train_time:451882ms step_avg:141.74ms step:3199/4578 train_loss:3.3838 train_time:452023ms step_avg:141.74ms step:3200/4578 train_loss:3.3171 train_time:452163ms step_avg:141.74ms step:3201/4578 train_loss:3.3948 train_time:452303ms step_avg:141.74ms step:3202/4578 train_loss:3.4934 train_time:452443ms step_avg:141.74ms step:3203/4578 train_loss:3.3468 train_time:452583ms step_avg:141.74ms step:3204/4578 train_loss:3.3906 train_time:452724ms step_avg:141.74ms step:3205/4578 train_loss:3.4768 train_time:452863ms step_avg:141.74ms step:3206/4578 train_loss:3.6273 train_time:453003ms step_avg:141.74ms step:3207/4578 train_loss:3.2372 train_time:453142ms step_avg:141.74ms step:3208/4578 train_loss:3.5834 train_time:453282ms step_avg:141.74ms step:3209/4578 train_loss:3.4321 train_time:453422ms step_avg:141.74ms step:3210/4578 train_loss:3.5065 train_time:453561ms step_avg:141.74ms step:3211/4578 train_loss:3.5929 train_time:453701ms step_avg:141.74ms step:3212/4578 train_loss:3.2684 train_time:453842ms step_avg:141.74ms step:3213/4578 train_loss:3.3275 train_time:454114ms step_avg:141.78ms step:3214/4578 train_loss:3.5405 train_time:454249ms step_avg:141.78ms step:3215/4578 train_loss:3.3262 train_time:454388ms step_avg:141.77ms step:3216/4578 train_loss:3.3987 train_time:454528ms step_avg:141.77ms step:3217/4578 train_loss:3.2987 train_time:454667ms step_avg:141.77ms step:3218/4578 train_loss:3.4330 train_time:454806ms step_avg:141.77ms step:3219/4578 train_loss:3.4725 train_time:454946ms step_avg:141.77ms step:3220/4578 train_loss:3.5201 train_time:455090ms step_avg:141.77ms step:3221/4578 train_loss:3.4615 train_time:455233ms step_avg:141.77ms step:3222/4578 train_loss:3.4634 train_time:455373ms step_avg:141.77ms step:3223/4578 train_loss:3.3336 train_time:455513ms step_avg:141.77ms step:3224/4578 train_loss:3.3558 train_time:455652ms step_avg:141.77ms step:3225/4578 train_loss:3.3605 train_time:455794ms step_avg:141.77ms step:3226/4578 train_loss:3.3909 train_time:455933ms step_avg:141.77ms step:3227/4578 train_loss:3.3207 train_time:456075ms step_avg:141.77ms step:3228/4578 train_loss:3.2523 train_time:456217ms step_avg:141.77ms step:3229/4578 train_loss:3.3668 train_time:456357ms step_avg:141.77ms step:3230/4578 train_loss:3.1357 train_time:456662ms step_avg:141.82ms step:3231/4578 train_loss:3.3168 train_time:456796ms step_avg:141.82ms step:3232/4578 train_loss:3.3173 train_time:456934ms step_avg:141.82ms step:3233/4578 train_loss:3.5533 train_time:457072ms step_avg:141.82ms step:3234/4578 train_loss:3.5429 train_time:457211ms step_avg:141.81ms step:3235/4578 train_loss:3.5030 train_time:457349ms step_avg:141.81ms step:3236/4578 train_loss:3.3918 train_time:457489ms step_avg:141.81ms step:3237/4578 train_loss:3.5493 train_time:457636ms step_avg:141.81ms step:3238/4578 train_loss:3.4091 train_time:457778ms step_avg:141.81ms step:3239/4578 train_loss:3.5364 train_time:457918ms step_avg:141.81ms step:3240/4578 train_loss:3.5113 train_time:458056ms step_avg:141.81ms step:3241/4578 train_loss:3.4061 train_time:458200ms step_avg:141.81ms step:3242/4578 train_loss:3.3638 train_time:458335ms step_avg:141.81ms step:3243/4578 train_loss:3.5864 train_time:458476ms step_avg:141.81ms step:3244/4578 train_loss:3.4586 train_time:458618ms step_avg:141.81ms step:3245/4578 train_loss:3.5006 train_time:458761ms step_avg:141.81ms step:3246/4578 train_loss:3.3922 train_time:458901ms step_avg:141.81ms step:3247/4578 train_loss:3.5227 train_time:459041ms step_avg:141.81ms step:3248/4578 train_loss:3.4470 train_time:459180ms step_avg:141.81ms step:3249/4578 train_loss:3.3946 train_time:459319ms step_avg:141.81ms step:3250/4578 train_loss:3.2717 train_time:459458ms step_avg:141.81ms step:3250/4578 val_loss:3.4129 train_time:459516ms step_avg:141.83ms step:3251/4578 train_loss:3.4712 train_time:459611ms step_avg:141.81ms step:3252/4578 train_loss:3.4626 train_time:459757ms step_avg:141.81ms step:3253/4578 train_loss:3.4270 train_time:459901ms step_avg:141.81ms step:3254/4578 train_loss:3.3394 train_time:460038ms step_avg:141.81ms step:3255/4578 train_loss:3.4955 train_time:460178ms step_avg:141.81ms step:3256/4578 train_loss:3.5271 train_time:460319ms step_avg:141.81ms step:3257/4578 train_loss:3.4652 train_time:460457ms step_avg:141.81ms step:3258/4578 train_loss:3.4965 train_time:460604ms step_avg:141.81ms step:3259/4578 train_loss:3.3389 train_time:460744ms step_avg:141.81ms step:3260/4578 train_loss:3.4189 train_time:460885ms step_avg:141.81ms step:3261/4578 train_loss:3.2888 train_time:461025ms step_avg:141.81ms step:3262/4578 train_loss:3.3189 train_time:461164ms step_avg:141.81ms step:3263/4578 train_loss:3.3539 train_time:461304ms step_avg:141.81ms step:3264/4578 train_loss:3.5057 train_time:461443ms step_avg:141.81ms step:3265/4578 train_loss:3.3902 train_time:461584ms step_avg:141.81ms step:3266/4578 train_loss:3.4477 train_time:461727ms step_avg:141.81ms step:3267/4578 train_loss:3.4638 train_time:461869ms step_avg:141.81ms step:3268/4578 train_loss:3.5491 train_time:462007ms step_avg:141.81ms step:3269/4578 train_loss:3.3546 train_time:462146ms step_avg:141.81ms step:3270/4578 train_loss:3.4831 train_time:462287ms step_avg:141.81ms step:3271/4578 train_loss:3.3542 train_time:462429ms step_avg:141.81ms step:3272/4578 train_loss:3.2609 train_time:462568ms step_avg:141.81ms step:3273/4578 train_loss:3.3643 train_time:462708ms step_avg:141.80ms step:3274/4578 train_loss:3.4996 train_time:462848ms step_avg:141.80ms step:3275/4578 train_loss:3.3041 train_time:463000ms step_avg:141.81ms step:3276/4578 train_loss:3.4513 train_time:463128ms step_avg:141.80ms step:3277/4578 train_loss:3.4485 train_time:463268ms step_avg:141.80ms step:3278/4578 train_loss:3.4392 train_time:463407ms step_avg:141.80ms step:3279/4578 train_loss:3.4103 train_time:463546ms step_avg:141.80ms step:3280/4578 train_loss:3.5662 train_time:463688ms step_avg:141.80ms step:3281/4578 train_loss:3.4126 train_time:463832ms step_avg:141.80ms step:3282/4578 train_loss:3.4588 train_time:463967ms step_avg:141.80ms step:3283/4578 train_loss:3.3179 train_time:464107ms step_avg:141.80ms step:3284/4578 train_loss:3.4482 train_time:464247ms step_avg:141.80ms step:3285/4578 train_loss:3.4952 train_time:464387ms step_avg:141.80ms step:3286/4578 train_loss:3.4818 train_time:464526ms step_avg:141.80ms step:3287/4578 train_loss:3.5127 train_time:464668ms step_avg:141.80ms step:3288/4578 train_loss:3.3934 train_time:464812ms step_avg:141.80ms step:3289/4578 train_loss:3.5110 train_time:464951ms step_avg:141.80ms step:3290/4578 train_loss:3.4330 train_time:465089ms step_avg:141.80ms step:3291/4578 train_loss:3.3083 train_time:465227ms step_avg:141.79ms step:3292/4578 train_loss:3.4386 train_time:465367ms step_avg:141.79ms step:3293/4578 train_loss:3.4696 train_time:465511ms step_avg:141.79ms step:3294/4578 train_loss:3.4553 train_time:465647ms step_avg:141.79ms step:3295/4578 train_loss:3.3327 train_time:465787ms step_avg:141.79ms step:3296/4578 train_loss:3.3965 train_time:465927ms step_avg:141.79ms step:3297/4578 train_loss:3.4398 train_time:466087ms step_avg:141.80ms step:3298/4578 train_loss:3.4335 train_time:466208ms step_avg:141.79ms step:3299/4578 train_loss:3.4179 train_time:466347ms step_avg:141.79ms step:3300/4578 train_loss:3.4782 train_time:466487ms step_avg:141.79ms step:3301/4578 train_loss:3.3784 train_time:466627ms step_avg:141.79ms step:3302/4578 train_loss:3.4488 train_time:466767ms step_avg:141.79ms step:3303/4578 train_loss:3.3926 train_time:466908ms step_avg:141.79ms step:3304/4578 train_loss:3.4066 train_time:467047ms step_avg:141.79ms step:3305/4578 train_loss:3.3990 train_time:467188ms step_avg:141.79ms step:3306/4578 train_loss:3.4989 train_time:467331ms step_avg:141.79ms step:3307/4578 train_loss:3.4298 train_time:467467ms step_avg:141.79ms step:3308/4578 train_loss:3.3958 train_time:467608ms step_avg:141.79ms step:3309/4578 train_loss:3.5139 train_time:467747ms step_avg:141.78ms step:3310/4578 train_loss:3.3900 train_time:467887ms step_avg:141.78ms step:3311/4578 train_loss:3.3337 train_time:468027ms step_avg:141.78ms step:3312/4578 train_loss:3.4413 train_time:468167ms step_avg:141.78ms step:3313/4578 train_loss:3.4155 train_time:468315ms step_avg:141.78ms step:3314/4578 train_loss:3.6143 train_time:468451ms step_avg:141.78ms step:3315/4578 train_loss:3.4413 train_time:468593ms step_avg:141.78ms step:3316/4578 train_loss:3.4049 train_time:468735ms step_avg:141.78ms step:3317/4578 train_loss:3.0398 train_time:468871ms step_avg:141.78ms step:3318/4578 train_loss:3.5484 train_time:469012ms step_avg:141.78ms step:3319/4578 train_loss:3.3858 train_time:469153ms step_avg:141.78ms step:3320/4578 train_loss:3.4628 train_time:469288ms step_avg:141.78ms step:3321/4578 train_loss:3.3903 train_time:469428ms step_avg:141.78ms step:3322/4578 train_loss:3.4581 train_time:469570ms step_avg:141.78ms step:3323/4578 train_loss:3.3980 train_time:469707ms step_avg:141.78ms step:3324/4578 train_loss:3.3249 train_time:469853ms step_avg:141.78ms step:3325/4578 train_loss:3.2613 train_time:469988ms step_avg:141.78ms step:3326/4578 train_loss:3.4196 train_time:470128ms step_avg:141.78ms step:3327/4578 train_loss:3.3820 train_time:470267ms step_avg:141.77ms step:3328/4578 train_loss:3.3044 train_time:470407ms step_avg:141.77ms step:3329/4578 train_loss:3.3428 train_time:470547ms step_avg:141.77ms step:3330/4578 train_loss:3.3062 train_time:470691ms step_avg:141.77ms step:3331/4578 train_loss:3.5462 train_time:470829ms step_avg:141.77ms step:3332/4578 train_loss:3.4405 train_time:470969ms step_avg:141.77ms step:3333/4578 train_loss:3.4301 train_time:471109ms step_avg:141.77ms step:3334/4578 train_loss:3.2781 train_time:471249ms step_avg:141.77ms step:3335/4578 train_loss:3.3631 train_time:471389ms step_avg:141.77ms step:3336/4578 train_loss:3.4647 train_time:471530ms step_avg:141.77ms step:3337/4578 train_loss:3.4334 train_time:471670ms step_avg:141.77ms step:3338/4578 train_loss:3.4663 train_time:471810ms step_avg:141.77ms step:3339/4578 train_loss:3.4004 train_time:471949ms step_avg:141.77ms step:3340/4578 train_loss:3.4176 train_time:472094ms step_avg:141.77ms step:3341/4578 train_loss:3.4388 train_time:472237ms step_avg:141.77ms step:3342/4578 train_loss:3.4493 train_time:472369ms step_avg:141.77ms step:3343/4578 train_loss:3.4341 train_time:472510ms step_avg:141.77ms step:3344/4578 train_loss:3.3631 train_time:472649ms step_avg:141.77ms step:3345/4578 train_loss:3.2802 train_time:472790ms step_avg:141.77ms step:3346/4578 train_loss:3.6072 train_time:472929ms step_avg:141.77ms step:3347/4578 train_loss:3.3653 train_time:473071ms step_avg:141.77ms step:3348/4578 train_loss:3.5278 train_time:473211ms step_avg:141.76ms step:3349/4578 train_loss:3.3934 train_time:473354ms step_avg:141.77ms step:3350/4578 train_loss:3.4809 train_time:473489ms step_avg:141.76ms step:3351/4578 train_loss:3.2204 train_time:473629ms step_avg:141.76ms step:3352/4578 train_loss:3.2478 train_time:473768ms step_avg:141.76ms step:3353/4578 train_loss:3.4218 train_time:473908ms step_avg:141.76ms step:3354/4578 train_loss:3.2965 train_time:474048ms step_avg:141.76ms step:3355/4578 train_loss:3.4459 train_time:474189ms step_avg:141.76ms step:3356/4578 train_loss:3.3123 train_time:474329ms step_avg:141.76ms step:3357/4578 train_loss:3.4767 train_time:474469ms step_avg:141.76ms step:3358/4578 train_loss:3.3391 train_time:474610ms step_avg:141.76ms step:3359/4578 train_loss:3.5038 train_time:474749ms step_avg:141.76ms step:3360/4578 train_loss:3.3177 train_time:474889ms step_avg:141.76ms step:3361/4578 train_loss:3.9873 train_time:475029ms step_avg:141.76ms step:3362/4578 train_loss:3.4750 train_time:475169ms step_avg:141.76ms step:3363/4578 train_loss:3.5034 train_time:475309ms step_avg:141.76ms step:3364/4578 train_loss:3.3797 train_time:475449ms step_avg:141.76ms step:3365/4578 train_loss:3.4966 train_time:475589ms step_avg:141.76ms step:3366/4578 train_loss:3.3990 train_time:475728ms step_avg:141.75ms step:3367/4578 train_loss:3.5700 train_time:475867ms step_avg:141.75ms step:3368/4578 train_loss:3.3786 train_time:476008ms step_avg:141.75ms step:3369/4578 train_loss:3.3964 train_time:476146ms step_avg:141.75ms step:3370/4578 train_loss:3.3650 train_time:476287ms step_avg:141.75ms step:3371/4578 train_loss:3.3279 