==================================================================================================== 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) self.lamb = 0.5 def forward(self, x, v1=None): B, T, C = x.size() # batch size, sequence length, embedding dimensionality (n_embd) q = self.c_q(x).view(B, T, self.n_head, self.head_dim) k = self.c_k(x).view(B, T, self.n_head, self.head_dim) v = self.c_v(x).view(B, T, self.n_head, self.head_dim) if v1 is None: v1 = v # This happens if we are in the first block. v needs to be accessed by subsequent blocks v = (1 - self.lamb) * v + self.lamb * v1.view_as(v) # @Grad62304977 cos, sin = self.rotary(q) q, k = F.rms_norm(q, (q.size(-1),)), F.rms_norm(k, (k.size(-1),)) # QK norm suggested by @Grad62304977 q, k = apply_rotary_emb(q, cos, sin), apply_rotary_emb(k, cos, sin) y = 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, v1 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, v1): x1, v1 = self.attn(F.rms_norm(x, (x.size(-1),)), v1) x = x + x1 x = x + self.mlp(F.rms_norm(x, (x.size(-1),))) return x, v1 # ----------------------------------------------------------------------------- # The main GPT-2 model @dataclass class GPTConfig: vocab_size : int = 50304 n_layer : int = 12 n_head : int = 6 # head dim 128 suggested by @Grad62304977 n_embd : int = 768 class GPT(nn.Module): def __init__(self, config): super().__init__() 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_() # @Grad62304977 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),)) # @Grad62304977 v1 = None for block in self.transformer.h: x, v1 = block(x, v1) 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 = 3200 # number of iterations to run warmup_iters : int = 0 warmdown_iters : int = 914 # 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) params = list(raw_model.transformer.h.parameters()) optimizer3 = Muon(params, 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: Wed Nov 6 19:46:00 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 33C P0 141W / 700W | 5304MiB / 81559MiB | 4% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 1 NVIDIA H100 80GB HBM3 Off | 00000000:2A:00.0 Off | 0 | | N/A 34C P0 130W / 700W | 5352MiB / 81559MiB | 4% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 2 NVIDIA H100 80GB HBM3 Off | 00000000:3A:00.0 Off | 0 | | N/A 34C P0 126W / 700W | 5352MiB / 81559MiB | 5% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 3 NVIDIA H100 80GB HBM3 Off | 00000000:5D:00.0 Off | 0 | | N/A 32C P0 136W / 700W | 5352MiB / 81559MiB | 5% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 4 NVIDIA H100 80GB HBM3 Off | 00000000:9A:00.0 Off | 0 | | N/A 34C P0 142W / 700W | 5352MiB / 81559MiB | 9% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 5 NVIDIA H100 80GB HBM3 Off | 00000000:AB:00.0 Off | 0 | | N/A 36C P0 142W / 700W | 5352MiB / 81559MiB | 3% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 6 NVIDIA H100 80GB HBM3 Off | 00000000:BA:00.0 Off | 0 | | N/A 35C P0 142W / 700W | 5352MiB / 81559MiB | 11% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ | 7 NVIDIA H100 80GB HBM3 Off | 00000000:DB:00.0 Off | 0 | | N/A 34C P0 146W / 700W | 5112MiB / 81559MiB | 0% Default | | | | Disabled | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 27499 C /usr/bin/python3 0MiB | | 1 N/A N/A 27500 C /usr/bin/python3 0MiB | | 2 N/A N/A 27501 C /usr/bin/python3 0MiB | | 3 N/A N/A 27502 C /usr/bin/python3 0MiB | | 4 N/A N/A 27503 C /usr/bin/python3 0MiB | | 5 N/A N/A 27504 C /usr/bin/python3 0MiB | | 6 N/A N/A 27505 C /usr/bin/python3 0MiB | | 7 N/A N/A 27506 C /usr/bin/python3 0MiB | +-----------------------------------------------------------------------------------------+ ==================================================================================================== step:0/3200 val_loss:10.8258 train_time:487ms step_avg:nanms step:1/3200 train_loss:10.8258 train_time:6565ms step_avg:nanms step:2/3200 train_loss:10.4281 train_time:6656ms step_avg:nanms step:3/3200 train_loss:9.9267 train_time:6792ms step_avg:nanms step:4/3200 train_loss:9.0276 train_time:6933ms step_avg:nanms step:5/3200 train_loss:8.0822 train_time:7073ms step_avg:nanms step:6/3200 train_loss:7.5352 train_time:7215ms step_avg:nanms step:7/3200 train_loss:7.0169 train_time:7357ms step_avg:nanms step:8/3200 train_loss:7.1894 train_time:7498ms step_avg:nanms step:9/3200 train_loss:6.8909 train_time:7646ms step_avg:nanms step:10/3200 train_loss:6.7493 train_time:7792ms step_avg:nanms step:11/3200 train_loss:6.7375 train_time:88ms step_avg:nanms step:12/3200 train_loss:6.6820 train_time:230ms step_avg:nanms step:13/3200 train_loss:6.5060 train_time:372ms step_avg:123.92ms step:14/3200 train_loss:6.4972 train_time:514ms step_avg:128.59ms step:15/3200 train_loss:6.4949 train_time:657ms step_avg:131.43ms step:16/3200 train_loss:6.4511 train_time:801ms step_avg:133.47ms step:17/3200 train_loss:6.4496 train_time:947ms step_avg:135.28ms step:18/3200 train_loss:6.4936 train_time:1089ms step_avg:136.08ms step:19/3200 train_loss:6.3283 train_time:1232ms step_avg:136.84ms step:20/3200 train_loss:6.3472 train_time:1375ms step_avg:137.47ms step:21/3200 train_loss:6.0540 train_time:1517ms step_avg:137.92ms step:22/3200 train_loss:6.3806 train_time:1660ms step_avg:138.33ms step:23/3200 train_loss:6.6051 train_time:1810ms step_avg:139.24ms step:24/3200 train_loss:6.2684 train_time:1948ms step_avg:139.17ms step:25/3200 train_loss:6.4253 train_time:2093ms step_avg:139.50ms step:26/3200 train_loss:6.1635 train_time:2237ms step_avg:139.79ms step:27/3200 train_loss:6.0617 train_time:2379ms step_avg:139.96ms step:28/3200 train_loss:6.2460 train_time:2526ms step_avg:140.31ms step:29/3200 train_loss:5.9026 train_time:2664ms step_avg:140.20ms step:30/3200 train_loss:6.1584 train_time:2809ms step_avg:140.46ms step:31/3200 train_loss:5.9918 train_time:2952ms step_avg:140.55ms step:32/3200 train_loss:5.9601 train_time:3095ms step_avg:140.68ms step:33/3200 train_loss:5.7996 train_time:3239ms step_avg:140.85ms step:34/3200 train_loss:6.0977 train_time:3383ms step_avg:140.95ms step:35/3200 train_loss:6.0087 train_time:3526ms step_avg:141.02ms step:36/3200 train_loss:6.1512 train_time:3669ms step_avg:141.12ms step:37/3200 train_loss:6.0768 train_time:3813ms step_avg:141.22ms step:38/3200 train_loss:5.9758 train_time:3958ms step_avg:141.35ms step:39/3200 train_loss:5.8602 train_time:4099ms step_avg:141.35ms step:40/3200 train_loss:5.8890 train_time:4242ms step_avg:141.40ms step:41/3200 train_loss:5.7995 train_time:4386ms step_avg:141.48ms step:42/3200 train_loss:5.8018 train_time:4530ms step_avg:141.57ms step:43/3200 train_loss:5.6983 train_time:4673ms step_avg:141.62ms step:44/3200 train_loss:5.7748 train_time:4824ms step_avg:141.87ms step:45/3200 train_loss:5.7682 train_time:4962ms step_avg:141.78ms step:46/3200 train_loss:5.9300 train_time:5104ms step_avg:141.78ms step:47/3200 train_loss:5.7245 train_time:5247ms step_avg:141.81ms step:48/3200 train_loss:5.5775 train_time:5390ms step_avg:141.84ms step:49/3200 train_loss:5.7771 train_time:5534ms step_avg:141.90ms step:50/3200 train_loss:5.6572 train_time:5678ms step_avg:141.94ms step:51/3200 train_loss:5.7957 train_time:5821ms step_avg:141.99ms step:52/3200 train_loss:5.6622 train_time:5966ms step_avg:142.04ms step:53/3200 train_loss:5.5080 train_time:6109ms step_avg:142.08ms step:54/3200 train_loss:5.6556 train_time:6252ms step_avg:142.10ms step:55/3200 train_loss:5.5447 train_time:6396ms step_avg:142.14ms step:56/3200 train_loss:5.8556 train_time:6539ms step_avg:142.16ms step:57/3200 train_loss:5.5045 train_time:6683ms step_avg:142.19ms step:58/3200 train_loss:5.3852 train_time:6827ms step_avg:142.22ms step:59/3200 train_loss:5.5273 train_time:6970ms step_avg:142.24ms step:60/3200 train_loss:5.4904 train_time:7113ms step_avg:142.26ms step:61/3200 train_loss:5.5934 train_time:7257ms step_avg:142.29ms step:62/3200 train_loss:5.3555 train_time:7400ms step_avg:142.30ms step:63/3200 train_loss:5.4544 train_time:7543ms step_avg:142.32ms step:64/3200 train_loss:5.4234 train_time:7687ms step_avg:142.35ms step:65/3200 train_loss:5.2083 train_time:7829ms step_avg:142.35ms step:66/3200 train_loss:5.2402 train_time:7973ms step_avg:142.38ms step:67/3200 train_loss:5.4073 train_time:8118ms step_avg:142.41ms step:68/3200 train_loss:5.2730 train_time:8260ms step_avg:142.42ms step:69/3200 train_loss:5.4973 train_time:8405ms step_avg:142.45ms step:70/3200 train_loss:5.1656 train_time:8554ms step_avg:142.57ms step:71/3200 train_loss:5.2384 train_time:8692ms step_avg:142.49ms step:72/3200 train_loss:5.3938 train_time:8838ms step_avg:142.55ms step:73/3200 train_loss:5.3234 train_time:8980ms step_avg:142.54ms step:74/3200 train_loss:5.2137 train_time:9124ms step_avg:142.56ms step:75/3200 train_loss:5.3253 train_time:9267ms step_avg:142.57ms step:76/3200 train_loss:5.3047 train_time:9412ms step_avg:142.61ms step:77/3200 train_loss:5.2480 train_time:9556ms step_avg:142.62ms step:78/3200 train_loss:5.3403 train_time:9699ms step_avg:142.63ms step:79/3200 train_loss:5.4402 train_time:9843ms step_avg:142.65ms step:80/3200 train_loss:5.1836 train_time:9985ms step_avg:142.65ms step:81/3200 train_loss:5.2772 train_time:10128ms step_avg:142.65ms step:82/3200 train_loss:5.0387 train_time:10271ms step_avg:142.66ms step:83/3200 train_loss:5.2198 train_time:10415ms step_avg:142.68ms step:84/3200 train_loss:5.1717 train_time:10559ms step_avg:142.68ms step:85/3200 train_loss:5.1532 train_time:10703ms step_avg:142.70ms step:86/3200 train_loss:5.0263 train_time:10846ms step_avg:142.71ms step:87/3200 train_loss:5.2197 train_time:10988ms step_avg:142.70ms step:88/3200 train_loss:5.1252 train_time:11133ms step_avg:142.72ms step:89/3200 train_loss:5.1758 train_time:11276ms step_avg:142.73ms step:90/3200 train_loss:5.1489 train_time:11421ms step_avg:142.76ms step:91/3200 train_loss:5.0618 train_time:11563ms step_avg:142.75ms step:92/3200 train_loss:5.0638 train_time:11706ms step_avg:142.76ms step:93/3200 train_loss:5.1853 train_time:11849ms step_avg:142.76ms step:94/3200 train_loss:5.0143 train_time:11990ms step_avg:142.74ms step:95/3200 train_loss:5.0051 train_time:12133ms step_avg:142.74ms step:96/3200 train_loss:5.0602 train_time:12276ms step_avg:142.74ms step:97/3200 train_loss:4.9582 train_time:12421ms step_avg:142.77ms step:98/3200 train_loss:5.0333 train_time:12562ms step_avg:142.75ms step:99/3200 train_loss:4.9553 train_time:12705ms step_avg:142.76ms step:100/3200 train_loss:5.0792 train_time:12847ms step_avg:142.75ms step:101/3200 train_loss:5.0388 train_time:12990ms step_avg:142.75ms step:102/3200 train_loss:4.9190 train_time:13132ms step_avg:142.74ms step:103/3200 train_loss:5.0495 train_time:13277ms step_avg:142.77ms step:104/3200 train_loss:4.9941 train_time:13418ms step_avg:142.74ms step:105/3200 train_loss:4.8622 train_time:13560ms step_avg:142.74ms step:106/3200 train_loss:4.9252 train_time:13709ms step_avg:142.80ms step:107/3200 train_loss:5.1201 train_time:13846ms step_avg:142.74ms step:108/3200 train_loss:4.9099 train_time:13988ms step_avg:142.74ms step:109/3200 train_loss:4.7019 train_time:14134ms step_avg:142.76ms step:110/3200 train_loss:4.8705 train_time:14277ms step_avg:142.77ms step:111/3200 train_loss:4.8681 train_time:14421ms step_avg:142.78ms step:112/3200 train_loss:4.8312 train_time:14566ms step_avg:142.80ms step:113/3200 train_loss:4.9528 train_time:14709ms step_avg:142.80ms step:114/3200 train_loss:4.8628 train_time:14850ms step_avg:142.79ms step:115/3200 train_loss:4.7210 train_time:14993ms step_avg:142.79ms step:116/3200 train_loss:4.8786 train_time:15138ms step_avg:142.81ms step:117/3200 train_loss:4.7916 train_time:15280ms step_avg:142.80ms step:118/3200 train_loss:4.7324 train_time:15424ms step_avg:142.81ms step:119/3200 train_loss:4.9023 train_time:15567ms step_avg:142.81ms step:120/3200 train_loss:4.8276 train_time:15712ms step_avg:142.83ms step:121/3200 train_loss:4.7494 train_time:15855ms step_avg:142.83ms step:122/3200 train_loss:4.6571 train_time:15997ms step_avg:142.83ms step:123/3200 train_loss:4.7781 train_time:16140ms step_avg:142.84ms step:124/3200 train_loss:4.6423 train_time:16283ms step_avg:142.83ms step:125/3200 train_loss:4.9395 train_time:16426ms step_avg:142.83ms step:125/3200 val_loss:4.7597 train_time:16482ms step_avg:143.32ms step:126/3200 train_loss:4.8053 train_time:16581ms step_avg:142.94ms step:127/3200 train_loss:4.7547 train_time:16730ms step_avg:142.99ms step:128/3200 train_loss:4.8036 train_time:16872ms step_avg:142.98ms step:129/3200 train_loss:4.6910 train_time:17014ms step_avg:142.97ms step:130/3200 train_loss:4.9959 train_time:17156ms step_avg:142.96ms step:131/3200 train_loss:4.7245 train_time:17299ms step_avg:142.96ms step:132/3200 train_loss:4.7254 train_time:17441ms step_avg:142.96ms step:133/3200 train_loss:4.6778 train_time:17585ms step_avg:142.97ms step:134/3200 train_loss:4.7392 train_time:17733ms step_avg:143.01ms step:135/3200 train_loss:4.6075 train_time:17877ms step_avg:143.02ms step:136/3200 train_loss:4.7397 train_time:18020ms step_avg:143.02ms step:137/3200 train_loss:4.5240 train_time:18163ms step_avg:143.02ms step:138/3200 train_loss:4.6827 train_time:18306ms step_avg:143.01ms step:139/3200 train_loss:4.6200 train_time:18453ms step_avg:143.05ms step:140/3200 train_loss:4.6664 train_time:18592ms step_avg:143.02ms step:141/3200 train_loss:4.7293 train_time:18737ms step_avg:143.03ms step:142/3200 train_loss:4.6096 train_time:18881ms step_avg:143.03ms step:143/3200 train_loss:4.6325 train_time:19024ms step_avg:143.04ms step:144/3200 train_loss:4.5204 train_time:19166ms step_avg:143.03ms step:145/3200 train_loss:4.6289 train_time:19308ms step_avg:143.03ms step:146/3200 train_loss:4.5887 train_time:19451ms step_avg:143.02ms step:147/3200 train_loss:4.4632 train_time:19595ms step_avg:143.03ms step:148/3200 train_loss:4.6023 train_time:19740ms step_avg:143.04ms step:149/3200 train_loss:4.6184 train_time:19884ms step_avg:143.05ms step:150/3200 train_loss:4.5950 train_time:20028ms step_avg:143.06ms step:151/3200 train_loss:4.6795 train_time:20171ms step_avg:143.06ms step:152/3200 train_loss:4.5447 train_time:20313ms step_avg:143.05ms step:153/3200 train_loss:4.5343 train_time:20455ms step_avg:143.04ms step:154/3200 train_loss:4.6104 train_time:20599ms step_avg:143.05ms step:155/3200 train_loss:4.5930 train_time:20744ms step_avg:143.06ms step:156/3200 train_loss:4.5226 train_time:20888ms step_avg:143.07ms step:157/3200 train_loss:4.5728 train_time:21031ms step_avg:143.07ms step:158/3200 train_loss:4.6515 train_time:21174ms step_avg:143.07ms step:159/3200 train_loss:4.4722 train_time:21317ms step_avg:143.07ms step:160/3200 train_loss:4.5402 train_time:21460ms step_avg:143.07ms step:161/3200 train_loss:4.3549 train_time:21603ms step_avg:143.07ms step:162/3200 train_loss:4.5509 train_time:21745ms step_avg:143.06ms step:163/3200 train_loss:4.5652 train_time:21889ms step_avg:143.06ms step:164/3200 train_loss:4.5597 train_time:22032ms step_avg:143.06ms step:165/3200 train_loss:4.4044 train_time:22175ms step_avg:143.07ms step:166/3200 train_loss:4.4932 train_time:22318ms step_avg:143.07ms step:167/3200 train_loss:4.5861 train_time:22462ms step_avg:143.07ms step:168/3200 train_loss:4.4063 train_time:22605ms step_avg:143.07ms step:169/3200 train_loss:4.4814 train_time:22747ms step_avg:143.06ms step:170/3200 train_loss:4.3736 train_time:22890ms step_avg:143.06ms step:171/3200 train_loss:4.2511 train_time:23034ms step_avg:143.07ms step:172/3200 train_loss:4.3941 train_time:23177ms step_avg:143.07ms step:173/3200 train_loss:4.4045 train_time:23320ms step_avg:143.07ms step:174/3200 train_loss:4.4508 train_time:23463ms step_avg:143.07ms step:175/3200 train_loss:4.6151 train_time:23606ms step_avg:143.07ms step:176/3200 train_loss:4.4423 train_time:23748ms step_avg:143.06ms step:177/3200 train_loss:4.3014 train_time:23893ms step_avg:143.07ms step:178/3200 train_loss:4.2690 train_time:24037ms step_avg:143.07ms step:179/3200 train_loss:4.3662 train_time:24180ms step_avg:143.08ms step:180/3200 train_loss:4.3335 train_time:24322ms step_avg:143.07ms step:181/3200 train_loss:4.3063 train_time:24464ms step_avg:143.07ms step:182/3200 train_loss:4.4711 train_time:24608ms step_avg:143.07ms step:183/3200 train_loss:4.3346 train_time:24752ms step_avg:143.08ms step:184/3200 train_loss:4.3190 train_time:24895ms step_avg:143.07ms step:185/3200 train_loss:4.3060 train_time:25039ms step_avg:143.08ms step:186/3200 train_loss:4.4032 train_time:25183ms step_avg:143.08ms step:187/3200 train_loss:4.3514 train_time:25325ms step_avg:143.08ms step:188/3200 train_loss:4.4472 train_time:25468ms