train_time:476435ms step_avg:141.75ms step:3372/4578 train_loss:3.3311 train_time:476575ms step_avg:141.75ms step:3373/4578 train_loss:3.3932 train_time:476707ms step_avg:141.75ms step:3374/4578 train_loss:3.4263 train_time:476847ms step_avg:141.75ms step:3375/4578 train_loss:3.3927 train_time:476987ms step_avg:141.75ms step:3375/4578 val_loss:3.4020 train_time:477043ms step_avg:141.77ms step:3376/4578 train_loss:3.4382 train_time:477140ms step_avg:141.75ms step:3377/4578 train_loss:3.4329 train_time:477288ms step_avg:141.75ms step:3378/4578 train_loss:3.5158 train_time:477435ms step_avg:141.76ms step:3379/4578 train_loss:3.3685 train_time:477571ms step_avg:141.75ms step:3380/4578 train_loss:3.3796 train_time:477704ms step_avg:141.75ms step:3381/4578 train_loss:3.3922 train_time:477849ms step_avg:141.75ms step:3382/4578 train_loss:3.4909 train_time:477980ms step_avg:141.75ms step:3383/4578 train_loss:3.3300 train_time:478122ms step_avg:141.75ms step:3384/4578 train_loss:3.5022 train_time:478272ms step_avg:141.75ms step:3385/4578 train_loss:3.3534 train_time:478413ms step_avg:141.75ms step:3386/4578 train_loss:3.3820 train_time:478550ms step_avg:141.75ms step:3387/4578 train_loss:3.3240 train_time:478693ms step_avg:141.75ms step:3388/4578 train_loss:3.4964 train_time:478832ms step_avg:141.75ms step:3389/4578 train_loss:3.4589 train_time:478968ms step_avg:141.75ms step:3390/4578 train_loss:3.4764 train_time:479109ms step_avg:141.75ms step:3391/4578 train_loss:3.4593 train_time:479250ms step_avg:141.75ms step:3392/4578 train_loss:3.4000 train_time:479390ms step_avg:141.75ms step:3393/4578 train_loss:3.5199 train_time:479532ms step_avg:141.75ms step:3394/4578 train_loss:3.4823 train_time:479670ms step_avg:141.75ms step:3395/4578 train_loss:3.5654 train_time:479811ms step_avg:141.75ms step:3396/4578 train_loss:3.4292 train_time:479951ms step_avg:141.75ms step:3397/4578 train_loss:3.4142 train_time:480091ms step_avg:141.75ms step:3398/4578 train_loss:3.3685 train_time:480237ms step_avg:141.75ms step:3399/4578 train_loss:3.4294 train_time:480374ms step_avg:141.74ms step:3400/4578 train_loss:3.4188 train_time:480516ms step_avg:141.75ms step:3401/4578 train_loss:3.5047 train_time:480660ms step_avg:141.75ms step:3402/4578 train_loss:3.3775 train_time:480933ms step_avg:141.78ms step:3403/4578 train_loss:3.5588 train_time:481070ms step_avg:141.78ms step:3404/4578 train_loss:3.3848 train_time:481209ms step_avg:141.78ms step:3405/4578 train_loss:3.4003 train_time:481348ms step_avg:141.78ms step:3406/4578 train_loss:3.3453 train_time:481487ms step_avg:141.78ms step:3407/4578 train_loss:3.4131 train_time:481625ms step_avg:141.78ms step:3408/4578 train_loss:3.4121 train_time:481764ms step_avg:141.78ms step:3409/4578 train_loss:3.3936 train_time:481910ms step_avg:141.78ms step:3410/4578 train_loss:3.4073 train_time:482051ms step_avg:141.78ms step:3411/4578 train_loss:3.3752 train_time:482192ms step_avg:141.78ms step:3412/4578 train_loss:3.4077 train_time:482334ms step_avg:141.78ms step:3413/4578 train_loss:3.3338 train_time:482474ms step_avg:141.78ms step:3414/4578 train_loss:3.5474 train_time:482614ms step_avg:141.78ms step:3415/4578 train_loss:3.2997 train_time:482755ms step_avg:141.78ms step:3416/4578 train_loss:3.4643 train_time:482895ms step_avg:141.78ms step:3417/4578 train_loss:3.3413 train_time:483038ms step_avg:141.78ms step:3418/4578 train_loss:3.4358 train_time:483179ms step_avg:141.78ms step:3419/4578 train_loss:3.4419 train_time:483321ms step_avg:141.78ms step:3420/4578 train_loss:3.4670 train_time:483631ms step_avg:141.83ms step:3421/4578 train_loss:3.3472 train_time:483763ms step_avg:141.82ms step:3422/4578 train_loss:3.3823 train_time:483909ms step_avg:141.83ms step:3423/4578 train_loss:3.3229 train_time:484042ms step_avg:141.82ms step:3424/4578 train_loss:3.6392 train_time:484186ms step_avg:141.82ms step:3425/4578 train_loss:3.5228 train_time:484317ms step_avg:141.82ms step:3426/4578 train_loss:3.3972 train_time:484459ms step_avg:141.82ms step:3427/4578 train_loss:3.3538 train_time:484605ms step_avg:141.82ms step:3428/4578 train_loss:3.3322 train_time:484749ms step_avg:141.82ms step:3429/4578 train_loss:3.3356 train_time:484890ms step_avg:141.82ms step:3430/4578 train_loss:3.3938 train_time:485031ms step_avg:141.82ms step:3431/4578 train_loss:3.4172 train_time:485170ms step_avg:141.82ms step:3432/4578 train_loss:3.5111 train_time:485309ms step_avg:141.82ms step:3433/4578 train_loss:3.3295 train_time:485451ms step_avg:141.82ms step:3434/4578 train_loss:3.5591 train_time:485590ms step_avg:141.82ms step:3435/4578 train_loss:3.4704 train_time:485731ms step_avg:141.82ms step:3436/4578 train_loss:3.3133 train_time:485871ms step_avg:141.82ms step:3437/4578 train_loss:3.3570 train_time:486011ms step_avg:141.82ms step:3438/4578 train_loss:3.4056 train_time:486152ms step_avg:141.82ms step:3439/4578 train_loss:3.4956 train_time:486300ms step_avg:141.82ms step:3440/4578 train_loss:3.2684 train_time:486432ms step_avg:141.82ms step:3441/4578 train_loss:3.4549 train_time:486571ms step_avg:141.82ms step:3442/4578 train_loss:3.3512 train_time:486717ms step_avg:141.82ms step:3443/4578 train_loss:3.5377 train_time:486854ms step_avg:141.82ms step:3444/4578 train_loss:3.4026 train_time:486995ms step_avg:141.82ms step:3445/4578 train_loss:3.2894 train_time:487139ms step_avg:141.82ms step:3446/4578 train_loss:3.4985 train_time:487276ms step_avg:141.82ms step:3447/4578 train_loss:3.5760 train_time:487418ms step_avg:141.81ms step:3448/4578 train_loss:3.3895 train_time:487558ms step_avg:141.81ms step:3449/4578 train_loss:3.4061 train_time:487698ms step_avg:141.81ms step:3450/4578 train_loss:3.4938 train_time:487840ms step_avg:141.81ms step:3451/4578 train_loss:3.4912 train_time:487979ms step_avg:141.81ms step:3452/4578 train_loss:3.4941 train_time:488121ms step_avg:141.81ms step:3453/4578 train_loss:3.2975 train_time:488261ms step_avg:141.81ms step:3454/4578 train_loss:3.4189 train_time:488402ms step_avg:141.81ms step:3455/4578 train_loss:3.2991 train_time:488546ms step_avg:141.81ms step:3456/4578 train_loss:3.5919 train_time:488687ms step_avg:141.81ms step:3457/4578 train_loss:3.2685 train_time:488822ms step_avg:141.81ms step:3458/4578 train_loss:3.4124 train_time:488962ms step_avg:141.81ms step:3459/4578 train_loss:3.3565 train_time:489105ms step_avg:141.81ms step:3460/4578 train_loss:3.3592 train_time:489241ms step_avg:141.81ms step:3461/4578 train_loss:3.3551 train_time:489384ms step_avg:141.81ms step:3462/4578 train_loss:3.3585 train_time:489522ms step_avg:141.81ms step:3463/4578 train_loss:3.4625 train_time:489662ms step_avg:141.81ms step:3464/4578 train_loss:3.3365 train_time:489802ms step_avg:141.81ms step:3465/4578 train_loss:3.3551 train_time:489942ms step_avg:141.81ms step:3466/4578 train_loss:3.3235 train_time:490083ms step_avg:141.81ms step:3467/4578 train_loss:3.4831 train_time:490228ms step_avg:141.81ms step:3468/4578 train_loss:3.3801 train_time:490362ms step_avg:141.81ms step:3469/4578 train_loss:3.3907 train_time:490506ms step_avg:141.81ms step:3470/4578 train_loss:3.5733 train_time:490646ms step_avg:141.81ms step:3471/4578 train_loss:3.4715 train_time:490788ms step_avg:141.81ms step:3472/4578 train_loss:3.5202 train_time:490922ms step_avg:141.80ms step:3473/4578 train_loss:4.1450 train_time:491063ms step_avg:141.80ms step:3474/4578 train_loss:3.3977 train_time:491203ms step_avg:141.80ms step:3475/4578 train_loss:3.3995 train_time:491343ms step_avg:141.80ms step:3476/4578 train_loss:3.3843 train_time:491482ms step_avg:141.80ms step:3477/4578 train_loss:3.3331 train_time:491622ms step_avg:141.80ms step:3478/4578 train_loss:3.3990 train_time:491762ms step_avg:141.80ms step:3479/4578 train_loss:3.3933 train_time:491903ms step_avg:141.80ms step:3480/4578 train_loss:3.2910 train_time:492045ms step_avg:141.80ms step:3481/4578 train_loss:3.5975 train_time:492182ms step_avg:141.80ms step:3482/4578 train_loss:3.4583 train_time:492325ms step_avg:141.80ms step:3483/4578 train_loss:3.4173 train_time:492462ms step_avg:141.80ms step:3484/4578 train_loss:3.4326 train_time:492602ms step_avg:141.80ms step:3485/4578 train_loss:3.4023 train_time:492748ms step_avg:141.80ms step:3486/4578 train_loss:3.5897 train_time:492882ms step_avg:141.80ms step:3487/4578 train_loss:3.6137 train_time:493029ms step_avg:141.80ms step:3488/4578 train_loss:3.4656 train_time:493162ms step_avg:141.79ms step:3489/4578 train_loss:3.3270 train_time:493303ms step_avg:141.79ms step:3490/4578 train_loss:3.4956 train_time:493444ms step_avg:141.79ms step:3491/4578 train_loss:3.4006 train_time:493588ms step_avg:141.79ms step:3492/4578 train_loss:3.4429 train_time:493724ms step_avg:141.79ms step:3493/4578 train_loss:3.2845 train_time:493863ms step_avg:141.79ms step:3494/4578 train_loss:3.4229 train_time:494003ms step_avg:141.79ms step:3495/4578 train_loss:3.3725 train_time:494143ms step_avg:141.79ms step:3496/4578 train_loss:3.3891 train_time:494282ms step_avg:141.79ms step:3497/4578 train_loss:3.5593 train_time:494422ms step_avg:141.79ms step:3498/4578 train_loss:3.3881 train_time:494562ms step_avg:141.79ms step:3499/4578 train_loss:3.4214 train_time:494703ms step_avg:141.79ms step:3500/4578 train_loss:3.4130 train_time:494842ms step_avg:141.79ms step:3500/4578 val_loss:3.3846 train_time:494897ms step_avg:141.80ms step:3501/4578 train_loss:3.4564 train_time:494994ms step_avg:141.79ms step:3502/4578 train_loss:3.5244 train_time:495140ms step_avg:141.79ms step:3503/4578 train_loss:3.2335 train_time:495280ms step_avg:141.79ms step:3504/4578 train_loss:3.3977 train_time:495418ms step_avg:141.79ms step:3505/4578 train_loss:3.4237 train_time:495562ms step_avg:141.79ms step:3506/4578 train_loss:3.4441 train_time:495694ms step_avg:141.79ms step:3507/4578 train_loss:3.3156 train_time:495843ms step_avg:141.79ms step:3508/4578 train_loss:3.4741 train_time:495986ms step_avg:141.79ms step:3509/4578 train_loss:3.3624 train_time:496117ms step_avg:141.79ms step:3510/4578 train_loss:3.5785 train_time:496259ms step_avg:141.79ms step:3511/4578 train_loss:3.3794 train_time:496398ms step_avg:141.79ms step:3512/4578 train_loss:3.3354 train_time:496543ms step_avg:141.79ms step:3513/4578 train_loss:3.3955 train_time:496681ms step_avg:141.79ms step:3514/4578 train_loss:3.3560 train_time:496816ms step_avg:141.79ms step:3515/4578 train_loss:3.4181 train_time:496962ms step_avg:141.79ms step:3516/4578 train_loss:3.4126 train_time:497097ms step_avg:141.78ms step:3517/4578 train_loss:3.3858 train_time:497238ms step_avg:141.78ms step:3518/4578 train_loss:3.4021 train_time:497381ms step_avg:141.78ms step:3519/4578 train_loss:3.3956 train_time:497518ms step_avg:141.78ms step:3520/4578 train_loss:3.4170 train_time:497663ms step_avg:141.78ms step:3521/4578 train_loss:3.4840 train_time:497797ms step_avg:141.78ms step:3522/4578 train_loss:3.4011 train_time:497937ms step_avg:141.78ms step:3523/4578 train_loss:3.3233 train_time:498077ms step_avg:141.78ms step:3524/4578 train_loss:3.3663 train_time:498220ms step_avg:141.78ms step:3525/4578 train_loss:3.3620 train_time:498363ms step_avg:141.78ms step:3526/4578 train_loss:3.3625 train_time:498502ms step_avg:141.78ms step:3527/4578 train_loss:3.4949 train_time:498641ms step_avg:141.78ms step:3528/4578 train_loss:3.3196 train_time:498780ms step_avg:141.78ms step:3529/4578 train_loss:3.2014 train_time:498920ms step_avg:141.78ms step:3530/4578 train_loss:3.5001 train_time:499061ms step_avg:141.78ms step:3531/4578 train_loss:3.3069 train_time:499201ms step_avg:141.78ms step:3532/4578 train_loss:3.3524 train_time:499343ms step_avg:141.78ms step:3533/4578 train_loss:3.2641 train_time:499481ms step_avg:141.78ms step:3534/4578 train_loss:3.3248 train_time:499621ms step_avg:141.78ms step:3535/4578 train_loss:3.2934 train_time:499762ms step_avg:141.78ms step:3536/4578 train_loss:3.4669 train_time:499905ms step_avg:141.78ms step:3537/4578 train_loss:3.4295 train_time:500041ms step_avg:141.78ms step:3538/4578 train_loss:3.4498 train_time:500186ms step_avg:141.78ms step:3539/4578 train_loss:3.3483 train_time:500324ms step_avg:141.78ms step:3540/4578 train_loss:3.3236 train_time:500467ms step_avg:141.78ms step:3541/4578 train_loss:3.4535 train_time:500602ms step_avg:141.77ms step:3542/4578 train_loss:3.3093 train_time:500742ms step_avg:141.77ms step:3543/4578 train_loss:3.5337 