step_avg:143.08ms step:189/3200 train_loss:4.3482 train_time:25764ms step_avg:143.93ms step:190/3200 train_loss:4.2805 train_time:26069ms step_avg:144.83ms step:191/3200 train_loss:4.3873 train_time:26207ms step_avg:144.79ms step:192/3200 train_loss:4.2565 train_time:26347ms step_avg:144.76ms step:193/3200 train_loss:4.1929 train_time:26488ms step_avg:144.74ms step:194/3200 train_loss:4.4156 train_time:26631ms step_avg:144.73ms step:195/3200 train_loss:4.3301 train_time:26772ms step_avg:144.71ms step:196/3200 train_loss:4.5321 train_time:26914ms step_avg:144.70ms step:197/3200 train_loss:4.3665 train_time:27062ms step_avg:144.72ms step:198/3200 train_loss:4.2103 train_time:27206ms step_avg:144.72ms step:199/3200 train_loss:4.3302 train_time:27348ms step_avg:144.70ms step:200/3200 train_loss:4.1906 train_time:27491ms step_avg:144.69ms step:201/3200 train_loss:4.2905 train_time:27632ms step_avg:144.67ms step:202/3200 train_loss:4.1680 train_time:27774ms step_avg:144.65ms step:203/3200 train_loss:4.4010 train_time:27917ms step_avg:144.65ms step:204/3200 train_loss:4.2326 train_time:28061ms step_avg:144.65ms step:205/3200 train_loss:4.3451 train_time:28206ms step_avg:144.65ms step:206/3200 train_loss:4.4018 train_time:28348ms step_avg:144.63ms step:207/3200 train_loss:4.1017 train_time:28491ms step_avg:144.62ms step:208/3200 train_loss:4.2494 train_time:28634ms step_avg:144.62ms step:209/3200 train_loss:4.2430 train_time:28777ms step_avg:144.61ms step:210/3200 train_loss:4.3951 train_time:28920ms step_avg:144.60ms step:211/3200 train_loss:4.3283 train_time:29066ms step_avg:144.61ms step:212/3200 train_loss:4.2122 train_time:29209ms step_avg:144.60ms step:213/3200 train_loss:4.2374 train_time:29353ms step_avg:144.59ms step:214/3200 train_loss:4.1886 train_time:29495ms step_avg:144.58ms step:215/3200 train_loss:4.2636 train_time:29637ms step_avg:144.57ms step:216/3200 train_loss:4.0840 train_time:29780ms step_avg:144.56ms step:217/3200 train_loss:4.1538 train_time:29923ms step_avg:144.55ms step:218/3200 train_loss:4.1555 train_time:30066ms step_avg:144.55ms step:219/3200 train_loss:4.2256 train_time:30208ms step_avg:144.53ms step:220/3200 train_loss:4.2100 train_time:30351ms step_avg:144.53ms step:221/3200 train_loss:4.2320 train_time:30494ms step_avg:144.52ms step:222/3200 train_loss:4.2476 train_time:30638ms step_avg:144.52ms step:223/3200 train_loss:4.1652 train_time:30781ms step_avg:144.51ms step:224/3200 train_loss:4.1275 train_time:30923ms step_avg:144.50ms step:225/3200 train_loss:4.4249 train_time:31067ms step_avg:144.50ms step:226/3200 train_loss:4.0541 train_time:31210ms step_avg:144.49ms step:227/3200 train_loss:4.1231 train_time:31353ms step_avg:144.48ms step:228/3200 train_loss:4.1339 train_time:31498ms step_avg:144.49ms step:229/3200 train_loss:4.2783 train_time:31642ms step_avg:144.48ms step:230/3200 train_loss:4.0653 train_time:31784ms step_avg:144.47ms step:231/3200 train_loss:4.1892 train_time:31927ms step_avg:144.47ms step:232/3200 train_loss:4.0426 train_time:32070ms step_avg:144.46ms step:233/3200 train_loss:4.1068 train_time:32212ms step_avg:144.45ms step:234/3200 train_loss:4.2400 train_time:32355ms step_avg:144.44ms step:235/3200 train_loss:4.1581 train_time:32497ms step_avg:144.43ms step:236/3200 train_loss:4.0497 train_time:32640ms step_avg:144.42ms step:237/3200 train_loss:4.2104 train_time:32783ms step_avg:144.42ms step:238/3200 train_loss:4.2210 train_time:32926ms step_avg:144.41ms step:239/3200 train_loss:4.0767 train_time:33071ms step_avg:144.42ms step:240/3200 train_loss:4.2219 train_time:33213ms step_avg:144.41ms step:241/3200 train_loss:4.2447 train_time:33357ms step_avg:144.40ms step:242/3200 train_loss:4.0969 train_time:33500ms step_avg:144.40ms step:243/3200 train_loss:4.2729 train_time:33644ms step_avg:144.39ms step:244/3200 train_loss:4.1468 train_time:33787ms step_avg:144.39ms step:245/3200 train_loss:4.2018 train_time:33930ms step_avg:144.38ms step:246/3200 train_loss:4.2768 train_time:34073ms step_avg:144.38ms step:247/3200 train_loss:4.1964 train_time:34217ms step_avg:144.38ms step:248/3200 train_loss:4.1372 train_time:34360ms step_avg:144.37ms step:249/3200 train_loss:4.2492 train_time:34505ms step_avg:144.37ms step:250/3200 train_loss:4.0515 train_time:34647ms step_avg:144.36ms step:250/3200 val_loss:4.1357 train_time:34703ms step_avg:144.60ms step:251/3200 train_loss:4.0993 train_time:34803ms step_avg:144.41ms step:252/3200 train_loss:4.2041 train_time:34949ms step_avg:144.42ms step:253/3200 train_loss:4.2758 train_time:35093ms step_avg:144.42ms step:254/3200 train_loss:4.0629 train_time:35235ms step_avg:144.41ms step:255/3200 train_loss:4.0152 train_time:35376ms step_avg:144.39ms step:256/3200 train_loss:4.1954 train_time:35518ms step_avg:144.38ms step:257/3200 train_loss:4.1030 train_time:35660ms step_avg:144.37ms step:258/3200 train_loss:4.1132 train_time:35804ms step_avg:144.37ms step:259/3200 train_loss:4.0958 train_time:35949ms step_avg:144.37ms step:260/3200 train_loss:4.1475 train_time:36093ms step_avg:144.37ms step:261/3200 train_loss:4.1834 train_time:36236ms step_avg:144.37ms step:262/3200 train_loss:4.1488 train_time:36378ms step_avg:144.36ms step:263/3200 train_loss:4.1141 train_time:36521ms step_avg:144.35ms step:264/3200 train_loss:4.0274 train_time:36664ms step_avg:144.35ms step:265/3200 train_loss:4.1076 train_time:36809ms step_avg:144.35ms step:266/3200 train_loss:3.9833 train_time:36952ms step_avg:144.35ms step:267/3200 train_loss:4.0344 train_time:37096ms step_avg:144.34ms step:268/3200 train_loss:4.0450 train_time:37239ms step_avg:144.34ms step:269/3200 train_loss:4.0681 train_time:37381ms step_avg:144.33ms step:270/3200 train_loss:3.9812 train_time:37523ms step_avg:144.32ms step:271/3200 train_loss:4.2153 train_time:37666ms step_avg:144.31ms step:272/3200 train_loss:4.0999 train_time:37810ms step_avg:144.31ms step:273/3200 train_loss:4.0305 train_time:37953ms step_avg:144.31ms step:274/3200 train_loss:4.0767 train_time:38097ms step_avg:144.31ms step:275/3200 train_loss:4.1587 train_time:38239ms step_avg:144.30ms step:276/3200 train_loss:4.1812 train_time:38382ms step_avg:144.29ms step:277/3200 train_loss:4.3450 train_time:38523ms step_avg:144.28ms step:278/3200 train_loss:4.1477 train_time:38666ms step_avg:144.28ms step:279/3200 train_loss:4.1975 train_time:38809ms step_avg:144.27ms step:280/3200 train_loss:4.1136 train_time:38953ms step_avg:144.27ms step:281/3200 train_loss:4.2416 train_time:39098ms step_avg:144.27ms step:282/3200 train_loss:4.0739 train_time:39240ms step_avg:144.26ms step:283/3200 train_loss:4.0734 train_time:39382ms step_avg:144.26ms step:284/3200 train_loss:4.0277 train_time:39525ms step_avg:144.25ms step:285/3200 train_loss:4.1610 train_time:39668ms step_avg:144.25ms step:286/3200 train_loss:4.1673 train_time:39811ms step_avg:144.24ms step:287/3200 train_loss:4.2032 train_time:39953ms step_avg:144.24ms step:288/3200 train_loss:4.0346 train_time:40097ms step_avg:144.23ms step:289/3200 train_loss:4.1346 train_time:40240ms step_avg:144.23ms step:290/3200 train_loss:3.9810 train_time:40383ms step_avg:144.22ms step:291/3200 train_loss:3.9796 train_time:40529ms step_avg:144.23ms step:292/3200 train_loss:4.0532 train_time:40670ms step_avg:144.22ms step:293/3200 train_loss:3.9854 train_time:40813ms step_avg:144.22ms step:294/3200 train_loss:4.0272 train_time:40955ms step_avg:144.21ms step:295/3200 train_loss:4.0721 train_time:41099ms step_avg:144.21ms step:296/3200 train_loss:3.9510 train_time:41244ms step_avg:144.21ms step:297/3200 train_loss:3.9743 train_time:41389ms step_avg:144.21ms step:298/3200 train_loss:3.9753 train_time:41532ms step_avg:144.21ms step:299/3200 train_loss:4.0824 train_time:41675ms step_avg:144.20ms step:300/3200 train_loss:3.9421 train_time:41819ms step_avg:144.20ms step:301/3200 train_loss:4.0853 train_time:41962ms step_avg:144.20ms step:302/3200 train_loss:4.0952 train_time:42104ms step_avg:144.19ms step:303/3200 train_loss:4.0459 train_time:42249ms step_avg:144.19ms step:304/3200 train_loss:4.0882 train_time:42392ms step_avg:144.19ms step:305/3200 train_loss:4.0732 train_time:42535ms step_avg:144.19ms step:306/3200 train_loss:4.5666 train_time:42679ms step_avg:144.18ms step:307/3200 train_loss:4.0496 train_time:42821ms step_avg:144.18ms step:308/3200 