train_time:500881ms step_avg:141.77ms step:3544/4578 train_loss:3.6378 train_time:501021ms step_avg:141.77ms step:3545/4578 train_loss:3.4648 train_time:501161ms step_avg:141.77ms step:3546/4578 train_loss:3.4891 train_time:501302ms step_avg:141.77ms step:3547/4578 train_loss:3.2624 train_time:501443ms step_avg:141.77ms step:3548/4578 train_loss:3.3369 train_time:501583ms step_avg:141.77ms step:3549/4578 train_loss:3.3710 train_time:501724ms step_avg:141.77ms step:3550/4578 train_loss:3.4570 train_time:501864ms step_avg:141.77ms step:3551/4578 train_loss:3.4161 train_time:502005ms step_avg:141.77ms step:3552/4578 train_loss:3.3449 train_time:502149ms step_avg:141.77ms step:3553/4578 train_loss:3.4584 train_time:502287ms step_avg:141.77ms step:3554/4578 train_loss:3.3929 train_time:502436ms step_avg:141.77ms step:3555/4578 train_loss:3.3502 train_time:502569ms step_avg:141.77ms step:3556/4578 train_loss:3.2867 train_time:502709ms step_avg:141.77ms step:3557/4578 train_loss:3.2434 train_time:502851ms step_avg:141.77ms step:3558/4578 train_loss:3.3537 train_time:503006ms step_avg:141.77ms step:3559/4578 train_loss:3.3834 train_time:503131ms step_avg:141.77ms step:3560/4578 train_loss:3.5928 train_time:503271ms step_avg:141.77ms step:3561/4578 train_loss:3.4594 train_time:503414ms step_avg:141.77ms step:3562/4578 train_loss:3.3625 train_time:503552ms step_avg:141.77ms step:3563/4578 train_loss:3.2391 train_time:503691ms step_avg:141.76ms step:3564/4578 train_loss:3.7483 train_time:503834ms step_avg:141.77ms step:3565/4578 train_loss:3.3198 train_time:503973ms step_avg:141.76ms step:3566/4578 train_loss:3.2635 train_time:504113ms step_avg:141.76ms step:3567/4578 train_loss:3.2882 train_time:504254ms step_avg:141.76ms step:3568/4578 train_loss:3.4290 train_time:504397ms step_avg:141.76ms step:3569/4578 train_loss:3.3616 train_time:504534ms step_avg:141.76ms step:3570/4578 train_loss:3.4791 train_time:504674ms step_avg:141.76ms step:3571/4578 train_loss:3.4064 train_time:504814ms step_avg:141.76ms step:3572/4578 train_loss:3.6771 train_time:504955ms step_avg:141.76ms step:3573/4578 train_loss:3.3031 train_time:505094ms step_avg:141.76ms step:3574/4578 train_loss:3.3856 train_time:505234ms step_avg:141.76ms step:3575/4578 train_loss:3.5477 train_time:505374ms step_avg:141.76ms step:3576/4578 train_loss:3.4247 train_time:505514ms step_avg:141.76ms step:3577/4578 train_loss:3.3631 train_time:505654ms step_avg:141.76ms step:3578/4578 train_loss:3.3323 train_time:505796ms step_avg:141.76ms step:3579/4578 train_loss:3.4054 train_time:505937ms step_avg:141.76ms step:3580/4578 train_loss:3.3586 train_time:506075ms step_avg:141.76ms step:3581/4578 train_loss:3.2569 train_time:506215ms step_avg:141.76ms step:3582/4578 train_loss:3.3361 train_time:506355ms step_avg:141.76ms step:3583/4578 train_loss:3.2875 train_time:506495ms step_avg:141.76ms step:3584/4578 train_loss:3.3882 train_time:506634ms step_avg:141.76ms step:3585/4578 train_loss:3.4732 train_time:506774ms step_avg:141.75ms step:3586/4578 train_loss:3.3356 train_time:506914ms step_avg:141.75ms step:3587/4578 train_loss:3.3777 train_time:507060ms step_avg:141.76ms step:3588/4578 train_loss:3.3844 train_time:507202ms step_avg:141.76ms step:3589/4578 train_loss:3.3549 train_time:507335ms step_avg:141.75ms step:3590/4578 train_loss:3.3569 train_time:507475ms step_avg:141.75ms step:3591/4578 train_loss:3.4763 train_time:507750ms step_avg:141.79ms step:3592/4578 train_loss:3.3519 train_time:507884ms step_avg:141.79ms step:3593/4578 train_loss:3.4430 train_time:508023ms step_avg:141.79ms step:3594/4578 train_loss:3.4245 train_time:508162ms step_avg:141.79ms step:3595/4578 train_loss:3.3878 train_time:508300ms step_avg:141.79ms step:3596/4578 train_loss:3.3079 train_time:508439ms step_avg:141.78ms step:3597/4578 train_loss:3.3103 train_time:508576ms step_avg:141.78ms step:3598/4578 train_loss:3.5803 train_time:508723ms step_avg:141.78ms step:3599/4578 train_loss:3.3704 train_time:508863ms step_avg:141.78ms step:3600/4578 train_loss:3.3759 train_time:509003ms step_avg:141.78ms step:3601/4578 train_loss:3.2356 train_time:509143ms step_avg:141.78ms step:3602/4578 train_loss:3.4193 train_time:509282ms step_avg:141.78ms step:3603/4578 train_loss:3.3635 train_time:509422ms step_avg:141.78ms step:3604/4578 train_loss:3.5003 train_time:509561ms step_avg:141.78ms step:3605/4578 train_loss:3.5576 train_time:509700ms step_avg:141.78ms step:3606/4578 train_loss:3.3587 train_time:509844ms step_avg:141.78ms step:3607/4578 train_loss:3.3926 train_time:509994ms step_avg:141.78ms step:3608/4578 train_loss:3.6462 train_time:510122ms step_avg:141.78ms step:3609/4578 train_loss:3.3543 train_time:510261ms step_avg:141.78ms step:3610/4578 train_loss:3.4978 train_time:510591ms step_avg:141.83ms step:3611/4578 train_loss:3.2497 train_time:510726ms step_avg:141.83ms step:3612/4578 train_loss:3.3534 train_time:510866ms step_avg:141.83ms step:3613/4578 train_loss:3.4178 train_time:511005ms step_avg:141.83ms step:3614/4578 train_loss:3.6260 train_time:511144ms step_avg:141.83ms step:3615/4578 train_loss:3.6453 train_time:511288ms step_avg:141.83ms step:3616/4578 train_loss:3.2933 train_time:511422ms step_avg:141.83ms step:3617/4578 train_loss:3.3827 train_time:511565ms step_avg:141.83ms step:3618/4578 train_loss:3.3517 train_time:511716ms step_avg:141.83ms step:3619/4578 train_loss:3.4745 train_time:511860ms step_avg:141.83ms step:3620/4578 train_loss:3.4117 train_time:511987ms step_avg:141.82ms step:3621/4578 train_loss:3.2382 train_time:512128ms step_avg:141.82ms step:3622/4578 train_loss:3.3839 train_time:512275ms step_avg:141.83ms step:3623/4578 train_loss:3.3821 train_time:512409ms step_avg:141.82ms step:3624/4578 train_loss:3.3190 train_time:512551ms step_avg:141.82ms step:3625/4578 train_loss:3.4455 train_time:512697ms step_avg:141.82ms step:3625/4578 val_loss:3.3708 train_time:512749ms step_avg:141.84ms step:3626/4578 train_loss:3.5169 train_time:512844ms step_avg:141.83ms step:3627/4578 train_loss:3.5156 train_time:512993ms step_avg:141.83ms step:3628/4578 train_loss:3.3997 train_time:513130ms step_avg:141.83ms step:3629/4578 train_loss:3.5570 train_time:513269ms step_avg:141.83ms step:3630/4578 train_loss:3.3827 train_time:513411ms step_avg:141.83ms step:3631/4578 train_loss:3.3832 train_time:513547ms step_avg:141.82ms step:3632/4578 train_loss:3.4613 train_time:513686ms step_avg:141.82ms step:3633/4578 train_loss:3.4466 train_time:513836ms step_avg:141.83ms step:3634/4578 train_loss:3.3735 train_time:513972ms step_avg:141.82ms step:3635/4578 train_loss:3.3669 train_time:514117ms step_avg:141.83ms step:3636/4578 train_loss:3.4189 train_time:514254ms step_avg:141.82ms step:3637/4578 train_loss:3.5967 train_time:514393ms step_avg:141.82ms step:3638/4578 train_loss:3.3954 train_time:514533ms step_avg:141.82ms step:3639/4578 train_loss:3.3592 train_time:514681ms step_avg:141.82ms step:3640/4578 train_loss:3.3538 train_time:514814ms step_avg:141.82ms step:3641/4578 train_loss:3.6435 train_time:514956ms step_avg:141.82ms step:3642/4578 train_loss:3.3762 train_time:515097ms step_avg:141.82ms step:3643/4578 train_loss:3.4260 train_time:515237ms step_avg:141.82ms step:3644/4578 train_loss:3.3942 train_time:515384ms step_avg:141.82ms step:3645/4578 train_loss:3.3138 train_time:515515ms step_avg:141.82ms step:3646/4578 train_loss:3.5076 train_time:515660ms step_avg:141.82ms step:3647/4578 train_loss:3.2880 train_time:515808ms step_avg:141.82ms step:3648/4578 train_loss:3.3697 train_time:515936ms step_avg:141.82ms step:3649/4578 train_loss:3.4241 train_time:516082ms step_avg:141.82ms step:3650/4578 train_loss:3.3888 train_time:516229ms step_avg:141.82ms step:3651/4578 train_loss:3.4274 train_time:516359ms step_avg:141.82ms step:3652/4578 train_loss:3.4779 train_time:516498ms step_avg:141.82ms step:3653/4578 train_loss:3.3030 train_time:516638ms step_avg:141.82ms step:3654/4578 train_loss:3.4129 train_time:516779ms step_avg:141.82ms step:3655/4578 train_loss:3.4346 train_time:516917ms step_avg:141.82ms step:3656/4578 train_loss:4.1155 train_time:517058ms step_avg:141.82ms step:3657/4578 train_loss:3.4831 train_time:517198ms step_avg:141.81ms step:3658/4578 train_loss:3.4013 train_time:517339ms step_avg:141.81ms step:3659/4578 train_loss:3.4009 train_time:517478ms step_avg:141.81ms step:3660/4578 train_loss:3.2833 train_time:517620ms step_avg:141.81ms step:3661/4578 train_loss:3.4015 train_time:517766ms step_avg:141.81ms step:3662/4578 train_loss:3.2788 train_time:517910ms step_avg:141.82ms step:3663/4578 train_loss:3.4278 train_time:518040ms step_avg:141.81ms step:3664/4578 train_loss:3.4395 train_time:518182ms step_avg:141.81ms step:3665/4578 train_loss:3.2807 train_time:518320ms step_avg:141.81ms step:3666/4578 train_loss:3.2144 train_time:518460ms step_avg:141.81ms step:3667/4578 train_loss:3.6404 train_time:518600ms step_avg:141.81ms step:3668/4578 train_loss:3.4213 train_time:518739ms step_avg:141.81ms step:3669/4578 train_loss:3.4465 train_time:518883ms step_avg:141.81ms step:3670/4578 train_loss:3.3662 train_time:519019ms step_avg:141.81ms step:3671/4578 train_loss:3.4350 train_time:519159ms step_avg:141.81ms step:3672/4578 train_loss:3.3252 train_time:519299ms step_avg:141.81ms step:3673/4578 train_loss:3.3369 train_time:519438ms step_avg:141.81ms step:3674/4578 train_loss:3.2212 train_time:519578ms step_avg:141.81ms step:3675/4578 train_loss:3.3129 train_time:519717ms step_avg:141.81ms step:3676/4578 train_loss:3.4717 train_time:519857ms step_avg:141.81ms step:3677/4578 train_loss:3.2726 train_time:519997ms step_avg:141.80ms step:3678/4578 train_loss:3.4389 train_time:520137ms step_avg:141.80ms step:3679/4578 train_loss:3.4188 train_time:520277ms step_avg:141.80ms step:3680/4578 train_loss:3.3271 train_time:520417ms step_avg:141.80ms step:3681/4578 train_loss:3.3891 train_time:520557ms step_avg:141.80ms step:3682/4578 train_loss:3.4534 train_time:520696ms step_avg:141.80ms step:3683/4578 train_loss:3.5475 train_time:520838ms step_avg:141.80ms step:3684/4578 train_loss:3.2934 train_time:520977ms step_avg:141.80ms step:3685/4578 train_loss:3.3705 train_time:521118ms step_avg:141.80ms step:3686/4578 train_loss:3.5198 train_time:521258ms step_avg:141.80ms step:3687/4578 train_loss:3.3083 train_time:521398ms step_avg:141.80ms step:3688/4578 train_loss:3.5066 train_time:521538ms step_avg:141.80ms step:3689/4578 train_loss:3.2391 train_time:521679ms step_avg:141.80ms step:3690/4578 train_loss:3.3345 train_time:521818ms step_avg:141.80ms step:3691/4578 train_loss:3.4625 train_time:521957ms step_avg:141.80ms step:3692/4578 train_loss:3.2556 train_time:522098ms step_avg:141.80ms step:3693/4578 train_loss:3.4047 train_time:522239ms step_avg:141.80ms step:3694/4578 train_loss:3.3822 train_time:522379ms step_avg:141.80ms step:3695/4578 train_loss:3.3861 train_time:522520ms step_avg:141.80ms step:3696/4578 train_loss:3.4124 train_time:522659ms step_avg:141.80ms step:3697/4578 train_loss:3.2578 train_time:522799ms step_avg:141.80ms step:3698/4578 train_loss:3.3965 train_time:522939ms step_avg:141.79ms step:3699/4578 train_loss:3.4027 train_time:523079ms step_avg:141.79ms step:3700/4578 train_loss:3.3811 train_time:523220ms step_avg:141.79ms step:3701/4578 train_loss:3.4416 train_time:523360ms step_avg:141.79ms step:3702/4578 train_loss:3.4142 train_time:523505ms step_avg:141.79ms step:3703/4578 train_loss:3.3278 train_time:523641ms step_avg:141.79ms step:3704/4578 train_loss:3.2992 train_time:523785ms step_avg:141.79ms step:3705/4578 train_loss:3.4465 train_time:523920ms step_avg:141.79ms step:3706/4578 train_loss:3.4431 train_time:524059ms step_avg:141.79ms step:3707/4578 train_loss:3.4520 train_time:524199ms step_avg:141.79ms step:3708/4578 train_loss:3.4024 train_time:524340ms step_avg:141.79ms step:3709/4578 train_loss:3.2760 train_time:524480ms step_avg:141.79ms step:3710/4578 train_loss:3.5826 train_time:524620ms step_avg:141.79ms step:3711/4578 train_loss:3.1611 train_time:524760ms step_avg:141.79ms step:3712/4578 train_loss:3.4351 train_time:524899ms step_avg:141.79ms step:3713/4578 train_loss:3.3212 train_time:525040ms step_avg:141.79ms step:3714/4578 train_loss:3.3546 train_time:525180ms step_avg:141.79ms step:3715/4578 train_loss:3.7284 train_time:525320ms step_avg:141.79ms step:3716/4578 train_loss:3.5686 train_time:525461ms