train_loss:3.9529 train_time:42964ms step_avg:144.17ms step:309/3200 train_loss:4.1009 train_time:43107ms step_avg:144.17ms step:310/3200 train_loss:3.9673 train_time:43251ms step_avg:144.17ms step:311/3200 train_loss:4.1934 train_time:43394ms step_avg:144.16ms step:312/3200 train_loss:4.0403 train_time:43536ms step_avg:144.16ms step:313/3200 train_loss:3.9839 train_time:43678ms step_avg:144.15ms step:314/3200 train_loss:4.0711 train_time:43821ms step_avg:144.15ms step:315/3200 train_loss:4.1968 train_time:43964ms step_avg:144.14ms step:316/3200 train_loss:4.0703 train_time:44108ms step_avg:144.14ms step:317/3200 train_loss:3.9045 train_time:44252ms step_avg:144.14ms step:318/3200 train_loss:3.9878 train_time:44397ms step_avg:144.15ms step:319/3200 train_loss:4.0234 train_time:44540ms step_avg:144.14ms step:320/3200 train_loss:4.0026 train_time:44682ms step_avg:144.14ms step:321/3200 train_loss:4.1100 train_time:44825ms step_avg:144.13ms step:322/3200 train_loss:4.0541 train_time:44970ms step_avg:144.13ms step:323/3200 train_loss:4.0369 train_time:45114ms step_avg:144.13ms step:324/3200 train_loss:4.1212 train_time:45256ms step_avg:144.13ms step:325/3200 train_loss:4.0641 train_time:45400ms step_avg:144.13ms step:326/3200 train_loss:4.1379 train_time:45542ms step_avg:144.12ms step:327/3200 train_loss:4.0035 train_time:45685ms step_avg:144.12ms step:328/3200 train_loss:4.5028 train_time:45828ms step_avg:144.11ms step:329/3200 train_loss:4.1828 train_time:45971ms step_avg:144.11ms step:330/3200 train_loss:3.9222 train_time:46113ms step_avg:144.10ms step:331/3200 train_loss:3.8650 train_time:46266ms step_avg:144.13ms step:332/3200 train_loss:4.0809 train_time:46399ms step_avg:144.10ms step:333/3200 train_loss:4.0213 train_time:46542ms step_avg:144.09ms step:334/3200 train_loss:3.9847 train_time:46684ms step_avg:144.09ms step:335/3200 train_loss:3.9475 train_time:46827ms step_avg:144.08ms step:336/3200 train_loss:4.1253 train_time:46970ms step_avg:144.08ms step:337/3200 train_loss:4.0582 train_time:47114ms step_avg:144.08ms step:338/3200 train_loss:4.5321 train_time:47256ms step_avg:144.07ms step:339/3200 train_loss:4.0424 train_time:47398ms step_avg:144.07ms step:340/3200 train_loss:3.9943 train_time:47542ms step_avg:144.07ms step:341/3200 train_loss:4.0386 train_time:47684ms step_avg:144.06ms step:342/3200 train_loss:3.9601 train_time:47828ms step_avg:144.06ms step:343/3200 train_loss:3.9209 train_time:47972ms step_avg:144.06ms step:344/3200 train_loss:3.9584 train_time:48116ms step_avg:144.06ms step:345/3200 train_loss:4.1061 train_time:48259ms step_avg:144.06ms step:346/3200 train_loss:3.9455 train_time:48402ms step_avg:144.05ms step:347/3200 train_loss:3.8749 train_time:48546ms step_avg:144.05ms step:348/3200 train_loss:3.9173 train_time:48689ms step_avg:144.05ms step:349/3200 train_loss:3.9722 train_time:48831ms step_avg:144.05ms step:350/3200 train_loss:3.9304 train_time:48975ms step_avg:144.04ms step:351/3200 train_loss:3.6717 train_time:49118ms step_avg:144.04ms step:352/3200 train_loss:3.9296 train_time:49262ms step_avg:144.04ms step:353/3200 train_loss:4.2662 train_time:49406ms step_avg:144.04ms step:354/3200 train_loss:3.7656 train_time:49550ms step_avg:144.04ms step:355/3200 train_loss:4.0371 train_time:49693ms step_avg:144.04ms step:356/3200 train_loss:3.8958 train_time:49836ms step_avg:144.03ms step:357/3200 train_loss:3.9968 train_time:49979ms step_avg:144.03ms step:358/3200 train_loss:3.9178 train_time:50121ms step_avg:144.03ms step:359/3200 train_loss:3.9559 train_time:50264ms step_avg:144.02ms step:360/3200 train_loss:3.9692 train_time:50408ms step_avg:144.02ms step:361/3200 train_loss:3.5544 train_time:50552ms step_avg:144.02ms step:362/3200 train_loss:4.1293 train_time:50697ms step_avg:144.03ms step:363/3200 train_loss:4.0326 train_time:50839ms step_avg:144.02ms step:364/3200 train_loss:3.9502 train_time:50981ms step_avg:144.01ms step:365/3200 train_loss:3.8509 train_time:51125ms step_avg:144.01ms step:366/3200 train_loss:4.0195 train_time:51269ms step_avg:144.01ms step:367/3200 train_loss:3.9761 train_time:51411ms step_avg:144.01ms step:368/3200 train_loss:3.9706 train_time:51554ms step_avg:144.01ms step:369/3200 train_loss:3.9536 train_time:51698ms step_avg:144.00ms step:370/3200 train_loss:3.8540 train_time:51841ms step_avg:144.00ms step:371/3200 train_loss:3.9931 train_time:51982ms step_avg:144.00ms step:372/3200 train_loss:3.8585 train_time:52125ms step_avg:143.99ms step:373/3200 train_loss:3.8036 train_time:52270ms step_avg:143.99ms step:374/3200 train_loss:4.0290 train_time:52412ms step_avg:143.99ms step:375/3200 train_loss:3.9421 train_time:52555ms step_avg:143.99ms step:375/3200 val_loss:3.9380 train_time:52611ms step_avg:144.14ms step:376/3200 train_loss:3.9176 train_time:52710ms step_avg:144.02ms step:377/3200 train_loss:3.9705 train_time:52856ms step_avg:144.02ms step:378/3200 train_loss:3.8923 train_time:53159ms step_avg:144.45ms step:379/3200 train_loss:3.9489 train_time:53300ms step_avg:144.44ms step:380/3200 train_loss:3.9864 train_time:53605ms step_avg:144.88ms step:381/3200 train_loss:4.0561 train_time:53747ms step_avg:144.87ms step:382/3200 train_loss:3.9558 train_time:53890ms step_avg:144.87ms step:383/3200 train_loss:3.9231 train_time:54032ms step_avg:144.86ms step:384/3200 train_loss:3.9024 train_time:54173ms step_avg:144.85ms step:385/3200 train_loss:3.9791 train_time:54315ms step_avg:144.84ms step:386/3200 train_loss:3.8993 train_time:54457ms step_avg:144.83ms step:387/3200 train_loss:4.0023 train_time:54606ms step_avg:144.84ms step:388/3200 train_loss:4.1827 train_time:54751ms step_avg:144.84ms step:389/3200 train_loss:3.9098 train_time:54894ms step_avg:144.84ms step:390/3200 train_loss:3.8968 train_time:55035ms step_avg:144.83ms step:391/3200 train_loss:4.0052 train_time:55178ms step_avg:144.82ms step:392/3200 train_loss:3.9219 train_time:55320ms step_avg:144.82ms step:393/3200 train_loss:4.0310 train_time:55463ms step_avg:144.81ms step:394/3200 train_loss:3.8745 train_time:55609ms step_avg:144.82ms step:395/3200 train_loss:4.0044 train_time:55754ms step_avg:144.82ms step:396/3200 train_loss:3.7435 train_time:55898ms step_avg:144.81ms step:397/3200 train_loss:3.9579 train_time:56040ms step_avg:144.81ms step:398/3200 train_loss:3.9855 train_time:56182ms step_avg:144.80ms step:399/3200 train_loss:3.9989 train_time:56324ms step_avg:144.79ms step:400/3200 train_loss:3.8940 train_time:56467ms step_avg:144.79ms step:401/3200 train_loss:3.9468 train_time:56611ms step_avg:144.79ms step:402/3200 train_loss:4.0229 train_time:56755ms step_avg:144.78ms step:403/3200 train_loss:3.9566 train_time:56899ms step_avg:144.78ms step:404/3200 train_loss:4.0636 train_time:57040ms step_avg:144.77ms step:405/3200 train_loss:3.8105 train_time:57183ms step_avg:144.77ms step:406/3200 train_loss:3.9068 train_time:57325ms step_avg:144.76ms step:407/3200 train_loss:4.1982 train_time:57469ms step_avg:144.76ms step:408/3200 train_loss:3.9038 train_time:57612ms step_avg:144.75ms step:409/3200 train_loss:3.9296 train_time:57756ms step_avg:144.75ms step:410/3200 train_loss:3.9752 train_time:57899ms step_avg:144.75ms step:411/3200 train_loss:3.8669 train_time:58040ms step_avg:144.74ms step:412/3200 train_loss:3.8758 train_time:58182ms step_avg:144.73ms step:413/3200 train_loss:4.2952 train_time:58325ms step_avg:144.73ms step:414/3200 train_loss:3.7563 train_time:58469ms step_avg:144.72ms step:415/3200 train_loss:4.1256 train_time:58612ms step_avg:144.72ms step:416/3200 train_loss:3.8754 train_time:58755ms step_avg:144.72ms step:417/3200 train_loss:3.8806 train_time:58902ms step_avg:144.72ms step:418/3200 train_loss:4.0647 train_time:59043ms step_avg:144.71ms step:419/3200 train_loss:3.8011 train_time:59186ms step_avg:144.71ms step:420/3200 train_loss:3.9228 train_time:59330ms step_avg:144.71ms step:421/3200 train_loss:3.8349 train_time:59472ms step_avg:144.70ms step:422/3200 train_loss:3.7628 train_time:59614ms step_avg:144.69ms step:423/3200 train_loss:3.8929 train_time:59758ms step_avg:144.69ms step:424/3200 train_loss:3.9874 train_time:59901ms step_avg:144.69ms step:425/3200 train_loss:3.7374 train_time:60045ms step_avg:144.69ms step:426/3200 train_loss:3.9147 train_time:60187ms step_avg:144.68ms step:427/3200 train_loss:3.7992 train_time:60330ms step_avg:144.68ms