step_avg:141.79ms step:3717/4578 train_loss:3.8462 train_time:525600ms step_avg:141.79ms step:3718/4578 train_loss:3.3498 train_time:525742ms step_avg:141.79ms step:3719/4578 train_loss:3.2623 train_time:525886ms step_avg:141.79ms step:3720/4578 train_loss:3.5228 train_time:526027ms step_avg:141.79ms step:3721/4578 train_loss:3.2850 train_time:526168ms step_avg:141.79ms step:3722/4578 train_loss:3.3813 train_time:526303ms step_avg:141.78ms step:3723/4578 train_loss:3.2508 train_time:526446ms step_avg:141.78ms step:3724/4578 train_loss:3.2408 train_time:526586ms step_avg:141.78ms step:3725/4578 train_loss:3.3715 train_time:526724ms step_avg:141.78ms step:3726/4578 train_loss:3.3195 train_time:526863ms step_avg:141.78ms step:3727/4578 train_loss:3.5832 train_time:527003ms step_avg:141.78ms step:3728/4578 train_loss:3.3168 train_time:527144ms step_avg:141.78ms step:3729/4578 train_loss:3.3091 train_time:527283ms step_avg:141.78ms step:3730/4578 train_loss:3.6542 train_time:527423ms step_avg:141.78ms step:3731/4578 train_loss:3.4220 train_time:527562ms step_avg:141.78ms step:3732/4578 train_loss:3.3303 train_time:527702ms step_avg:141.78ms step:3733/4578 train_loss:3.2381 train_time:527842ms step_avg:141.78ms step:3734/4578 train_loss:3.4326 train_time:527983ms step_avg:141.78ms step:3735/4578 train_loss:3.2940 train_time:528125ms step_avg:141.78ms step:3736/4578 train_loss:3.3968 train_time:528265ms step_avg:141.78ms step:3737/4578 train_loss:3.3033 train_time:528405ms step_avg:141.78ms step:3738/4578 train_loss:3.4040 train_time:528546ms step_avg:141.78ms step:3739/4578 train_loss:3.3075 train_time:528687ms step_avg:141.78ms step:3740/4578 train_loss:3.3478 train_time:528829ms step_avg:141.78ms step:3741/4578 train_loss:3.6373 train_time:528970ms step_avg:141.78ms step:3742/4578 train_loss:3.3027 train_time:529114ms step_avg:141.78ms step:3743/4578 train_loss:3.3543 train_time:529252ms step_avg:141.78ms step:3744/4578 train_loss:3.5720 train_time:529392ms step_avg:141.78ms step:3745/4578 train_loss:3.2823 train_time:529538ms step_avg:141.78ms step:3746/4578 train_loss:3.2241 train_time:529672ms step_avg:141.78ms step:3747/4578 train_loss:3.4060 train_time:529813ms step_avg:141.77ms step:3748/4578 train_loss:3.2546 train_time:529954ms step_avg:141.77ms step:3749/4578 train_loss:3.3019 train_time:530094ms step_avg:141.77ms step:3750/4578 train_loss:3.4953 train_time:530234ms step_avg:141.77ms step:3750/4578 val_loss:3.3547 train_time:530289ms step_avg:141.79ms step:3751/4578 train_loss:3.3852 train_time:530386ms step_avg:141.78ms step:3752/4578 train_loss:3.6192 train_time:530534ms step_avg:141.78ms step:3753/4578 train_loss:3.3313 train_time:530674ms step_avg:141.78ms step:3754/4578 train_loss:3.3406 train_time:530812ms step_avg:141.78ms step:3755/4578 train_loss:3.2999 train_time:530951ms step_avg:141.78ms step:3756/4578 train_loss:3.4068 train_time:531088ms step_avg:141.77ms step:3757/4578 train_loss:3.3506 train_time:531227ms step_avg:141.77ms step:3758/4578 train_loss:3.3532 train_time:531367ms step_avg:141.77ms step:3759/4578 train_loss:3.5410 train_time:531517ms step_avg:141.78ms step:3760/4578 train_loss:3.4255 train_time:531654ms step_avg:141.77ms step:3761/4578 train_loss:3.5463 train_time:531795ms step_avg:141.77ms step:3762/4578 train_loss:3.3034 train_time:531935ms step_avg:141.77ms step:3763/4578 train_loss:3.3191 train_time:532073ms step_avg:141.77ms step:3764/4578 train_loss:3.4901 train_time:532213ms step_avg:141.77ms step:3765/4578 train_loss:3.2420 train_time:532353ms step_avg:141.77ms step:3766/4578 train_loss:3.3361 train_time:532495ms step_avg:141.77ms step:3767/4578 train_loss:3.4295 train_time:532636ms step_avg:141.77ms step:3768/4578 train_loss:3.2333 train_time:532778ms step_avg:141.77ms step:3769/4578 train_loss:3.5011 train_time:532918ms step_avg:141.77ms step:3770/4578 train_loss:3.3091 train_time:533059ms step_avg:141.77ms step:3771/4578 train_loss:3.1957 train_time:533207ms step_avg:141.77ms step:3772/4578 train_loss:3.4397 train_time:533344ms step_avg:141.77ms step:3773/4578 train_loss:3.3620 train_time:533479ms step_avg:141.77ms step:3774/4578 train_loss:3.3567 train_time:533621ms step_avg:141.77ms step:3775/4578 train_loss:3.3494 train_time:533768ms step_avg:141.77ms step:3776/4578 train_loss:3.4134 train_time:533907ms step_avg:141.77ms step:3777/4578 train_loss:3.2379 train_time:534052ms step_avg:141.77ms step:3778/4578 train_loss:3.3491 train_time:534185ms step_avg:141.77ms step:3779/4578 train_loss:3.4683 train_time:534325ms step_avg:141.77ms step:3780/4578 train_loss:3.4366 train_time:534604ms step_avg:141.80ms step:3781/4578 train_loss:3.4389 train_time:534739ms step_avg:141.80ms step:3782/4578 train_loss:3.3864 train_time:534878ms step_avg:141.80ms step:3783/4578 train_loss:3.3874 train_time:535017ms step_avg:141.80ms step:3784/4578 train_loss:3.3590 train_time:535157ms step_avg:141.80ms step:3785/4578 train_loss:3.2381 train_time:535295ms step_avg:141.80ms step:3786/4578 train_loss:3.3223 train_time:535433ms step_avg:141.80ms step:3787/4578 train_loss:3.3662 train_time:535577ms step_avg:141.80ms step:3788/4578 train_loss:3.3496 train_time:535719ms step_avg:141.80ms step:3789/4578 train_loss:3.3079 train_time:535860ms step_avg:141.80ms step:3790/4578 train_loss:3.3217 train_time:536000ms step_avg:141.80ms step:3791/4578 train_loss:3.2023 train_time:536141ms step_avg:141.80ms step:3792/4578 train_loss:3.4196 train_time:536280ms step_avg:141.80ms step:3793/4578 train_loss:3.4088 train_time:536421ms step_avg:141.80ms step:3794/4578 train_loss:3.3312 train_time:536564ms step_avg:141.80ms step:3795/4578 train_loss:3.2853 train_time:536707ms step_avg:141.80ms step:3796/4578 train_loss:3.1417 train_time:536847ms step_avg:141.80ms step:3797/4578 train_loss:3.3380 train_time:536986ms step_avg:141.80ms step:3798/4578 train_loss:3.3311 train_time:537126ms step_avg:141.80ms step:3799/4578 train_loss:3.3998 train_time:537265ms step_avg:141.80ms step:3800/4578 train_loss:3.3021 train_time:537559ms step_avg:141.84ms step:3801/4578 train_loss:3.2828 train_time:537700ms step_avg:141.84ms step:3802/4578 train_loss:3.2430 train_time:537838ms step_avg:141.83ms step:3803/4578 train_loss:3.5497 train_time:537972ms step_avg:141.83ms step:3804/4578 train_loss:3.4028 train_time:538111ms step_avg:141.83ms step:3805/4578 train_loss:3.2462 train_time:538248ms step_avg:141.83ms step:3806/4578 train_loss:3.4870 train_time:538387ms step_avg:141.83ms step:3807/4578 train_loss:3.4532 train_time:538532ms step_avg:141.83ms step:3808/4578 train_loss:3.3289 train_time:538673ms step_avg:141.83ms step:3809/4578 train_loss:3.3897 train_time:538814ms step_avg:141.83ms step:3810/4578 train_loss:3.2945 train_time:538954ms step_avg:141.83ms step:3811/4578 train_loss:3.3674 train_time:539092ms step_avg:141.83ms step:3812/4578 train_loss:3.3408 train_time:539231ms step_avg:141.83ms step:3813/4578 train_loss:3.3816 train_time:539370ms step_avg:141.83ms step:3814/4578 train_loss:3.3787 train_time:539512ms step_avg:141.83ms step:3815/4578 train_loss:3.3552 train_time:539653ms step_avg:141.83ms step:3816/4578 train_loss:3.6037 train_time:539794ms step_avg:141.83ms step:3817/4578 train_loss:3.1990 train_time:539934ms step_avg:141.83ms step:3818/4578 train_loss:3.3810 train_time:540076ms step_avg:141.83ms step:3819/4578 train_loss:3.3568 train_time:540214ms step_avg:141.83ms step:3820/4578 train_loss:3.3343 train_time:540355ms step_avg:141.83ms step:3821/4578 train_loss:3.2730 train_time:540495ms step_avg:141.82ms step:3822/4578 train_loss:3.4432 train_time:540636ms step_avg:141.82ms step:3823/4578 train_loss:3.1738 train_time:540776ms step_avg:141.82ms step:3824/4578 train_loss:3.2880 train_time:540916ms step_avg:141.82ms step:3825/4578 train_loss:3.3305 train_time:541057ms step_avg:141.82ms step:3826/4578 train_loss:3.4845 train_time:541199ms step_avg:141.82ms step:3827/4578 train_loss:3.4227 train_time:541336ms step_avg:141.82ms step:3828/4578 train_loss:3.7957 train_time:541478ms step_avg:141.82ms step:3829/4578 train_loss:3.4174 train_time:541616ms step_avg:141.82ms step:3830/4578 train_loss:3.2407 train_time:541757ms step_avg:141.82ms step:3831/4578 train_loss:3.2874 train_time:541898ms step_avg:141.82ms step:3832/4578 train_loss:3.4978 train_time:542040ms step_avg:141.82ms step:3833/4578 train_loss:3.3130 train_time:542181ms step_avg:141.82ms step:3834/4578 train_loss:3.4337 train_time:542321ms step_avg:141.82ms step:3835/4578 train_loss:3.3688 train_time:542469ms step_avg:141.82ms step:3836/4578 train_loss:3.1679 train_time:542608ms step_avg:141.82ms step:3837/4578 train_loss:3.4535 train_time:542744ms step_avg:141.82ms step:3838/4578 train_loss:3.4408 train_time:542884ms step_avg:141.82ms step:3839/4578 train_loss:3.3853 train_time:543027ms step_avg:141.82ms step:3840/4578 train_loss:3.4364 train_time:543166ms step_avg:141.82ms step:3841/4578 train_loss:3.5689 train_time:543306ms step_avg:141.82ms step:3842/4578 train_loss:3.3142 train_time:543446ms step_avg:141.82ms step:3843/4578 train_loss:3.3808 train_time:543585ms step_avg:141.82ms step:3844/4578 train_loss:3.5094 train_time:543726ms step_avg:141.82ms step:3845/4578 train_loss:3.3127 train_time:543868ms step_avg:141.82ms step:3846/4578 train_loss:3.1894 train_time:544007ms step_avg:141.82ms step:3847/4578 train_loss:3.4209 train_time:544147ms step_avg:141.82ms step:3848/4578 train_loss:3.3456 train_time:544291ms step_avg:141.82ms step:3849/4578 train_loss:3.3910 train_time:544428ms step_avg:141.82ms step:3850/4578 train_loss:3.2780 train_time:544569ms step_avg:141.81ms step:3851/4578 train_loss:3.2689 train_time:544712ms step_avg:141.82ms step:3852/4578 train_loss:3.4236 train_time:544848ms step_avg:141.81ms step:3853/4578 train_loss:3.2807 train_time:544988ms step_avg:141.81ms step:3854/4578 train_loss:3.2513 train_time:545130ms step_avg:141.81ms step:3855/4578 train_loss:3.3258 train_time:545268ms step_avg:141.81ms step:3856/4578 train_loss:3.3647 train_time:545408ms step_avg:141.81ms step:3857/4578 train_loss:3.3391 train_time:545550ms step_avg:141.81ms step:3858/4578 train_loss:3.3699 train_time:545688ms step_avg:141.81ms step:3859/4578 train_loss:3.3421 train_time:545828ms step_avg:141.81ms step:3860/4578 train_loss:3.3105 train_time:545968ms step_avg:141.81ms step:3861/4578 train_loss:3.5105 train_time:546110ms step_avg:141.81ms step:3862/4578 train_loss:3.3137 train_time:546249ms step_avg:141.81ms step:3863/4578 train_loss:3.4392 train_time:546388ms step_avg:141.81ms step:3864/4578 train_loss:3.3948 train_time:546528ms step_avg:141.81ms step:3865/4578 train_loss:3.4393 train_time:546668ms step_avg:141.81ms step:3866/4578 train_loss:3.4118 train_time:546809ms step_avg:141.81ms step:3867/4578 train_loss:3.3535 train_time:546952ms step_avg:141.81ms step:3868/4578 train_loss:3.4350 train_time:547089ms step_avg:141.81ms step:3869/4578 train_loss:3.5923 train_time:547229ms step_avg:141.81ms step:3870/4578 train_loss:3.4410 train_time:547369ms step_avg:141.81ms step:3871/4578 train_loss:3.3378 train_time:547509ms step_avg:141.80ms step:3872/4578 train_loss:3.4801 train_time:547649ms step_avg:141.80ms step:3873/4578 train_loss:3.3776 train_time:547789ms step_avg:141.80ms step:3874/4578 train_loss:3.3232 train_time:547929ms step_avg:141.80ms step:3875/4578 train_loss:3.4261 train_time:548068ms step_avg:141.80ms step:3875/4578 val_loss:3.3404 train_time:548124ms step_avg:141.82ms step:3876/4578 train_loss:3.9449 train_time:548217ms step_avg:141.80ms step:3877/4578 train_loss:3.3638 train_time:548365ms step_avg:141.81ms step:3878/4578 train_loss:3.3593 train_time:548506ms step_avg:141.81ms step:3879/4578 train_loss:3.3387 train_time:548644ms step_avg:141.81ms step:3880/4578 train_loss:3.5417 train_time:548785ms step_avg:141.80ms step:3881/4578 train_loss:3.3462 train_time:548922ms step_avg:141.80ms step:3882/4578 train_loss:3.4184 train_time:549059ms step_avg:141.80ms step:3883/4578 train_loss:3.4627 train_time:549214ms step_avg:141.81ms step:3884/4578 train_loss:3.2820 train_time:549346ms step_avg:141.80ms step:3885/4578 train_loss:3.2790 train_time:549486ms step_avg:141.80ms step:3886/4578 train_loss:3.3139 train_time:549633ms step_avg:141.80ms step:3887/4578 train_loss:3.3510 train_time:549763ms step_avg:141.80ms step:3888/4578 train_loss:3.5242 train_time:549903ms