step:428/3200 train_loss:4.0180 train_time:60472ms step_avg:144.67ms step:429/3200 train_loss:3.9324 train_time:60615ms step_avg:144.67ms step:430/3200 train_loss:3.8730 train_time:60759ms step_avg:144.66ms step:431/3200 train_loss:3.8355 train_time:60902ms step_avg:144.66ms step:432/3200 train_loss:3.7452 train_time:61045ms step_avg:144.66ms step:433/3200 train_loss:3.8788 train_time:61189ms step_avg:144.66ms step:434/3200 train_loss:3.9394 train_time:61331ms step_avg:144.65ms step:435/3200 train_loss:3.8894 train_time:61474ms step_avg:144.64ms step:436/3200 train_loss:3.9283 train_time:61616ms step_avg:144.64ms step:437/3200 train_loss:3.9477 train_time:61760ms step_avg:144.64ms step:438/3200 train_loss:3.8252 train_time:61904ms step_avg:144.64ms step:439/3200 train_loss:3.8304 train_time:62046ms step_avg:144.63ms step:440/3200 train_loss:3.8183 train_time:62191ms step_avg:144.63ms step:441/3200 train_loss:3.9997 train_time:62333ms step_avg:144.62ms step:442/3200 train_loss:3.8808 train_time:62476ms step_avg:144.62ms step:443/3200 train_loss:3.8587 train_time:62619ms step_avg:144.62ms step:444/3200 train_loss:3.7614 train_time:62764ms step_avg:144.62ms step:445/3200 train_loss:4.0315 train_time:62908ms step_avg:144.62ms step:446/3200 train_loss:3.9592 train_time:63051ms step_avg:144.61ms step:447/3200 train_loss:3.9520 train_time:63193ms step_avg:144.61ms step:448/3200 train_loss:3.8675 train_time:63335ms step_avg:144.60ms step:449/3200 train_loss:3.9700 train_time:63478ms step_avg:144.60ms step:450/3200 train_loss:3.8018 train_time:63621ms step_avg:144.59ms step:451/3200 train_loss:3.8392 train_time:63765ms step_avg:144.59ms step:452/3200 train_loss:3.7049 train_time:63909ms step_avg:144.59ms step:453/3200 train_loss:3.8285 train_time:64052ms step_avg:144.59ms step:454/3200 train_loss:3.7957 train_time:64195ms step_avg:144.58ms step:455/3200 train_loss:3.7504 train_time:64337ms step_avg:144.58ms step:456/3200 train_loss:3.9657 train_time:64480ms step_avg:144.57ms step:457/3200 train_loss:3.8405 train_time:64625ms step_avg:144.58ms step:458/3200 train_loss:3.9117 train_time:64769ms step_avg:144.57ms step:459/3200 train_loss:3.9522 train_time:64911ms step_avg:144.57ms step:460/3200 train_loss:3.7554 train_time:65054ms step_avg:144.56ms step:461/3200 train_loss:3.9184 train_time:65197ms step_avg:144.56ms step:462/3200 train_loss:3.8237 train_time:65339ms step_avg:144.56ms step:463/3200 train_loss:3.8403 train_time:65481ms step_avg:144.55ms step:464/3200 train_loss:3.8948 train_time:65624ms step_avg:144.55ms step:465/3200 train_loss:3.8348 train_time:65767ms step_avg:144.54ms step:466/3200 train_loss:3.8426 train_time:65910ms step_avg:144.54ms step:467/3200 train_loss:3.9288 train_time:66052ms step_avg:144.53ms step:468/3200 train_loss:3.9391 train_time:66195ms step_avg:144.53ms step:469/3200 train_loss:3.9218 train_time:66337ms step_avg:144.52ms step:470/3200 train_loss:3.8137 train_time:66479ms step_avg:144.52ms step:471/3200 train_loss:3.8934 train_time:66624ms step_avg:144.52ms step:472/3200 train_loss:3.9437 train_time:66769ms step_avg:144.52ms step:473/3200 train_loss:3.8913 train_time:66912ms step_avg:144.52ms step:474/3200 train_loss:3.8412 train_time:67054ms step_avg:144.51ms step:475/3200 train_loss:3.7076 train_time:67197ms step_avg:144.51ms step:476/3200 train_loss:4.1392 train_time:67339ms step_avg:144.50ms step:477/3200 train_loss:3.8929 train_time:67481ms step_avg:144.50ms step:478/3200 train_loss:3.7066 train_time:67624ms step_avg:144.50ms step:479/3200 train_loss:3.9385 train_time:67767ms step_avg:144.49ms step:480/3200 train_loss:3.8898 train_time:67911ms step_avg:144.49ms step:481/3200 train_loss:4.0340 train_time:68054ms step_avg:144.49ms step:482/3200 train_loss:3.8435 train_time:68198ms step_avg:144.49ms step:483/3200 train_loss:3.6488 train_time:68341ms step_avg:144.48ms step:484/3200 train_loss:3.9340 train_time:68483ms step_avg:144.48ms step:485/3200 train_loss:3.7873 train_time:68625ms step_avg:144.47ms step:486/3200 train_loss:3.7894 train_time:68768ms step_avg:144.47ms step:487/3200 train_loss:3.7189 train_time:68912ms step_avg:144.47ms step:488/3200 train_loss:3.7967 train_time:69056ms step_avg:144.47ms step:489/3200 train_loss:3.9962 train_time:69199ms step_avg:144.46ms step:490/3200 train_loss:3.8438 train_time:69341ms step_avg:144.46ms step:491/3200 train_loss:3.7279 train_time:69483ms step_avg:144.46ms step:492/3200 train_loss:3.7380 train_time:69627ms step_avg:144.45ms step:493/3200 train_loss:3.8527 train_time:69771ms step_avg:144.45ms step:494/3200 train_loss:3.7035 train_time:69914ms step_avg:144.45ms step:495/3200 train_loss:3.8299 train_time:70058ms step_avg:144.45ms step:496/3200 train_loss:3.7805 train_time:70200ms step_avg:144.44ms step:497/3200 train_loss:3.6497 train_time:70344ms step_avg:144.44ms step:498/3200 train_loss:3.8561 train_time:70487ms step_avg:144.44ms step:499/3200 train_loss:3.9287 train_time:70629ms step_avg:144.44ms step:500/3200 train_loss:3.9570 train_time:70772ms step_avg:144.43ms step:500/3200 val_loss:3.8334 train_time:70827ms step_avg:144.55ms step:501/3200 train_loss:3.8714 train_time:70926ms step_avg:144.45ms step:502/3200 train_loss:3.9238 train_time:71073ms step_avg:144.46ms step:503/3200 train_loss:3.8740 train_time:71213ms step_avg:144.45ms step:504/3200 train_loss:3.9076 train_time:71355ms step_avg:144.44ms step:505/3200 train_loss:3.8529 train_time:71497ms step_avg:144.44ms step:506/3200 train_loss:3.9405 train_time:71638ms step_avg:144.43ms step:507/3200 train_loss:3.7781 train_time:71779ms step_avg:144.42ms step:508/3200 train_loss:3.8884 train_time:71924ms step_avg:144.43ms step:509/3200 train_loss:3.9561 train_time:72071ms step_avg:144.43ms step:510/3200 train_loss:3.8954 train_time:72214ms step_avg:144.43ms step:511/3200 train_loss:3.7123 train_time:72356ms step_avg:144.42ms step:512/3200 train_loss:3.9080 train_time:72498ms step_avg:144.42ms step:513/3200 train_loss:3.8440 train_time:72639ms step_avg:144.41ms step:514/3200 train_loss:3.8110 train_time:72782ms step_avg:144.41ms step:515/3200 train_loss:3.8885 train_time:72925ms step_avg:144.41ms step:516/3200 train_loss:3.8598 train_time:73071ms step_avg:144.41ms step:517/3200 train_loss:4.2107 train_time:73214ms step_avg:144.41ms step:518/3200 train_loss:3.8140 train_time:73357ms step_avg:144.40ms step:519/3200 train_loss:3.9087 train_time:73499ms step_avg:144.40ms step:520/3200 train_loss:3.8014 train_time:73641ms step_avg:144.39ms step:521/3200 train_loss:3.8173 train_time:73783ms step_avg:144.39ms step:522/3200 train_loss:3.7717 train_time:73926ms step_avg:144.39ms step:523/3200 train_loss:3.7867 train_time:74069ms step_avg:144.38ms step:524/3200 train_loss:4.4148 train_time:74213ms step_avg:144.38ms step:525/3200 train_loss:3.8682 train_time:74357ms step_avg:144.38ms step:526/3200 train_loss:3.8075 train_time:74499ms step_avg:144.38ms step:527/3200 train_loss:3.8173 train_time:74642ms step_avg:144.38ms step:528/3200 train_loss:3.7818 train_time:74786ms step_avg:144.37ms step:529/3200 train_loss:3.7491 train_time:74928ms step_avg:144.37ms step:530/3200 train_loss:3.9750 train_time:75071ms step_avg:144.37ms step:531/3200 train_loss:3.7744 train_time:75213ms step_avg:144.36ms step:532/3200 train_loss:4.0378 train_time:75356ms step_avg:144.36ms step:533/3200 train_loss:3.8589 train_time:75499ms step_avg:144.36ms step:534/3200 train_loss:3.7830 train_time:75641ms step_avg:144.35ms step:535/3200 train_loss:3.8085 train_time:75785ms step_avg:144.35ms step:536/3200 train_loss:3.7473 train_time:75928ms step_avg:144.35ms step:537/3200 train_loss:3.8695 train_time:76071ms step_avg:144.35ms step:538/3200 train_loss:3.8567 train_time:76215ms step_avg:144.35ms step:539/3200 train_loss:3.7568 train_time:76358ms step_avg:144.34ms step:540/3200 train_loss:4.2592 train_time:76502ms step_avg:144.34ms step:541/3200 train_loss:3.7981 train_time:76646ms step_avg:144.34ms step:542/3200 train_loss:3.9039 train_time:76790ms step_avg:144.34ms step:543/3200 train_loss:3.7319 train_time:76934ms step_avg:144.34ms step:544/3200 train_loss:3.7087 train_time:77077ms step_avg:144.34ms step:545/3200 train_loss:3.7894 train_time:77221ms step_avg:144.34ms step:546/3200 train_loss:3.7209 train_time:77364ms step_avg:144.34ms step:547/3200 train_loss:3.7668 