step_avg:141.80ms step:3889/4578 train_loss:3.3722 train_time:550041ms step_avg:141.80ms step:3890/4578 train_loss:3.3107 train_time:550184ms step_avg:141.80ms step:3891/4578 train_loss:3.4530 train_time:550336ms step_avg:141.80ms step:3892/4578 train_loss:3.3125 train_time:550471ms step_avg:141.80ms step:3893/4578 train_loss:3.5658 train_time:550610ms step_avg:141.80ms step:3894/4578 train_loss:3.3010 train_time:550750ms step_avg:141.80ms step:3895/4578 train_loss:3.3070 train_time:550889ms step_avg:141.80ms step:3896/4578 train_loss:3.3878 train_time:551030ms step_avg:141.80ms step:3897/4578 train_loss:3.6264 train_time:551169ms step_avg:141.80ms step:3898/4578 train_loss:3.1930 train_time:551310ms step_avg:141.80ms step:3899/4578 train_loss:3.3141 train_time:551452ms step_avg:141.80ms step:3900/4578 train_loss:3.4503 train_time:551593ms step_avg:141.80ms step:3901/4578 train_loss:3.3831 train_time:551734ms step_avg:141.80ms step:3902/4578 train_loss:3.4129 train_time:551874ms step_avg:141.80ms step:3903/4578 train_loss:3.7029 train_time:552015ms step_avg:141.80ms step:3904/4578 train_loss:3.3028 train_time:552155ms step_avg:141.80ms step:3905/4578 train_loss:3.3143 train_time:552296ms step_avg:141.80ms step:3906/4578 train_loss:3.2742 train_time:552437ms step_avg:141.80ms step:3907/4578 train_loss:3.4334 train_time:552578ms step_avg:141.80ms step:3908/4578 train_loss:3.4460 train_time:552719ms step_avg:141.80ms step:3909/4578 train_loss:3.4400 train_time:552860ms step_avg:141.80ms step:3910/4578 train_loss:3.3861 train_time:553000ms step_avg:141.79ms step:3911/4578 train_loss:3.3162 train_time:553139ms step_avg:141.79ms step:3912/4578 train_loss:3.3377 train_time:553280ms step_avg:141.79ms step:3913/4578 train_loss:3.3275 train_time:553420ms step_avg:141.79ms step:3914/4578 train_loss:3.4549 train_time:553560ms step_avg:141.79ms step:3915/4578 train_loss:3.2902 train_time:553700ms step_avg:141.79ms step:3916/4578 train_loss:3.2648 train_time:553840ms step_avg:141.79ms step:3917/4578 train_loss:3.2658 train_time:553981ms step_avg:141.79ms step:3918/4578 train_loss:3.3746 train_time:554121ms step_avg:141.79ms step:3919/4578 train_loss:3.4968 train_time:554260ms step_avg:141.79ms step:3920/4578 train_loss:3.2757 train_time:554402ms step_avg:141.79ms step:3921/4578 train_loss:3.2606 train_time:554542ms step_avg:141.79ms step:3922/4578 train_loss:3.3337 train_time:554683ms step_avg:141.79ms step:3923/4578 train_loss:3.3236 train_time:554822ms step_avg:141.79ms step:3924/4578 train_loss:3.3423 train_time:554963ms step_avg:141.79ms step:3925/4578 train_loss:3.4215 train_time:555101ms step_avg:141.79ms step:3926/4578 train_loss:3.3828 train_time:555241ms step_avg:141.79ms step:3927/4578 train_loss:3.4791 train_time:555382ms step_avg:141.79ms step:3928/4578 train_loss:3.3670 train_time:555522ms step_avg:141.79ms step:3929/4578 train_loss:3.2296 train_time:555661ms step_avg:141.79ms step:3930/4578 train_loss:3.5521 train_time:555801ms step_avg:141.79ms step:3931/4578 train_loss:3.3355 train_time:555942ms step_avg:141.79ms step:3932/4578 train_loss:3.3810 train_time:556082ms step_avg:141.79ms step:3933/4578 train_loss:3.4179 train_time:556222ms step_avg:141.78ms step:3934/4578 train_loss:3.2898 train_time:556361ms step_avg:141.78ms step:3935/4578 train_loss:3.4200 train_time:556503ms step_avg:141.78ms step:3936/4578 train_loss:3.4310 train_time:556642ms step_avg:141.78ms step:3937/4578 train_loss:3.3573 train_time:556783ms step_avg:141.78ms step:3938/4578 train_loss:3.4105 train_time:556922ms step_avg:141.78ms step:3939/4578 train_loss:3.3297 train_time:557063ms step_avg:141.78ms step:3940/4578 train_loss:3.0917 train_time:557203ms step_avg:141.78ms step:3941/4578 train_loss:3.3083 train_time:557343ms step_avg:141.78ms step:3942/4578 train_loss:3.4144 train_time:557484ms step_avg:141.78ms step:3943/4578 train_loss:3.5116 train_time:557624ms step_avg:141.78ms step:3944/4578 train_loss:3.5458 train_time:557764ms step_avg:141.78ms step:3945/4578 train_loss:3.3958 train_time:557904ms step_avg:141.78ms step:3946/4578 train_loss:3.2976 train_time:558043ms step_avg:141.78ms step:3947/4578 train_loss:3.3161 train_time:558183ms step_avg:141.78ms step:3948/4578 train_loss:3.3934 train_time:558323ms step_avg:141.78ms step:3949/4578 train_loss:3.1940 train_time:558464ms step_avg:141.78ms step:3950/4578 train_loss:3.4019 train_time:558603ms step_avg:141.78ms step:3951/4578 train_loss:3.3386 train_time:558742ms step_avg:141.78ms step:3952/4578 train_loss:3.1378 train_time:558883ms step_avg:141.78ms step:3953/4578 train_loss:3.2188 train_time:559023ms step_avg:141.78ms step:3954/4578 train_loss:3.4833 train_time:559162ms step_avg:141.78ms step:3955/4578 train_loss:3.3893 train_time:559303ms step_avg:141.78ms step:3956/4578 train_loss:3.3289 train_time:559442ms step_avg:141.77ms step:3957/4578 train_loss:3.3835 train_time:559583ms step_avg:141.77ms step:3958/4578 train_loss:3.1047 train_time:559723ms step_avg:141.77ms step:3959/4578 train_loss:3.3883 train_time:559863ms step_avg:141.77ms step:3960/4578 train_loss:3.3465 train_time:560002ms step_avg:141.77ms step:3961/4578 train_loss:3.3140 train_time:560142ms step_avg:141.77ms step:3962/4578 train_loss:3.3312 train_time:560282ms step_avg:141.77ms step:3963/4578 train_loss:3.3534 train_time:560422ms step_avg:141.77ms step:3964/4578 train_loss:3.3855 train_time:560561ms step_avg:141.77ms step:3965/4578 train_loss:3.2324 train_time:560704ms step_avg:141.77ms step:3966/4578 train_loss:3.3503 train_time:560841ms step_avg:141.77ms step:3967/4578 train_loss:3.4224 train_time:560981ms step_avg:141.77ms step:3968/4578 train_loss:3.3352 train_time:561121ms step_avg:141.77ms step:3969/4578 train_loss:3.4288 train_time:561400ms step_avg:141.80ms step:3970/4578 train_loss:3.3136 train_time:561538ms step_avg:141.80ms step:3971/4578 train_loss:3.5026 train_time:561677ms step_avg:141.80ms step:3972/4578 train_loss:3.4299 train_time:561816ms step_avg:141.80ms step:3973/4578 train_loss:3.3823 train_time:561955ms step_avg:141.80ms step:3974/4578 train_loss:3.2709 train_time:562094ms step_avg:141.80ms step:3975/4578 train_loss:3.3241 train_time:562233ms step_avg:141.80ms step:3976/4578 train_loss:3.3924 train_time:562378ms step_avg:141.80ms step:3977/4578 train_loss:3.3015 train_time:562521ms step_avg:141.80ms step:3978/4578 train_loss:3.3670 train_time:562660ms step_avg:141.80ms step:3979/4578 train_loss:3.4381 train_time:562800ms step_avg:141.80ms step:3980/4578 train_loss:3.3818 train_time:562940ms step_avg:141.80ms step:3981/4578 train_loss:3.3886 train_time:563079ms step_avg:141.80ms step:3982/4578 train_loss:3.5806 train_time:563219ms step_avg:141.80ms step:3983/4578 train_loss:3.3263 train_time:563361ms step_avg:141.80ms step:3984/4578 train_loss:3.3998 train_time:563504ms step_avg:141.80ms step:3985/4578 train_loss:3.3181 train_time:563643ms step_avg:141.80ms step:3986/4578 train_loss:3.2545 train_time:563782ms step_avg:141.80ms step:3987/4578 train_loss:3.2931 train_time:563923ms step_avg:141.80ms step:3988/4578 train_loss:3.3108 train_time:564061ms step_avg:141.80ms step:3989/4578 train_loss:3.0478 train_time:564201ms step_avg:141.79ms step:3990/4578 train_loss:3.3628 train_time:564501ms step_avg:141.83ms step:3991/4578 train_loss:3.3374 train_time:564637ms step_avg:141.83ms step:3992/4578 train_loss:3.1825 train_time:564776ms step_avg:141.83ms step:3993/4578 train_loss:3.2968 train_time:564917ms step_avg:141.83ms step:3994/4578 train_loss:3.4887 train_time:565053ms step_avg:141.83ms step:3995/4578 train_loss:3.3100 train_time:565193ms step_avg:141.83ms step:3996/4578 train_loss:3.2258 train_time:565333ms step_avg:141.83ms step:3997/4578 train_loss:3.3713 train_time:565481ms step_avg:141.83ms step:3998/4578 train_loss:3.2939 train_time:565622ms step_avg:141.83ms step:3999/4578 train_loss:3.2489 train_time:565761ms step_avg:141.83ms step:4000/4578 train_loss:3.3213 train_time:565902ms step_avg:141.83ms step:4000/4578 val_loss:3.3250 train_time:565956ms step_avg:141.84ms step:4001/4578 train_loss:3.4448 train_time:566055ms step_avg:141.83ms step:4002/4578 train_loss:3.5117 train_time:566198ms step_avg:141.83ms step:4003/4578 train_loss:3.1849 train_time:566338ms step_avg:141.83ms step:4004/4578 train_loss:3.3877 train_time:566477ms step_avg:141.83ms step:4005/4578 train_loss:3.2784 train_time:566615ms step_avg:141.83ms step:4006/4578 train_loss:3.3360 train_time:566754ms step_avg:141.83ms step:4007/4578 train_loss:3.3165 train_time:566893ms step_avg:141.83ms step:4008/4578 train_loss:3.5165 train_time:567035ms step_avg:141.83ms step:4009/4578 train_loss:3.1064 train_time:567177ms step_avg:141.83ms step:4010/4578 train_loss:3.3053 train_time:567319ms step_avg:141.83ms step:4011/4578 train_loss:3.2817 train_time:567460ms step_avg:141.83ms step:4012/4578 train_loss:3.2578 train_time:567601ms step_avg:141.83ms step:4013/4578 train_loss:3.4269 train_time:567741ms step_avg:141.83ms step:4014/4578 train_loss:3.2911 train_time:567882ms step_avg:141.83ms step:4015/4578 train_loss:3.3949 train_time:568020ms step_avg:141.83ms step:4016/4578 train_loss:3.4761 train_time:568162ms step_avg:141.83ms step:4017/4578 train_loss:3.4685 train_time:568304ms step_avg:141.83ms step:4018/4578 train_loss:3.2158 train_time:568444ms step_avg:141.83ms step:4019/4578 train_loss:3.3565 train_time:568585ms step_avg:141.83ms step:4020/4578 train_loss:3.2786 train_time:568727ms step_avg:141.83ms step:4021/4578 train_loss:3.5502 train_time:568867ms step_avg:141.83ms step:4022/4578 train_loss:3.4252 train_time:569008ms step_avg:141.83ms step:4023/4578 train_loss:3.4036 train_time:569147ms step_avg:141.83ms step:4024/4578 train_loss:3.3764 train_time:569289ms step_avg:141.83ms step:4025/4578 train_loss:3.4054 train_time:569429ms step_avg:141.83ms step:4026/4578 train_loss:3.1667 train_time:569569ms step_avg:141.82ms step:4027/4578 train_loss:3.3792 train_time:569709ms step_avg:141.82ms step:4028/4578 train_loss:3.3318 train_time:569848ms step_avg:141.82ms step:4029/4578 train_loss:3.2188 train_time:569997ms step_avg:141.83ms step:4030/4578 train_loss:3.2432 train_time:570130ms step_avg:141.82ms step:4031/4578 train_loss:3.3003 train_time:570274ms step_avg:141.82ms step:4032/4578 train_loss:3.3896 train_time:570415ms step_avg:141.82ms step:4033/4578 train_loss:3.3543 train_time:570557ms step_avg:141.82ms step:4034/4578 train_loss:3.3339 train_time:570691ms step_avg:141.82ms step:4035/4578 train_loss:3.3095 train_time:570830ms step_avg:141.82ms step:4036/4578 train_loss:3.2516 train_time:570970ms step_avg:141.82ms step:4037/4578 train_loss:3.4266 train_time:571110ms step_avg:141.82ms step:4038/4578 train_loss:3.3637 train_time:571249ms step_avg:141.82ms step:4039/4578 train_loss:3.3487 train_time:571393ms step_avg:141.82ms step:4040/4578 train_loss:3.3377 train_time:571531ms step_avg:141.82ms step:4041/4578 train_loss:3.3926 train_time:571670ms step_avg:141.82ms step:4042/4578 train_loss:3.5768 train_time:571810ms step_avg:141.82ms step:4043/4578 train_loss:3.4693 train_time:571949ms step_avg:141.82ms step:4044/4578 train_loss:3.2577 train_time:572090ms step_avg:141.82ms step:4045/4578 train_loss:3.4243 train_time:572230ms step_avg:141.82ms step:4046/4578 train_loss:3.1340 train_time:572371ms step_avg:141.82ms step:4047/4578 train_loss:3.3982 train_time:572511ms step_avg:141.82ms step:4048/4578 train_loss:3.4755 train_time:572650ms step_avg:141.82ms step:4049/4578 train_loss:3.3507 train_time:572790ms step_avg:141.81ms step:4050/4578 train_loss:3.2791 train_time:572930ms step_avg:141.81ms step:4051/4578 train_loss:3.3192 train_time:573070ms step_avg:141.81ms step:4052/4578 train_loss:3.2539 train_time:573212ms step_avg:141.81ms step:4053/4578 train_loss:3.4603 train_time:573351ms step_avg:141.81ms step:4054/4578 train_loss:3.3249 train_time:573494ms step_avg:141.81ms step:4055/4578 train_loss:3.4059 train_time:573638ms step_avg:141.81ms step:4056/4578 train_loss:3.3728 train_time:573771ms step_avg:141.81ms step:4057/4578 train_loss:3.3561 train_time:573911ms step_avg:141.81ms step:4058/4578 train_loss:3.2273 train_time:574050ms step_avg:141.81ms step:4059/4578 train_loss:3.3798 train_time:574191ms step_avg:141.81ms step:4060/4578 train_loss:3.2319 train_time:574329ms step_avg:141.81ms step:4061/4578 train_loss:3.3232 train_time:574469ms