train_time:77508ms step_avg:144.34ms step:548/3200 train_loss:3.7717 train_time:77651ms step_avg:144.33ms step:549/3200 train_loss:3.7498 train_time:77794ms step_avg:144.33ms step:550/3200 train_loss:3.8541 train_time:77938ms step_avg:144.33ms step:551/3200 train_loss:3.7360 train_time:78080ms step_avg:144.33ms step:552/3200 train_loss:3.7575 train_time:78223ms step_avg:144.32ms step:553/3200 train_loss:4.0754 train_time:78368ms step_avg:144.32ms step:554/3200 train_loss:3.8789 train_time:78510ms step_avg:144.32ms step:555/3200 train_loss:3.8411 train_time:78653ms step_avg:144.32ms step:556/3200 train_loss:3.7794 train_time:78796ms step_avg:144.32ms step:557/3200 train_loss:3.8212 train_time:78938ms step_avg:144.31ms step:558/3200 train_loss:3.4758 train_time:79082ms step_avg:144.31ms step:559/3200 train_loss:3.7421 train_time:79225ms step_avg:144.31ms step:560/3200 train_loss:3.7895 train_time:79370ms step_avg:144.31ms step:561/3200 train_loss:3.8288 train_time:79513ms step_avg:144.31ms step:562/3200 train_loss:3.7414 train_time:79656ms step_avg:144.30ms step:563/3200 train_loss:3.6807 train_time:79799ms step_avg:144.30ms step:564/3200 train_loss:3.8839 train_time:79941ms step_avg:144.30ms step:565/3200 train_loss:3.6986 train_time:80085ms step_avg:144.30ms step:566/3200 train_loss:3.8197 train_time:80229ms step_avg:144.30ms step:567/3200 train_loss:3.7678 train_time:80534ms step_avg:144.59ms step:568/3200 train_loss:3.7247 train_time:80677ms step_avg:144.58ms step:569/3200 train_loss:3.8139 train_time:80818ms step_avg:144.58ms step:570/3200 train_loss:3.7856 train_time:81131ms step_avg:144.88ms step:571/3200 train_loss:3.8154 train_time:81272ms step_avg:144.87ms step:572/3200 train_loss:3.8976 train_time:81413ms step_avg:144.86ms step:573/3200 train_loss:3.8480 train_time:81555ms step_avg:144.86ms step:574/3200 train_loss:3.8547 train_time:81696ms step_avg:144.85ms step:575/3200 train_loss:3.9037 train_time:81837ms step_avg:144.84ms step:576/3200 train_loss:3.8594 train_time:81979ms step_avg:144.84ms step:577/3200 train_loss:3.8856 train_time:82126ms step_avg:144.84ms step:578/3200 train_loss:3.8033 train_time:82271ms step_avg:144.84ms step:579/3200 train_loss:3.8046 train_time:82414ms step_avg:144.84ms step:580/3200 train_loss:3.7933 train_time:82557ms step_avg:144.84ms step:581/3200 train_loss:3.7303 train_time:82698ms step_avg:144.83ms step:582/3200 train_loss:3.7589 train_time:82839ms step_avg:144.82ms step:583/3200 train_loss:3.9849 train_time:82981ms step_avg:144.82ms step:584/3200 train_loss:3.7555 train_time:83128ms step_avg:144.82ms step:585/3200 train_loss:3.7163 train_time:83271ms step_avg:144.82ms step:586/3200 train_loss:3.9083 train_time:83414ms step_avg:144.82ms step:587/3200 train_loss:3.6644 train_time:83557ms step_avg:144.81ms step:588/3200 train_loss:3.7986 train_time:83699ms step_avg:144.81ms step:589/3200 train_loss:3.7764 train_time:83841ms step_avg:144.80ms step:590/3200 train_loss:4.1259 train_time:83985ms step_avg:144.80ms step:591/3200 train_loss:3.9178 train_time:84133ms step_avg:144.81ms step:592/3200 train_loss:3.6477 train_time:84272ms step_avg:144.80ms step:593/3200 train_loss:3.6630 train_time:84416ms step_avg:144.80ms step:594/3200 train_loss:3.6496 train_time:84560ms step_avg:144.79ms step:595/3200 train_loss:3.6976 train_time:84702ms step_avg:144.79ms step:596/3200 train_loss:4.0605 train_time:84845ms step_avg:144.79ms step:597/3200 train_loss:3.7875 train_time:84990ms step_avg:144.79ms step:598/3200 train_loss:3.7194 train_time:85134ms step_avg:144.79ms step:599/3200 train_loss:3.7947 train_time:85279ms step_avg:144.79ms step:600/3200 train_loss:3.6091 train_time:85423ms step_avg:144.78ms step:601/3200 train_loss:3.7270 train_time:85567ms step_avg:144.78ms step:602/3200 train_loss:3.7661 train_time:85709ms step_avg:144.78ms step:603/3200 train_loss:3.7879 train_time:85852ms step_avg:144.78ms step:604/3200 train_loss:3.9116 train_time:85995ms step_avg:144.77ms step:605/3200 train_loss:3.7630 train_time:86137ms step_avg:144.77ms step:606/3200 train_loss:3.7486 train_time:86282ms step_avg:144.77ms step:607/3200 train_loss:3.6998 train_time:86426ms step_avg:144.77ms step:608/3200 train_loss:3.9484 train_time:86569ms step_avg:144.76ms step:609/3200 train_loss:3.7781 train_time:86712ms step_avg:144.76ms step:610/3200 train_loss:3.7499 train_time:86854ms step_avg:144.76ms step:611/3200 train_loss:3.8481 train_time:86996ms step_avg:144.75ms step:612/3200 train_loss:3.7465 train_time:87140ms step_avg:144.75ms step:613/3200 train_loss:3.7383 train_time:87284ms step_avg:144.75ms step:614/3200 train_loss:3.9006 train_time:87428ms step_avg:144.75ms step:615/3200 train_loss:3.8452 train_time:87570ms step_avg:144.74ms step:616/3200 train_loss:3.8209 train_time:87713ms step_avg:144.74ms step:617/3200 train_loss:3.7531 train_time:87856ms step_avg:144.74ms step:618/3200 train_loss:3.7077 train_time:87998ms step_avg:144.73ms step:619/3200 train_loss:3.8125 train_time:88140ms step_avg:144.73ms step:620/3200 train_loss:3.7077 train_time:88284ms step_avg:144.73ms step:621/3200 train_loss:3.7235 train_time:88429ms step_avg:144.73ms step:622/3200 train_loss:4.0360 train_time:88572ms step_avg:144.72ms step:623/3200 train_loss:3.7181 train_time:88714ms step_avg:144.72ms step:624/3200 train_loss:3.7442 train_time:88857ms step_avg:144.72ms step:625/3200 train_loss:3.8349 train_time:89000ms step_avg:144.72ms step:625/3200 val_loss:3.7592 train_time:89055ms step_avg:144.81ms step:626/3200 train_loss:3.8527 train_time:89152ms step_avg:144.73ms step:627/3200 train_loss:3.8761 train_time:89300ms step_avg:144.73ms step:628/3200 train_loss:3.8593 train_time:89442ms step_avg:144.73ms step:629/3200 train_loss:3.8997 train_time:89584ms step_avg:144.72ms step:630/3200 train_loss:3.7237 train_time:89726ms step_avg:144.72ms step:631/3200 train_loss:3.8564 train_time:89868ms step_avg:144.72ms step:632/3200 train_loss:3.8842 train_time:90010ms step_avg:144.71ms step:633/3200 train_loss:3.7867 train_time:90158ms step_avg:144.72ms step:634/3200 train_loss:3.7246 train_time:90304ms step_avg:144.72ms step:635/3200 train_loss:3.8197 train_time:90449ms step_avg:144.72ms step:636/3200 train_loss:4.0782 train_time:90592ms step_avg:144.71ms step:637/3200 train_loss:3.6703 train_time:90733ms step_avg:144.71ms step:638/3200 train_loss:3.4877 train_time:90875ms step_avg:144.71ms step:639/3200 train_loss:3.7180 train_time:91017ms step_avg:144.70ms step:640/3200 train_loss:3.7534 train_time:91162ms step_avg:144.70ms step:641/3200 train_loss:3.7014 train_time:91307ms step_avg:144.70ms step:642/3200 train_loss:3.7094 train_time:91450ms step_avg:144.70ms step:643/3200 train_loss:3.7560 train_time:91592ms step_avg:144.70ms step:644/3200 train_loss:3.7531 train_time:91735ms step_avg:144.69ms step:645/3200 train_loss:3.6871 train_time:91877ms step_avg:144.69ms step:646/3200 train_loss:3.9096 train_time:92020ms step_avg:144.69ms step:647/3200 train_loss:3.8072 train_time:92164ms step_avg:144.68ms step:648/3200 train_loss:3.8066 train_time:92308ms step_avg:144.68ms step:649/3200 train_loss:3.8337 train_time:92450ms step_avg:144.68ms step:650/3200 train_loss:3.8985 train_time:92594ms step_avg:144.68ms step:651/3200 train_loss:3.7544 train_time:92737ms step_avg:144.67ms step:652/3200 train_loss:3.8981 train_time:92880ms step_avg:144.67ms step:653/3200 train_loss:3.7186 train_time:93023ms step_avg:144.67ms step:654/3200 train_loss:3.7973 train_time:93166ms step_avg:144.67ms step:655/3200 train_loss:3.5607 train_time:93308ms step_avg:144.66ms step:656/3200 train_loss:3.7135 train_time:93451ms step_avg:144.66ms step:657/3200 train_loss:3.7118 train_time:93595ms step_avg:144.66ms step:658/3200 train_loss:3.6420 train_time:93738ms step_avg:144.66ms step:659/3200 train_loss:3.8221 train_time:93879ms step_avg:144.65ms step:660/3200 train_loss:3.7216 train_time:94022ms step_avg:144.65ms step:661/3200 train_loss:3.8106 train_time:94165ms step_avg:144.65ms step:662/3200 train_loss:3.8873 train_time:94308ms step_avg:144.64ms step:663/3200 train_loss:3.8047 train_time:94451ms step_avg:144.64ms step:664/3200 train_loss:3.6791 train_time:94595ms step_avg:144.64ms step:665/3200 train_loss:3.7646 train_time:94738ms step_avg:144.64ms step:666/3200 train_loss:3.6313 train_time:94881ms step_avg:144.64ms step:667/3200 