step_avg:141.81ms step:4062/4578 train_loss:3.4380 train_time:574610ms step_avg:141.81ms step:4063/4578 train_loss:3.5965 train_time:574749ms step_avg:141.81ms step:4064/4578 train_loss:3.0040 train_time:574891ms step_avg:141.81ms step:4065/4578 train_loss:3.3609 train_time:575029ms step_avg:141.81ms step:4066/4578 train_loss:3.2470 train_time:575169ms step_avg:141.81ms step:4067/4578 train_loss:3.4019 train_time:575311ms step_avg:141.81ms step:4068/4578 train_loss:3.4061 train_time:575450ms step_avg:141.81ms step:4069/4578 train_loss:3.2087 train_time:575597ms step_avg:141.81ms step:4070/4578 train_loss:3.3811 train_time:575730ms step_avg:141.81ms step:4071/4578 train_loss:3.1875 train_time:575871ms step_avg:141.81ms step:4072/4578 train_loss:3.3794 train_time:576012ms step_avg:141.81ms step:4073/4578 train_loss:3.4844 train_time:576149ms step_avg:141.80ms step:4074/4578 train_loss:3.3992 train_time:576293ms step_avg:141.80ms step:4075/4578 train_loss:3.3222 train_time:576430ms step_avg:141.80ms step:4076/4578 train_loss:3.3193 train_time:576570ms step_avg:141.80ms step:4077/4578 train_loss:3.1835 train_time:576711ms step_avg:141.80ms step:4078/4578 train_loss:3.3576 train_time:576850ms step_avg:141.80ms step:4079/4578 train_loss:3.3762 train_time:576991ms step_avg:141.80ms step:4080/4578 train_loss:3.1648 train_time:577131ms step_avg:141.80ms step:4081/4578 train_loss:3.3408 train_time:577273ms step_avg:141.80ms step:4082/4578 train_loss:3.2899 train_time:577411ms step_avg:141.80ms step:4083/4578 train_loss:3.3419 train_time:577550ms step_avg:141.80ms step:4084/4578 train_loss:3.3463 train_time:577691ms step_avg:141.80ms step:4085/4578 train_loss:3.3745 train_time:577833ms step_avg:141.80ms step:4086/4578 train_loss:3.3340 train_time:577972ms step_avg:141.80ms step:4087/4578 train_loss:3.3162 train_time:578112ms step_avg:141.80ms step:4088/4578 train_loss:3.4205 train_time:578250ms step_avg:141.80ms step:4089/4578 train_loss:3.2614 train_time:578391ms step_avg:141.80ms step:4090/4578 train_loss:3.2904 train_time:578531ms step_avg:141.80ms step:4091/4578 train_loss:3.2967 train_time:578678ms step_avg:141.80ms step:4092/4578 train_loss:3.2523 train_time:578810ms step_avg:141.80ms step:4093/4578 train_loss:3.2580 train_time:578953ms step_avg:141.80ms step:4094/4578 train_loss:3.4310 train_time:579100ms step_avg:141.80ms step:4095/4578 train_loss:3.4192 train_time:579229ms step_avg:141.79ms step:4096/4578 train_loss:3.3319 train_time:579369ms step_avg:141.79ms step:4097/4578 train_loss:3.3899 train_time:579512ms step_avg:141.79ms step:4098/4578 train_loss:3.1644 train_time:579653ms step_avg:141.79ms step:4099/4578 train_loss:3.3039 train_time:579799ms step_avg:141.79ms step:4100/4578 train_loss:3.2761 train_time:579930ms step_avg:141.79ms step:4101/4578 train_loss:3.0827 train_time:580070ms step_avg:141.79ms step:4102/4578 train_loss:3.3531 train_time:580210ms step_avg:141.79ms step:4103/4578 train_loss:3.3180 train_time:580350ms step_avg:141.79ms step:4104/4578 train_loss:3.1540 train_time:580490ms step_avg:141.79ms step:4105/4578 train_loss:3.2380 train_time:580632ms step_avg:141.79ms step:4106/4578 train_loss:3.4072 train_time:580770ms step_avg:141.79ms step:4107/4578 train_loss:3.4472 train_time:580915ms step_avg:141.79ms step:4108/4578 train_loss:3.3418 train_time:581049ms step_avg:141.79ms step:4109/4578 train_loss:3.4304 train_time:581191ms step_avg:141.79ms step:4110/4578 train_loss:3.4287 train_time:581330ms step_avg:141.79ms step:4111/4578 train_loss:3.5735 train_time:581470ms step_avg:141.79ms step:4112/4578 train_loss:3.2430 train_time:581610ms step_avg:141.79ms step:4113/4578 train_loss:3.3796 train_time:581749ms step_avg:141.79ms step:4114/4578 train_loss:3.2783 train_time:581890ms step_avg:141.79ms step:4115/4578 train_loss:3.3880 train_time:582032ms step_avg:141.79ms step:4116/4578 train_loss:3.3829 train_time:582172ms step_avg:141.79ms step:4117/4578 train_loss:3.6197 train_time:582312ms step_avg:141.79ms step:4118/4578 train_loss:3.1968 train_time:582451ms step_avg:141.78ms step:4119/4578 train_loss:3.3689 train_time:582592ms step_avg:141.78ms step:4120/4578 train_loss:3.2707 train_time:582732ms step_avg:141.78ms step:4121/4578 train_loss:3.3822 train_time:582873ms step_avg:141.78ms step:4122/4578 train_loss:3.3741 train_time:583021ms step_avg:141.79ms step:4123/4578 train_loss:3.3650 train_time:583154ms step_avg:141.78ms step:4124/4578 train_loss:3.2044 train_time:583293ms step_avg:141.78ms step:4125/4578 train_loss:3.2101 train_time:583434ms step_avg:141.78ms step:4125/4578 val_loss:3.3119 train_time:583489ms step_avg:141.80ms step:4126/4578 train_loss:3.3384 train_time:583591ms step_avg:141.79ms step:4127/4578 train_loss:3.2717 train_time:583728ms step_avg:141.78ms step:4128/4578 train_loss:3.3386 train_time:583866ms step_avg:141.78ms step:4129/4578 train_loss:3.3331 train_time:584003ms step_avg:141.78ms step:4130/4578 train_loss:3.1221 train_time:584142ms step_avg:141.78ms step:4131/4578 train_loss:3.4052 train_time:584285ms step_avg:141.78ms step:4132/4578 train_loss:3.3650 train_time:584420ms step_avg:141.78ms step:4133/4578 train_loss:3.2896 train_time:584565ms step_avg:141.78ms step:4134/4578 train_loss:3.4949 train_time:584712ms step_avg:141.78ms step:4135/4578 train_loss:3.3193 train_time:584850ms step_avg:141.78ms step:4136/4578 train_loss:3.3026 train_time:584990ms step_avg:141.78ms step:4137/4578 train_loss:3.4470 train_time:585131ms step_avg:141.78ms step:4138/4578 train_loss:3.2913 train_time:585269ms step_avg:141.78ms step:4139/4578 train_loss:3.3460 train_time:585407ms step_avg:141.78ms step:4140/4578 train_loss:3.4413 train_time:585548ms step_avg:141.78ms step:4141/4578 train_loss:3.4607 train_time:585690ms step_avg:141.78ms step:4142/4578 train_loss:3.4195 train_time:585831ms step_avg:141.78ms step:4143/4578 train_loss:3.4119 train_time:585972ms step_avg:141.78ms step:4144/4578 train_loss:3.3164 train_time:586112ms step_avg:141.78ms step:4145/4578 train_loss:3.2826 train_time:586259ms step_avg:141.78ms step:4146/4578 train_loss:3.3846 train_time:586391ms step_avg:141.78ms step:4147/4578 train_loss:2.9461 train_time:586531ms step_avg:141.78ms step:4148/4578 train_loss:3.3064 train_time:586673ms step_avg:141.78ms step:4149/4578 train_loss:3.3413 train_time:586812ms step_avg:141.78ms step:4150/4578 train_loss:3.1577 train_time:586951ms step_avg:141.78ms step:4151/4578 train_loss:3.1898 train_time:587093ms step_avg:141.78ms step:4152/4578 train_loss:3.2248 train_time:587233ms step_avg:141.78ms step:4153/4578 train_loss:3.2843 train_time:587373ms step_avg:141.77ms step:4154/4578 train_loss:3.3428 train_time:587513ms step_avg:141.77ms step:4155/4578 train_loss:3.4487 train_time:587653ms step_avg:141.77ms step:4156/4578 train_loss:3.2534 train_time:587795ms step_avg:141.77ms step:4157/4578 train_loss:3.2240 train_time:587936ms step_avg:141.77ms step:4158/4578 train_loss:3.3211 train_time:588215ms step_avg:141.81ms step:4159/4578 train_loss:3.3298 train_time:588350ms step_avg:141.81ms step:4160/4578 train_loss:3.2485 train_time:588489ms step_avg:141.80ms step:4161/4578 train_loss:3.3260 train_time:588626ms step_avg:141.80ms step:4162/4578 train_loss:3.2644 train_time:588765ms step_avg:141.80ms step:4163/4578 train_loss:3.4926 train_time:588904ms step_avg:141.80ms step:4164/4578 train_loss:3.1796 train_time:589043ms step_avg:141.80ms step:4165/4578 train_loss:3.2851 train_time:589187ms step_avg:141.80ms step:4166/4578 train_loss:3.2574 train_time:589330ms step_avg:141.80ms step:4167/4578 train_loss:3.3146 train_time:589470ms step_avg:141.80ms step:4168/4578 train_loss:3.2960 train_time:589612ms step_avg:141.80ms step:4169/4578 train_loss:3.3189 train_time:589750ms step_avg:141.80ms step:4170/4578 train_loss:3.1675 train_time:589887ms step_avg:141.80ms step:4171/4578 train_loss:3.2701 train_time:590036ms step_avg:141.80ms step:4172/4578 train_loss:3.3852 train_time:590169ms step_avg:141.80ms step:4173/4578 train_loss:3.4598 train_time:590309ms step_avg:141.80ms step:4174/4578 train_loss:3.8399 train_time:590451ms step_avg:141.80ms step:4175/4578 train_loss:3.2697 train_time:590591ms step_avg:141.80ms step:4176/4578 train_loss:3.4275 train_time:590730ms step_avg:141.80ms step:4177/4578 train_loss:3.2253 train_time:590870ms step_avg:141.80ms step:4178/4578 train_loss:3.2485 train_time:591009ms step_avg:141.80ms step:4179/4578 train_loss:3.4147 train_time:591160ms step_avg:141.80ms step:4180/4578 train_loss:3.3599 train_time:591444ms step_avg:141.83ms step:4181/4578 train_loss:3.3457 train_time:591580ms step_avg:141.83ms step:4182/4578 train_loss:3.3507 train_time:591720ms step_avg:141.83ms step:4183/4578 train_loss:3.3785 train_time:591857ms step_avg:141.83ms step:4184/4578 train_loss:3.8171 train_time:591996ms step_avg:141.83ms step:4185/4578 train_loss:3.3226 train_time:592135ms step_avg:141.83ms step:4186/4578 train_loss:3.3744 train_time:592276ms step_avg:141.83ms step:4187/4578 train_loss:3.4203 train_time:592420ms step_avg:141.83ms step:4188/4578 train_loss:3.4141 train_time:592567ms step_avg:141.83ms step:4189/4578 train_loss:3.0386 train_time:592702ms step_avg:141.83ms step:4190/4578 train_loss:3.4098 train_time:592840ms step_avg:141.83ms step:4191/4578 train_loss:3.4175 train_time:592981ms step_avg:141.83ms step:4192/4578 train_loss:3.3802 train_time:593120ms step_avg:141.83ms step:4193/4578 train_loss:3.3259 train_time:593260ms step_avg:141.83ms step:4194/4578 train_loss:3.3409 train_time:593402ms step_avg:141.83ms step:4195/4578 train_loss:3.3163 train_time:593544ms step_avg:141.83ms step:4196/4578 train_loss:3.2644 train_time:593684ms step_avg:141.83ms step:4197/4578 train_loss:3.6127 train_time:593823ms step_avg:141.83ms step:4198/4578 train_loss:3.0600 train_time:593963ms step_avg:141.82ms step:4199/4578 train_loss:3.4704 train_time:594102ms step_avg:141.82ms step:4200/4578 train_loss:3.3366 train_time:594242ms step_avg:141.82ms step:4201/4578 train_loss:3.2285 train_time:594384ms step_avg:141.82ms step:4202/4578 train_loss:3.3652 train_time:594524ms step_avg:141.82ms step:4203/4578 train_loss:3.2274 train_time:594664ms step_avg:141.82ms step:4204/4578 train_loss:3.2350 train_time:594804ms step_avg:141.82ms step:4205/4578 train_loss:3.2514 train_time:594944ms step_avg:141.82ms step:4206/4578 train_loss:3.2462 train_time:595083ms step_avg:141.82ms step:4207/4578 train_loss:3.7088 train_time:595223ms step_avg:141.82ms step:4208/4578 train_loss:3.2853 train_time:595364ms step_avg:141.82ms step:4209/4578 train_loss:3.4101 train_time:595505ms step_avg:141.82ms step:4210/4578 train_loss:3.3015 train_time:595645ms step_avg:141.82ms step:4211/4578 train_loss:3.6958 train_time:595785ms step_avg:141.82ms step:4212/4578 train_loss:3.3559 train_time:595928ms step_avg:141.82ms step:4213/4578 train_loss:3.3547 train_time:596066ms step_avg:141.82ms step:4214/4578 train_loss:3.2338 train_time:596205ms step_avg:141.82ms step:4215/4578 train_loss:3.2815 train_time:596346ms step_avg:141.82ms step:4216/4578 train_loss:3.3646 train_time:596486ms step_avg:141.82ms step:4217/4578 train_loss:3.2277 train_time:596626ms step_avg:141.82ms step:4218/4578 train_loss:3.2860 train_time:596768ms step_avg:141.82ms step:4219/4578 train_loss:3.3371 train_time:596906ms step_avg:141.82ms step:4220/4578 train_loss:3.1536 train_time:597047ms step_avg:141.82ms step:4221/4578 train_loss:3.3105 train_time:597186ms step_avg:141.82ms step:4222/4578 train_loss:3.3467 train_time:597325ms step_avg:141.82ms step:4223/4578 train_loss:3.3099 train_time:597467ms step_avg:141.82ms step:4224/4578 train_loss:3.5161 train_time:597607ms step_avg:141.81ms step:4225/4578 train_loss:3.4008 train_time:597748ms step_avg:141.81ms step:4226/4578 train_loss:3.4356 train_time:597887ms step_avg:141.81ms step:4227/4578 train_loss:3.2294 train_time:598031ms step_avg:141.81ms step:4228/4578 train_loss:3.2986 train_time:598169ms step_avg:141.81ms step:4229/4578 train_loss:3.3282 train_time:598309ms step_avg:141.81ms step:4230/4578 train_loss:3.2388 train_time:598449ms step_avg:141.81ms step:4231/4578 train_loss:3.4305 train_time:598589ms step_avg:141.81ms step:4232/4578 train_loss:3.4321 train_time:598728ms step_avg:141.81ms step:4233/4578 train_loss:3.4280 train_time:598869ms step_avg:141.81ms step:4234/4578 train_loss:3.5354 train_time:599009ms step_avg:141.81ms step:4235/4578 train_loss:3.3780 train_time:599150ms step_avg:141.81ms step:4236/4578 train_loss:3.3365 train_time:599290ms step_avg:141.81ms step:4237/4578 train_loss:3.1771 train_time:599430ms step_avg:141.81ms step:4238/4578 train_loss:3.3865 train_time:599570ms step_avg:141.81ms step:4239/4578 train_loss:3.3094 train_time:599710ms step_avg:141.81ms step:4240/4578 train_loss:3.2190 train_time:599850ms step_avg:141.81ms step:4241/4578 train_loss:3.2728 train_time:599992ms step_avg:141.81ms step:4242/4578 train_loss:3.2183 train_time:600131ms step_avg:141.81ms step:4243/4578 train_loss:3.2920 train_time:600272ms step_avg:141.81ms step:4244/4578 train_loss:3.2193 train_time:600411ms step_avg:141.81ms step:4245/4578 train_loss:3.1269 train_time:600551ms step_avg:141.81ms step:4246/4578 train_loss:3.4400 train_time:600695ms step_avg:141.81ms step:4247/4578 train_loss:3.2371 train_time:600833ms step_avg:141.81ms step:4248/4578 train_loss:3.1641 train_time:600972ms step_avg:141.81ms step:4249/4578 train_loss:3.3717 train_time:601112ms step_avg:141.81ms step:4250/4578 train_loss:3.6514 train_time:601252ms step_avg:141.80ms step:4250/4578 val_loss:3.2995 train_time:601307ms step_avg:141.82ms step:4251/4578 train_loss:3.2812 train_time:601404ms step_avg:141.81ms step:4252/4578 train_loss:3.5323 train_time:601551ms step_avg:141.81ms step:4253/4578 train_loss:3.3672 train_time:601691ms step_avg:141.81ms step:4254/4578 train_loss:3.1848 train_time:601832ms step_avg:141.81ms step:4255/4578 train_loss:3.2611 train_time:601968ms step_avg:141.81ms step:4256/4578 train_loss:3.1888 train_time:602106ms step_avg:141.81ms step:4257/4578 train_loss:3.4226 train_time:602245ms step_avg:141.80ms step:4258/4578 train_loss:3.3194 train_time:602388ms step_avg:141.81ms step:4259/4578 train_loss:3.3615 train_time:602531ms step_avg:141.81ms step:4260/4578 train_loss:3.1787 train_time:602673ms step_avg:141.81ms step:4261/4578 train_loss:3.4850 train_time:602813ms step_avg:141.80ms step:4262/4578 train_loss:3.3131 train_time:602953ms step_avg:141.80ms step:4263/4578 train_loss:3.3165 train_time:603093ms step_avg:141.80ms step:4264/4578 train_loss:3.3779 train_time:603232ms step_avg:141.80ms step:4265/4578 train_loss:3.3177 train_time:603374ms step_avg:141.80ms step:4266/4578 train_loss:3.3148 train_time:603516ms step_avg:141.80ms step:4267/4578 train_loss:3.4389 train_time:603658ms step_avg:141.80ms step:4268/4578 train_loss:3.2644 train_time:603797ms step_avg:141.80ms step:4269/4578 train_loss:3.8090 train_time:603937ms step_avg:141.80ms step:4270/4578 train_loss:3.2564 train_time:604077ms step_avg:141.80ms step:4271/4578 train_loss:3.3442 train_time:604217ms step_avg:141.80ms step:4272/4578 train_loss:3.2807 train_time:604360ms step_avg:141.80ms step:4273/4578 train_loss:3.4898 train_time:604498ms step_avg:141.80ms step:4274/4578 train_loss:3.4046 train_time:604637ms step_avg:141.80ms step:4275/4578 train_loss:3.2607 train_time:604778ms step_avg:141.80ms step:4276/4578 train_loss:3.3180 train_time:604918ms step_avg:141.80ms step:4277/4578 train_loss:3.2432 train_time:605058ms step_avg:141.80ms step:4278/4578 train_loss:3.2812 train_time:605202ms step_avg:141.80ms step:4279/4578 train_loss:3.2881 train_time:605338ms step_avg:141.80ms step:4280/4578 train_loss:3.3415 train_time:605477ms step_avg:141.80ms step:4281/4578 train_loss:3.3350 train_time:605618ms step_avg:141.80ms step:4282/4578 train_loss:3.3470 train_time:605760ms step_avg:141.80ms step:4283/4578 train_loss:3.2785 train_time:605902ms step_avg:141.80ms step:4284/4578 train_loss:3.3244 train_time:606043ms step_avg:141.80ms step:4285/4578 train_loss:3.3949 train_time:606184ms step_avg:141.80ms step:4286/4578 train_loss:3.3382 train_time:606325ms step_avg:141.80ms step:4287/4578 train_loss:3.2150 train_time:606465ms step_avg:141.80ms step:4288/4578 train_loss:3.2552 train_time:606607ms step_avg:141.80ms step:4289/4578 train_loss:3.3613 train_time:606749ms step_avg:141.80ms step:4290/4578 train_loss:3.3168 train_time:606890ms step_avg:141.80ms step:4291/4578 train_loss:3.2198 train_time:607030ms step_avg:141.80ms step:4292/4578 train_loss:3.2481 train_time:607170ms step_avg:141.80ms step:4293/4578 train_loss:3.3203 train_time:607311ms step_avg:141.80ms step:4294/4578 train_loss:3.0897 train_time:607450ms step_avg:141.80ms step:4295/4578 train_loss:3.4537 train_time:607591ms step_avg:141.79ms step:4296/4578 train_loss:3.3449 train_time:607731ms step_avg:141.79ms step:4297/4578 train_loss:3.2968 train_time:607872ms step_avg:141.79ms step:4298/4578 train_loss:3.4506 train_time:608012ms step_avg:141.79ms step:4299/4578 train_loss:3.3844 train_time:608153ms step_avg:141.79ms step:4300/4578 train_loss:3.2173 train_time:608292ms step_avg:141.79ms step:4301/4578 train_loss:3.2114 train_time:608431ms step_avg:141.79ms step:4302/4578 train_loss:3.3548 train_time:608573ms step_avg:141.79ms step:4303/4578 train_loss:3.2073 train_time:608714ms step_avg:141.79ms step:4304/4578 train_loss:3.3440 train_time:608853ms step_avg:141.79ms step:4305/4578 train_loss:3.4293 train_time:608994ms step_avg:141.79ms step:4306/4578 train_loss:3.1797 train_time:609134ms step_avg:141.79ms step:4307/4578 train_loss:3.6966 train_time:609274ms step_avg:141.79ms step:4308/4578 train_loss:3.3047 train_time:609413ms step_avg:141.79ms step:4309/4578 train_loss:3.2368 train_time:609554ms step_avg:141.79ms step:4310/4578 train_loss:3.2567 train_time:609694ms step_avg:141.79ms step:4311/4578 train_loss:3.5449 train_time:609833ms step_avg:141.79ms step:4312/4578 train_loss:3.3763 train_time:609974ms step_avg:141.79ms step:4313/4578 train_loss:3.2327 train_time:610114ms step_avg:141.79ms step:4314/4578 train_loss:3.4390 train_time:610257ms step_avg:141.79ms step:4315/4578 train_loss:3.3708 train_time:610394ms step_avg:141.79ms step:4316/4578 train_loss:3.2830 train_time:610533ms step_avg:141.79ms step:4317/4578 train_loss:3.3291 train_time:610674ms step_avg:141.79ms step:4318/4578 train_loss:3.2724 train_time:610814ms step_avg:141.79ms step:4319/4578 train_loss:3.3979 train_time:610953ms step_avg:141.79ms step:4320/4578 train_loss:3.4384 train_time:611095ms step_avg:141.79ms step:4321/4578 train_loss:3.2637 train_time:611236ms step_avg:141.79ms step:4322/4578 train_loss:3.4369 train_time:611375ms step_avg:141.78ms step:4323/4578 train_loss:3.2981 train_time:611514ms step_avg:141.78ms step:4324/4578 train_loss:3.2155 train_time:611656ms step_avg:141.78ms step:4325/4578 train_loss:3.1639 train_time:611797ms step_avg:141.78ms step:4326/4578 train_loss:3.2602 train_time:611936ms step_avg:141.78ms step:4327/4578 train_loss:3.1411 train_time:612076ms step_avg:141.78ms step:4328/4578 train_loss:3.2668 train_time:612216ms step_avg:141.78ms step:4329/4578 train_loss:3.2870 train_time:612356ms step_avg:141.78ms step:4330/4578 train_loss:3.2326 train_time:612497ms step_avg:141.78ms step:4331/4578 train_loss:3.4893 train_time:612636ms step_avg:141.78ms step:4332/4578 train_loss:3.2885 train_time:612777ms step_avg:141.78ms step:4333/4578 train_loss:3.4095 train_time:612917ms step_avg:141.78ms step:4334/4578 train_loss:3.7723 train_time:613056ms step_avg:141.78ms step:4335/4578 train_loss:3.3053 train_time:613198ms step_avg:141.78ms step:4336/4578 train_loss:3.4055 train_time:613340ms step_avg:141.78ms step:4337/4578 train_loss:3.2997 train_time:613477ms step_avg:141.78ms step:4338/4578 train_loss:3.1855 train_time:613617ms step_avg:141.78ms step:4339/4578 train_loss:3.3236 train_time:613760ms step_avg:141.78ms step:4340/4578 train_loss:3.2264 train_time:613899ms step_avg:141.78ms step:4341/4578 train_loss:3.3196 train_time:614039ms step_avg:141.78ms step:4342/4578 train_loss:3.3386 train_time:614180ms step_avg:141.78ms step:4343/4578 train_loss:3.3391 train_time:614319ms step_avg:141.78ms step:4344/4578 train_loss:3.3337 train_time:614460ms step_avg:141.78ms step:4345/4578 train_loss:3.9662 train_time:614600ms step_avg:141.78ms step:4346/4578 train_loss:3.4057 train_time:614739ms step_avg:141.78ms step:4347/4578 train_loss:3.2006 train_time:615044ms step_avg:141.81ms step:4348/4578 train_loss:3.3391 train_time:615178ms step_avg:141.81ms step:4349/4578 train_loss:3.2884 train_time:615317ms step_avg:141.81ms step:4350/4578 train_loss:3.2077 train_time:615456ms step_avg:141.81ms step:4351/4578 train_loss:3.3653 train_time:615595ms step_avg:141.81ms step:4352/4578 train_loss:3.3182 train_time:615740ms step_avg:141.81ms step:4353/4578 train_loss:3.3960 train_time:615873ms step_avg:141.81ms step:4354/4578 train_loss:3.4355 train_time:616017ms step_avg:141.81ms step:4355/4578 train_loss:3.2403 train_time:616158ms step_avg:141.81ms step:4356/4578 train_loss:3.2100 train_time:616306ms step_avg:141.81ms step:4357/4578 train_loss:3.3308 train_time:616439ms step_avg:141.81ms step:4358/4578 train_loss:3.2791 train_time:616579ms step_avg:141.81ms step:4359/4578 train_loss:3.4723 train_time:616717ms step_avg:141.81ms step:4360/4578 train_loss:3.3267 train_time:616858ms step_avg:141.81ms step:4361/4578 train_loss:3.4113 train_time:616999ms step_avg:141.81ms step:4362/4578 train_loss:3.5621 train_time:617139ms step_avg:141.81ms step:4363/4578 train_loss:3.3517 train_time:617279ms step_avg:141.81ms step:4364/4578 train_loss:3.3298 train_time:617419ms step_avg:141.80ms step:4365/4578 train_loss:3.5308 train_time:617558ms step_avg:141.80ms step:4366/4578 train_loss:3.4202 train_time:617700ms step_avg:141.80ms step:4367/4578 train_loss:3.2439 train_time:617836ms step_avg:141.80ms step:4368/4578 train_loss:3.2662 train_time:617980ms step_avg:141.80ms step:4369/4578 train_loss:3.3558 train_time:618118ms step_avg:141.80ms step:4370/4578 train_loss:3.3652 train_time:618484ms step_avg:141.85ms step:4371/4578 train_loss:3.5076 train_time:618615ms step_avg:141.85ms step:4372/4578 train_loss:3.2051 train_time:618753ms step_avg:141.85ms step:4373/4578 train_loss:3.1922 train_time:618892ms step_avg:141.85ms step:4374/4578 train_loss:3.3467 train_time:619029ms step_avg:141.85ms step:4375/4578 train_loss:3.3758 train_time:619169ms step_avg:141.85ms step:4375/4578 val_loss:3.2876 train_time:619224ms step_avg:141.86ms step:4376/4578 train_loss:3.4451 train_time:619322ms step_avg:141.85ms step:4377/4578 train_loss:3.2288 train_time:619465ms step_avg:141.85ms step:4378/4578 train_loss:3.2793 train_time:619604ms step_avg:141.85ms step:4379/4578 train_loss:3.2930 train_time:619744ms step_avg:141.85ms step:4380/4578 train_loss:3.3235 train_time:619884ms step_avg:141.85ms step:4381/4578 train_loss:3.2256 train_time:620023ms step_avg:141.85ms step:4382/4578 train_loss:3.4689 train_time:620164ms step_avg:141.85ms step:4383/4578 train_loss:3.3886 train_time:620307ms step_avg:141.85ms step:4384/4578 train_loss:3.3794 train_time:620449ms step_avg:141.85ms step:4385/4578 train_loss:3.2593 train_time:620590ms step_avg:141.85ms step:4386/4578 train_loss:3.3826 train_time:620734ms step_avg:141.85ms step:4387/4578 train_loss:3.2762 train_time:620869ms step_avg:141.85ms step:4388/4578 train_loss:3.4044 train_time:621009ms step_avg:141.85ms step:4389/4578 train_loss:3.2351 train_time:621149ms step_avg:141.85ms step:4390/4578 train_loss:3.3545 train_time:621290ms step_avg:141.85ms step:4391/4578 train_loss:3.3792 train_time:621438ms step_avg:141.85ms step:4392/4578 train_loss:3.1986 train_time:621575ms step_avg:141.85ms step:4393/4578 train_loss:3.8745 train_time:621713ms step_avg:141.85ms step:4394/4578 train_loss:3.2719 train_time:621853ms step_avg:141.85ms step:4395/4578 train_loss:3.4553 train_time:621994ms step_avg:141.85ms step:4396/4578 train_loss:3.2295 train_time:622134ms step_avg:141.85ms step:4397/4578 train_loss:3.3560 train_time:622273ms step_avg:141.84ms step:4398/4578 train_loss:3.1622 train_time:622415ms step_avg:141.84ms step:4399/4578 train_loss:3.3717 train_time:622553ms step_avg:141.84ms step:4400/4578 train_loss:3.2060 train_time:622693ms step_avg:141.84ms step:4401/4578 train_loss:3.2787 train_time:622833ms step_avg:141.84ms step:4402/4578 train_loss:3.3433 train_time:622973ms step_avg:141.84ms step:4403/4578 train_loss:3.1867 train_time:623112ms step_avg:141.84ms step:4404/4578 train_loss:3.2387 train_time:623254ms step_avg:141.84ms step:4405/4578 train_loss:3.4284 train_time:623394ms step_avg:141.84ms step:4406/4578 train_loss:3.2864 train_time:623534ms step_avg:141.84ms