train_loss:3.9135 train_time:95025ms step_avg:144.63ms step:668/3200 train_loss:3.7525 train_time:95169ms step_avg:144.63ms step:669/3200 train_loss:3.7650 train_time:95311ms step_avg:144.63ms step:670/3200 train_loss:3.6194 train_time:95453ms step_avg:144.63ms step:671/3200 train_loss:3.7332 train_time:95596ms step_avg:144.62ms step:672/3200 train_loss:3.6896 train_time:95741ms step_avg:144.62ms step:673/3200 train_loss:3.7107 train_time:95884ms step_avg:144.62ms step:674/3200 train_loss:3.9856 train_time:96029ms step_avg:144.62ms step:675/3200 train_loss:3.7756 train_time:96172ms step_avg:144.62ms step:676/3200 train_loss:3.8414 train_time:96315ms step_avg:144.62ms step:677/3200 train_loss:3.6292 train_time:96458ms step_avg:144.61ms step:678/3200 train_loss:3.7319 train_time:96601ms step_avg:144.61ms step:679/3200 train_loss:3.6833 train_time:96744ms step_avg:144.61ms step:680/3200 train_loss:3.8226 train_time:96889ms step_avg:144.61ms step:681/3200 train_loss:3.7156 train_time:97033ms step_avg:144.61ms step:682/3200 train_loss:3.7487 train_time:97177ms step_avg:144.61ms step:683/3200 train_loss:3.8322 train_time:97320ms step_avg:144.61ms step:684/3200 train_loss:3.8678 train_time:97462ms step_avg:144.60ms step:685/3200 train_loss:3.7686 train_time:97606ms step_avg:144.60ms step:686/3200 train_loss:3.8329 train_time:97749ms step_avg:144.60ms step:687/3200 train_loss:3.7743 train_time:97891ms step_avg:144.60ms step:688/3200 train_loss:3.8163 train_time:98034ms step_avg:144.59ms step:689/3200 train_loss:3.4163 train_time:98178ms step_avg:144.59ms step:690/3200 train_loss:3.5528 train_time:98321ms step_avg:144.59ms step:691/3200 train_loss:3.6911 train_time:98465ms step_avg:144.59ms step:692/3200 train_loss:3.5714 train_time:98609ms step_avg:144.59ms step:693/3200 train_loss:3.7835 train_time:98751ms step_avg:144.58ms step:694/3200 train_loss:3.7969 train_time:98893ms step_avg:144.58ms step:695/3200 train_loss:3.6847 train_time:99037ms step_avg:144.58ms step:696/3200 train_loss:3.6796 train_time:99181ms step_avg:144.58ms step:697/3200 train_loss:3.9925 train_time:99325ms step_avg:144.58ms step:698/3200 train_loss:3.7435 train_time:99468ms step_avg:144.58ms step:699/3200 train_loss:3.7777 train_time:99611ms step_avg:144.57ms step:700/3200 train_loss:3.9452 train_time:99754ms step_avg:144.57ms step:701/3200 train_loss:3.7126 train_time:99897ms step_avg:144.57ms step:702/3200 train_loss:3.6761 train_time:100041ms step_avg:144.57ms step:703/3200 train_loss:3.6573 train_time:100186ms step_avg:144.57ms step:704/3200 train_loss:3.6201 train_time:100328ms step_avg:144.57ms step:705/3200 train_loss:3.7038 train_time:100472ms step_avg:144.56ms step:706/3200 train_loss:3.6991 train_time:100615ms step_avg:144.56ms step:707/3200 train_loss:3.7111 train_time:100758ms step_avg:144.56ms step:708/3200 train_loss:3.7766 train_time:100902ms step_avg:144.56ms step:709/3200 train_loss:3.7338 train_time:101046ms step_avg:144.56ms step:710/3200 train_loss:3.7138 train_time:101188ms step_avg:144.55ms step:711/3200 train_loss:3.6788 train_time:101331ms step_avg:144.55ms step:712/3200 train_loss:3.7283 train_time:101474ms step_avg:144.55ms step:713/3200 train_loss:3.7797 train_time:101617ms step_avg:144.55ms step:714/3200 train_loss:3.7912 train_time:101760ms step_avg:144.55ms step:715/3200 train_loss:3.7095 train_time:101903ms step_avg:144.54ms step:716/3200 train_loss:3.7124 train_time:102048ms step_avg:144.54ms step:717/3200 train_loss:3.7265 train_time:102192ms step_avg:144.54ms step:718/3200 train_loss:3.8677 train_time:102335ms step_avg:144.54ms step:719/3200 train_loss:3.7300 train_time:102478ms step_avg:144.54ms step:720/3200 train_loss:3.8043 train_time:102622ms step_avg:144.54ms step:721/3200 train_loss:3.9729 train_time:102765ms step_avg:144.54ms step:722/3200 train_loss:3.5974 train_time:102909ms step_avg:144.53ms step:723/3200 train_loss:3.8617 train_time:103052ms step_avg:144.53ms step:724/3200 train_loss:3.9169 train_time:103194ms step_avg:144.53ms step:725/3200 train_loss:3.7047 train_time:103338ms step_avg:144.53ms step:726/3200 train_loss:3.7814 train_time:103481ms step_avg:144.53ms step:727/3200 train_loss:3.6804 train_time:103624ms step_avg:144.53ms step:728/3200 train_loss:3.6998 train_time:103769ms step_avg:144.52ms step:729/3200 train_loss:3.8724 train_time:103911ms step_avg:144.52ms step:730/3200 train_loss:3.8213 train_time:104054ms step_avg:144.52ms step:731/3200 train_loss:3.8075 train_time:104197ms step_avg:144.52ms step:732/3200 train_loss:3.6996 train_time:104340ms step_avg:144.52ms step:733/3200 train_loss:3.7254 train_time:104484ms step_avg:144.51ms step:734/3200 train_loss:3.9618 train_time:104628ms step_avg:144.51ms step:735/3200 train_loss:3.6946 train_time:104772ms step_avg:144.51ms step:736/3200 train_loss:3.7607 train_time:104913ms step_avg:144.51ms step:737/3200 train_loss:3.8788 train_time:105056ms step_avg:144.51ms step:738/3200 train_loss:3.8020 train_time:105199ms step_avg:144.50ms step:739/3200 train_loss:3.7388 train_time:105342ms step_avg:144.50ms step:740/3200 train_loss:3.6350 train_time:105485ms step_avg:144.50ms step:741/3200 train_loss:4.2734 train_time:105629ms step_avg:144.50ms step:742/3200 train_loss:3.6382 train_time:105771ms step_avg:144.50ms step:743/3200 train_loss:3.7125 train_time:105914ms step_avg:144.49ms step:744/3200 train_loss:3.7234 train_time:106058ms step_avg:144.49ms step:745/3200 train_loss:3.7806 train_time:106202ms step_avg:144.49ms step:746/3200 train_loss:3.7474 train_time:106347ms step_avg:144.49ms step:747/3200 train_loss:3.7325 train_time:106489ms step_avg:144.49ms step:748/3200 train_loss:3.7700 train_time:106632ms step_avg:144.49ms step:749/3200 train_loss:3.6990 train_time:106775ms step_avg:144.49ms step:750/3200 train_loss:3.6994 train_time:106918ms step_avg:144.48ms step:750/3200 val_loss:3.7090 train_time:106972ms step_avg:144.56ms step:751/3200 train_loss:3.7348 train_time:107070ms step_avg:144.49ms step:752/3200 train_loss:3.6964 train_time:107216ms step_avg:144.50ms step:753/3200 train_loss:3.7398 train_time:107358ms step_avg:144.49ms step:754/3200 train_loss:3.7573 train_time:107501ms step_avg:144.49ms step:755/3200 train_loss:3.7295 train_time:107643ms step_avg:144.49ms step:756/3200 train_loss:3.8049 train_time:107947ms step_avg:144.70ms step:757/3200 train_loss:3.6279 train_time:108088ms step_avg:144.70ms step:758/3200 train_loss:3.8692 train_time:108232ms step_avg:144.70ms step:759/3200 train_loss:3.7864 train_time:108375ms step_avg:144.69ms step:760/3200 train_loss:3.7208 train_time:108686ms step_avg:144.91ms step:761/3200 train_loss:3.8300 train_time:108828ms step_avg:144.91ms step:762/3200 train_loss:3.5462 train_time:108969ms step_avg:144.91ms step:763/3200 train_loss:3.6902 train_time:109110ms step_avg:144.90ms step:764/3200 train_loss:3.8028 train_time:109251ms step_avg:144.90ms step:765/3200 train_loss:3.4589 train_time:109393ms step_avg:144.89ms step:766/3200 train_loss:3.8865 train_time:109535ms step_avg:144.89ms step:767/3200 train_loss:3.7324 train_time:109683ms step_avg:144.89ms step:768/3200 train_loss:3.6971 train_time:109827ms step_avg:144.89ms step:769/3200 train_loss:3.7150 train_time:109970ms step_avg:144.89ms step:770/3200 train_loss:3.7411 train_time:110111ms step_avg:144.88ms step:771/3200 train_loss:3.7923 train_time:110253ms step_avg:144.88ms step:772/3200 train_loss:4.0213 train_time:110394ms step_avg:144.87ms step:773/3200 train_loss:3.5962 train_time:110538ms step_avg:144.87ms step:774/3200 train_loss:3.7900 train_time:110683ms step_avg:144.87ms step:775/3200 train_loss:3.7787 train_time:110828ms step_avg:144.87ms step:776/3200 train_loss:3.7463 train_time:110971ms step_avg:144.87ms step:777/3200 train_loss:3.5483 train_time:111113ms step_avg:144.87ms step:778/3200 train_loss:3.5460 train_time:111255ms step_avg:144.86ms step:779/3200 train_loss:3.6182 train_time:111398ms step_avg:144.86ms step:780/3200 train_loss:3.7078 train_time:111541ms step_avg:144.86ms step:781/3200 train_loss:3.7382 train_time:111685ms step_avg:144.86ms step:782/3200 train_loss:3.8020 train_time:111828ms step_avg:144.85ms step:783/3200 train_loss:3.7117 train_time:111972ms step_avg:144.85ms step:784/3200 train_loss:3.7061 train_time:112114ms step_avg:144.85ms step:785/3200 train_loss:3.7178 