step:4407/4578 train_loss:3.3086 train_time:623674ms step_avg:141.84ms step:4408/4578 train_loss:3.2798 train_time:623813ms step_avg:141.84ms step:4409/4578 train_loss:3.3437 train_time:623954ms step_avg:141.84ms step:4410/4578 train_loss:3.3403 train_time:624094ms step_avg:141.84ms step:4411/4578 train_loss:3.4617 train_time:624234ms step_avg:141.84ms step:4412/4578 train_loss:3.2997 train_time:624375ms step_avg:141.84ms step:4413/4578 train_loss:3.3227 train_time:624516ms step_avg:141.84ms step:4414/4578 train_loss:3.3175 train_time:624656ms step_avg:141.84ms step:4415/4578 train_loss:3.3605 train_time:624796ms step_avg:141.84ms step:4416/4578 train_loss:3.3047 train_time:624943ms step_avg:141.84ms step:4417/4578 train_loss:3.3760 train_time:625075ms step_avg:141.84ms step:4418/4578 train_loss:3.3108 train_time:625215ms step_avg:141.84ms step:4419/4578 train_loss:3.2064 train_time:625356ms step_avg:141.84ms step:4420/4578 train_loss:3.2700 train_time:625496ms step_avg:141.84ms step:4421/4578 train_loss:3.4963 train_time:625637ms step_avg:141.84ms step:4422/4578 train_loss:3.2978 train_time:625778ms step_avg:141.84ms step:4423/4578 train_loss:3.2178 train_time:625919ms step_avg:141.84ms step:4424/4578 train_loss:3.2417 train_time:626059ms step_avg:141.83ms step:4425/4578 train_loss:3.4206 train_time:626198ms step_avg:141.83ms step:4426/4578 train_loss:3.3632 train_time:626339ms step_avg:141.83ms step:4427/4578 train_loss:3.2812 train_time:626479ms step_avg:141.83ms step:4428/4578 train_loss:3.5023 train_time:626620ms step_avg:141.83ms step:4429/4578 train_loss:3.4025 train_time:626761ms step_avg:141.83ms step:4430/4578 train_loss:3.2046 train_time:626906ms step_avg:141.83ms step:4431/4578 train_loss:3.2054 train_time:627039ms step_avg:141.83ms step:4432/4578 train_loss:3.3314 train_time:627179ms step_avg:141.83ms step:4433/4578 train_loss:3.2117 train_time:627320ms step_avg:141.83ms step:4434/4578 train_loss:3.3295 train_time:627461ms step_avg:141.83ms step:4435/4578 train_loss:3.3875 train_time:627601ms step_avg:141.83ms step:4436/4578 train_loss:3.2777 train_time:627742ms step_avg:141.83ms step:4437/4578 train_loss:3.2042 train_time:627884ms step_avg:141.83ms step:4438/4578 train_loss:3.4427 train_time:628031ms step_avg:141.83ms step:4439/4578 train_loss:3.3672 train_time:628167ms step_avg:141.83ms step:4440/4578 train_loss:3.2713 train_time:628308ms step_avg:141.83ms step:4441/4578 train_loss:3.3746 train_time:628449ms step_avg:141.83ms step:4442/4578 train_loss:3.4029 train_time:628589ms step_avg:141.83ms step:4443/4578 train_loss:3.4551 train_time:628732ms step_avg:141.83ms step:4444/4578 train_loss:3.3414 train_time:628874ms step_avg:141.83ms step:4445/4578 train_loss:3.1624 train_time:629013ms step_avg:141.83ms step:4446/4578 train_loss:3.4365 train_time:629152ms step_avg:141.83ms step:4447/4578 train_loss:3.3186 train_time:629292ms step_avg:141.83ms step:4448/4578 train_loss:3.2055 train_time:629432ms step_avg:141.83ms step:4449/4578 train_loss:3.3437 train_time:629572ms step_avg:141.83ms step:4450/4578 train_loss:3.3237 train_time:629711ms step_avg:141.83ms step:4451/4578 train_loss:3.3428 train_time:629853ms step_avg:141.83ms step:4452/4578 train_loss:3.3744 train_time:629992ms step_avg:141.83ms step:4453/4578 train_loss:3.2419 train_time:630133ms step_avg:141.83ms step:4454/4578 train_loss:3.2633 train_time:630273ms step_avg:141.83ms step:4455/4578 train_loss:3.2681 train_time:630412ms step_avg:141.83ms step:4456/4578 train_loss:3.1605 train_time:630554ms step_avg:141.82ms step:4457/4578 train_loss:3.3725 train_time:630693ms step_avg:141.82ms step:4458/4578 train_loss:3.2402 train_time:630837ms step_avg:141.82ms step:4459/4578 train_loss:3.2141 train_time:630977ms step_avg:141.82ms step:4460/4578 train_loss:3.3408 train_time:631116ms step_avg:141.82ms step:4461/4578 train_loss:3.8295 train_time:631255ms step_avg:141.82ms step:4462/4578 train_loss:3.3356 train_time:631395ms step_avg:141.82ms step:4463/4578 train_loss:3.4350 train_time:631536ms step_avg:141.82ms step:4464/4578 train_loss:3.3448 train_time:631676ms step_avg:141.82ms step:4465/4578 train_loss:3.3211 train_time:631815ms step_avg:141.82ms step:4466/4578 train_loss:3.3924 train_time:631957ms step_avg:141.82ms step:4467/4578 train_loss:3.1974 train_time:632103ms step_avg:141.82ms step:4468/4578 train_loss:3.2461 train_time:632236ms step_avg:141.82ms step:4469/4578 train_loss:3.3759 train_time:632376ms step_avg:141.82ms step:4470/4578 train_loss:3.3536 train_time:632521ms step_avg:141.82ms step:4471/4578 train_loss:3.2779 train_time:632656ms step_avg:141.82ms step:4472/4578 train_loss:3.2412 train_time:632795ms step_avg:141.82ms step:4473/4578 train_loss:3.3217 train_time:632941ms step_avg:141.82ms step:4474/4578 train_loss:3.1651 train_time:633076ms step_avg:141.82ms step:4475/4578 train_loss:3.2483 train_time:633217ms step_avg:141.82ms step:4476/4578 train_loss:3.2654 train_time:633358ms step_avg:141.82ms step:4477/4578 train_loss:3.4367 train_time:633497ms step_avg:141.82ms step:4478/4578 train_loss:3.1672 train_time:633637ms step_avg:141.82ms step:4479/4578 train_loss:3.2949 train_time:633778ms step_avg:141.82ms step:4480/4578 train_loss:3.3294 train_time:633916ms step_avg:141.82ms step:4481/4578 train_loss:3.3009 train_time:634058ms step_avg:141.82ms step:4482/4578 train_loss:3.2916 train_time:634198ms step_avg:141.82ms step:4483/4578 train_loss:3.1133 train_time:634338ms step_avg:141.81ms step:4484/4578 train_loss:3.2604 train_time:634478ms step_avg:141.81ms step:4485/4578 train_loss:3.2111 train_time:634619ms step_avg:141.81ms step:4486/4578 train_loss:3.3447 train_time:634759ms step_avg:141.81ms step:4487/4578 train_loss:3.2200 train_time:634898ms step_avg:141.81ms step:4488/4578 train_loss:3.2932 train_time:635038ms step_avg:141.81ms step:4489/4578 train_loss:3.4328 train_time:635179ms step_avg:141.81ms step:4490/4578 train_loss:3.3940 train_time:635320ms step_avg:141.81ms step:4491/4578 train_loss:3.2747 train_time:635462ms step_avg:141.81ms step:4492/4578 train_loss:3.2352 train_time:635604ms step_avg:141.81ms step:4493/4578 train_loss:3.2763 train_time:635744ms step_avg:141.81ms step:4494/4578 train_loss:3.3080 train_time:635884ms step_avg:141.81ms step:4495/4578 train_loss:3.3078 train_time:636024ms step_avg:141.81ms step:4496/4578 train_loss:3.2506 train_time:636166ms step_avg:141.81ms step:4497/4578 train_loss:3.4108 train_time:636306ms step_avg:141.81ms step:4498/4578 train_loss:3.2866 train_time:636446ms step_avg:141.81ms step:4499/4578 train_loss:3.1351 train_time:636588ms step_avg:141.81ms step:4500/4578 train_loss:3.4400 train_time:636727ms step_avg:141.81ms step:4500/4578 val_loss:3.2790 train_time:636785ms step_avg:141.82ms step:4501/4578 train_loss:3.2332 train_time:636881ms step_avg:141.81ms step:4502/4578 train_loss:3.1998 train_time:637025ms step_avg:141.81ms step:4503/4578 train_loss:3.3866 train_time:637166ms step_avg:141.81ms step:4504/4578 train_loss:3.2647 train_time:637304ms step_avg:141.81ms step:4505/4578 train_loss:3.3633 train_time:637442ms step_avg:141.81ms step:4506/4578 train_loss:3.2876 train_time:637580ms step_avg:141.81ms step:4507/4578 train_loss:3.3536 train_time:637718ms step_avg:141.81ms step:4508/4578 train_loss:3.0707 train_time:637860ms step_avg:141.81ms step:4509/4578 train_loss:3.3567 train_time:638005ms step_avg:141.81ms step:4510/4578 train_loss:3.1957 train_time:638146ms step_avg:141.81ms step:4511/4578 train_loss:3.2690 train_time:638287ms step_avg:141.81ms step:4512/4578 train_loss:3.2111 train_time:638427ms step_avg:141.81ms step:4513/4578 train_loss:3.2178 train_time:638565ms step_avg:141.81ms step:4514/4578 train_loss:3.1641 train_time:638704ms step_avg:141.81ms step:4515/4578 train_loss:3.3065 train_time:638846ms step_avg:141.81ms step:4516/4578 train_loss:3.1756 train_time:638987ms step_avg:141.81ms step:4517/4578 train_loss:3.2760 train_time:639128ms step_avg:141.81ms step:4518/4578 train_loss:3.2792 train_time:639269ms step_avg:141.81ms step:4519/4578 train_loss:3.2893 train_time:639408ms step_avg:141.81ms step:4520/4578 train_loss:3.2105 train_time:639547ms step_avg:141.81ms step:4521/4578 train_loss:3.3986 train_time:639686ms step_avg:141.81ms step:4522/4578 train_loss:3.4610 train_time:639829ms step_avg:141.81ms step:4523/4578 train_loss:3.7967 train_time:639974ms step_avg:141.81ms step:4524/4578 train_loss:3.5508 train_time:640108ms step_avg:141.81ms step:4525/4578 train_loss:3.3076 train_time:640255ms step_avg:141.81ms step:4526/4578 train_loss:3.2593 train_time:640388ms step_avg:141.80ms step:4527/4578 train_loss:3.3214 train_time:640528ms step_avg:141.80ms step:4528/4578 train_loss:3.2861 train_time:640668ms step_avg:141.80ms step:4529/4578 train_loss:3.1911 train_time:640808ms step_avg:141.80ms step:4530/4578 train_loss:3.8893 train_time:640948ms step_avg:141.80ms step:4531/4578 train_loss:3.3653 train_time:641088ms step_avg:141.80ms step:4532/4578 train_loss:3.0924 train_time:641228ms step_avg:141.80ms step:4533/4578 train_loss:3.1984 train_time:641369ms step_avg:141.80ms step:4534/4578 train_loss:3.3128 train_time:641509ms step_avg:141.80ms step:4535/4578 train_loss:3.5363 train_time:641649ms step_avg:141.80ms step:4536/4578 train_loss:3.5322 train_time:641952ms step_avg:141.84ms step:4537/4578 train_loss:3.2648 train_time:642090ms step_avg:141.84ms step:4538/4578 train_loss:3.2638 train_time:642229ms step_avg:141.84ms step:4539/4578 train_loss:3.2967 train_time:642368ms step_avg:141.83ms step:4540/4578 train_loss:3.8622 train_time:642506ms step_avg:141.83ms step:4541/4578 train_loss:3.3509 train_time:642645ms step_avg:141.83ms step:4542/4578 train_loss:3.2603 train_time:642783ms step_avg:141.83ms step:4543/4578 train_loss:3.4281 train_time:642928ms step_avg:141.83ms step:4544/4578 train_loss:3.2020 train_time:643071ms step_avg:141.83ms step:4545/4578 train_loss:3.3194 train_time:643212ms step_avg:141.83ms step:4546/4578 train_loss:3.4915 train_time:643353ms step_avg:141.83ms step:4547/4578 train_loss:3.3501 train_time:643493ms step_avg:141.83ms step:4548/4578 train_loss:3.2950 train_time:643632ms step_avg:141.83ms step:4549/4578 train_loss:3.2980 train_time:643772ms step_avg:141.83ms step:4550/4578 train_loss:3.2415 train_time:643914ms step_avg:141.83ms step:4551/4578 train_loss:3.2158 train_time:644055ms step_avg:141.83ms step:4552/4578 train_loss:3.1943 train_time:644197ms step_avg:141.83ms step:4553/4578 train_loss:3.3093 train_time:644338ms step_avg:141.83ms step:4554/4578 train_loss:3.4898 train_time:644480ms step_avg:141.83ms step:4555/4578 train_loss:3.3777 train_time:644620ms step_avg:141.83ms step:4556/4578 train_loss:3.1361 train_time:644758ms step_avg:141.83ms step:4557/4578 train_loss:3.3372 train_time:644898ms step_avg:141.83ms step:4558/4578 train_loss:3.3435 train_time:645039ms step_avg:141.83ms step:4559/4578 train_loss:3.3344 train_time:645181ms step_avg:141.83ms step:4560/4578 train_loss:3.4304 train_time:645491ms step_avg:141.87ms step:4561/4578 train_loss:3.2629 train_time:645628ms step_avg:141.86ms step:4562/4578 train_loss:3.2670 train_time:645766ms step_avg:141.86ms step:4563/4578 train_loss:3.2898 train_time:645904ms step_avg:141.86ms step:4564/4578 train_loss:3.3409 train_time:646043ms step_avg:141.86ms step:4565/4578 train_loss:3.4216 train_time:646181ms step_avg:141.86ms step:4566/4578 train_loss:3.4807 train_time:646320ms step_avg:141.86ms step:4567/4578 train_loss:3.3341 train_time:646465ms step_avg:141.86ms step:4568/4578 train_loss:3.2026 train_time:646608ms step_avg:141.86ms step:4569/4578 train_loss:3.3066 train_time:646747ms step_avg:141.86ms step:4570/4578 train_loss:3.1914 train_time:646887ms step_avg:141.86ms step:4571/4578 train_loss:3.2219 train_time:647026ms step_avg:141.86ms step:4572/4578 train_loss:3.4253 train_time:647166ms step_avg:141.86ms step:4573/4578 train_loss:3.1306 train_time:647305ms step_avg:141.86ms step:4574/4578 train_loss:3.2136 train_time:647446ms step_avg:141.86ms step:4575/4578 train_loss:3.3390 train_time:647588ms step_avg:141.86ms step:4576/4578 train_loss:3.3697 train_time:647732ms step_avg:141.86ms step:4577/4578 train_loss:3.3235 train_time:647869ms step_avg:141.86ms step:4578/4578 train_loss:3.2914 train_time:648008ms step_avg:141.86ms step:4578/4578 val_loss:3.2762 train_time:648063ms step_avg:141.87ms