train_time:112256ms step_avg:144.85ms step:786/3200 train_loss:3.6928 train_time:112400ms step_avg:144.85ms step:787/3200 train_loss:3.5954 train_time:112543ms step_avg:144.84ms step:788/3200 train_loss:3.8411 train_time:112687ms step_avg:144.84ms step:789/3200 train_loss:3.6416 train_time:112829ms step_avg:144.84ms step:790/3200 train_loss:3.7018 train_time:112974ms step_avg:144.84ms step:791/3200 train_loss:3.7680 train_time:113117ms step_avg:144.84ms step:792/3200 train_loss:3.8969 train_time:113260ms step_avg:144.83ms step:793/3200 train_loss:3.9020 train_time:113404ms step_avg:144.83ms step:794/3200 train_loss:3.6174 train_time:113546ms step_avg:144.83ms step:795/3200 train_loss:3.7431 train_time:113688ms step_avg:144.83ms step:796/3200 train_loss:3.7995 train_time:113832ms step_avg:144.82ms step:797/3200 train_loss:3.9158 train_time:113975ms step_avg:144.82ms step:798/3200 train_loss:3.6535 train_time:114118ms step_avg:144.82ms step:799/3200 train_loss:3.8003 train_time:114262ms step_avg:144.82ms step:800/3200 train_loss:3.6892 train_time:114405ms step_avg:144.82ms step:801/3200 train_loss:3.6828 train_time:114547ms step_avg:144.81ms step:802/3200 train_loss:3.7736 train_time:114690ms step_avg:144.81ms step:803/3200 train_loss:3.6379 train_time:114833ms step_avg:144.81ms step:804/3200 train_loss:3.6530 train_time:114976ms step_avg:144.81ms step:805/3200 train_loss:3.7710 train_time:115118ms step_avg:144.80ms step:806/3200 train_loss:3.6732 train_time:115262ms step_avg:144.80ms step:807/3200 train_loss:3.6859 train_time:115404ms step_avg:144.80ms step:808/3200 train_loss:3.7801 train_time:115547ms step_avg:144.80ms step:809/3200 train_loss:3.7004 train_time:115689ms step_avg:144.79ms step:810/3200 train_loss:3.6243 train_time:115833ms step_avg:144.79ms step:811/3200 train_loss:3.7046 train_time:115977ms step_avg:144.79ms step:812/3200 train_loss:3.7400 train_time:116122ms step_avg:144.79ms step:813/3200 train_loss:3.7295 train_time:116265ms step_avg:144.79ms step:814/3200 train_loss:3.7711 train_time:116408ms step_avg:144.79ms step:815/3200 train_loss:3.7126 train_time:116551ms step_avg:144.78ms step:816/3200 train_loss:3.6999 train_time:116693ms step_avg:144.78ms step:817/3200 train_loss:3.8064 train_time:116836ms step_avg:144.78ms step:818/3200 train_loss:3.8949 train_time:116981ms step_avg:144.78ms step:819/3200 train_loss:3.6609 train_time:117125ms step_avg:144.78ms step:820/3200 train_loss:3.8637 train_time:117267ms step_avg:144.77ms step:821/3200 train_loss:3.6448 train_time:117410ms step_avg:144.77ms step:822/3200 train_loss:3.6881 train_time:117552ms step_avg:144.77ms step:823/3200 train_loss:3.8143 train_time:117695ms step_avg:144.77ms step:824/3200 train_loss:3.7221 train_time:117838ms step_avg:144.76ms step:825/3200 train_loss:3.6560 train_time:117982ms step_avg:144.76ms step:826/3200 train_loss:3.7564 train_time:118127ms step_avg:144.76ms step:827/3200 train_loss:3.6363 train_time:118270ms step_avg:144.76ms step:828/3200 train_loss:3.8730 train_time:118413ms step_avg:144.76ms step:829/3200 train_loss:3.7596 train_time:118555ms step_avg:144.76ms step:830/3200 train_loss:3.8100 train_time:118698ms step_avg:144.75ms step:831/3200 train_loss:3.6764 train_time:118842ms step_avg:144.75ms step:832/3200 train_loss:3.7270 train_time:118985ms step_avg:144.75ms step:833/3200 train_loss:3.6564 train_time:119128ms step_avg:144.75ms step:834/3200 train_loss:3.7886 train_time:119272ms step_avg:144.75ms step:835/3200 train_loss:3.6230 train_time:119413ms step_avg:144.74ms step:836/3200 train_loss:3.5971 train_time:119556ms step_avg:144.74ms step:837/3200 train_loss:3.8594 train_time:119698ms step_avg:144.74ms step:838/3200 train_loss:3.5549 train_time:119842ms step_avg:144.74ms step:839/3200 train_loss:3.7310 train_time:119985ms step_avg:144.73ms step:840/3200 train_loss:3.5753 train_time:120129ms step_avg:144.73ms step:841/3200 train_loss:3.6141 train_time:120273ms step_avg:144.73ms step:842/3200 train_loss:3.6998 train_time:120415ms step_avg:144.73ms step:843/3200 train_loss:3.7276 train_time:120558ms step_avg:144.73ms step:844/3200 train_loss:3.7166 train_time:120702ms step_avg:144.73ms step:845/3200 train_loss:3.5732 train_time:120844ms step_avg:144.72ms step:846/3200 train_loss:3.8105 train_time:120987ms step_avg:144.72ms step:847/3200 train_loss:3.6713 train_time:121131ms step_avg:144.72ms step:848/3200 train_loss:3.6342 train_time:121274ms step_avg:144.72ms step:849/3200 train_loss:3.7745 train_time:121417ms step_avg:144.72ms step:850/3200 train_loss:3.6348 train_time:121560ms step_avg:144.71ms step:851/3200 train_loss:3.5895 train_time:121706ms step_avg:144.72ms step:852/3200 train_loss:3.8795 train_time:121848ms step_avg:144.71ms step:853/3200 train_loss:3.5901 train_time:121991ms step_avg:144.71ms step:854/3200 train_loss:3.7067 train_time:122134ms step_avg:144.71ms step:855/3200 train_loss:3.7920 train_time:122279ms step_avg:144.71ms step:856/3200 train_loss:3.6684 train_time:122423ms step_avg:144.71ms step:857/3200 train_loss:3.6955 train_time:122566ms step_avg:144.71ms step:858/3200 train_loss:3.7438 train_time:122708ms step_avg:144.70ms step:859/3200 train_loss:3.6272 train_time:122850ms step_avg:144.70ms step:860/3200 train_loss:3.7073 train_time:122994ms step_avg:144.70ms step:861/3200 train_loss:3.7354 train_time:123138ms step_avg:144.70ms step:862/3200 train_loss:3.7808 train_time:123282ms step_avg:144.70ms step:863/3200 train_loss:3.7393 train_time:123426ms step_avg:144.70ms step:864/3200 train_loss:3.7167 train_time:123568ms step_avg:144.69ms step:865/3200 train_loss:3.5385 train_time:123711ms step_avg:144.69ms step:866/3200 train_loss:3.7324 train_time:123854ms step_avg:144.69ms step:867/3200 train_loss:4.0065 train_time:123997ms step_avg:144.69ms step:868/3200 train_loss:3.5967 train_time:124141ms step_avg:144.69ms step:869/3200 train_loss:3.7765 train_time:124287ms step_avg:144.69ms step:870/3200 train_loss:3.7590 train_time:124429ms step_avg:144.69ms step:871/3200 train_loss:3.5931 train_time:124573ms step_avg:144.68ms step:872/3200 train_loss:3.5626 train_time:124715ms step_avg:144.68ms step:873/3200 train_loss:3.8055 train_time:124857ms step_avg:144.68ms step:874/3200 train_loss:3.5945 train_time:125001ms step_avg:144.68ms step:875/3200 train_loss:3.3256 train_time:125143ms step_avg:144.67ms step:875/3200 val_loss:3.6673 train_time:125198ms step_avg:144.74ms step:876/3200 train_loss:3.7821 train_time:125296ms step_avg:144.68ms step:877/3200 train_loss:3.5887 train_time:125443ms step_avg:144.69ms step:878/3200 train_loss:3.7669 train_time:125587ms step_avg:144.69ms step:879/3200 train_loss:3.6265 train_time:125730ms step_avg:144.68ms step:880/3200 train_loss:3.8025 train_time:125870ms step_avg:144.68ms step:881/3200 train_loss:3.4623 train_time:126012ms step_avg:144.68ms step:882/3200 train_loss:3.6369 train_time:126154ms step_avg:144.67ms step:883/3200 train_loss:3.8262 train_time:126300ms step_avg:144.67ms step:884/3200 train_loss:3.9898 train_time:126447ms step_avg:144.68ms step:885/3200 train_loss:3.7096 train_time:126590ms step_avg:144.67ms step:886/3200 train_loss:3.6236 train_time:126733ms step_avg:144.67ms step:887/3200 train_loss:3.7176 train_time:126874ms step_avg:144.67ms step:888/3200 train_loss:4.2210 train_time:127016ms step_avg:144.67ms step:889/3200 train_loss:3.9808 train_time:127160ms step_avg:144.66ms step:890/3200 train_loss:3.6605 train_time:127304ms step_avg:144.66ms step:891/3200 train_loss:3.6764 train_time:127448ms step_avg:144.66ms step:892/3200 train_loss:3.5086 train_time:127593ms step_avg:144.66ms step:893/3200 train_loss:3.8578 train_time:127736ms step_avg:144.66ms step:894/3200 train_loss:3.5696 train_time:127878ms step_avg:144.66ms step:895/3200 train_loss:3.8244 train_time:128020ms step_avg:144.66ms step:896/3200 train_loss:3.8382 train_time:128163ms step_avg:144.65ms step:897/3200 train_loss:3.6382 train_time:128307ms step_avg:144.65ms step:898/3200 train_loss:3.6824 train_time:128450ms step_avg:144.65ms step:899/3200 train_loss:3.7369 train_time:128596ms step_avg:144.65ms step:900/3200 train_loss:3.6203 train_time:128741ms step_avg:144.65ms step:901/3200 train_loss:3.5630 train_time:128885ms step_avg:144.65ms step:902/3200 train_loss:3.7723 train_time:129028ms step_avg:144.65ms step:903/3200 train_loss:3.7728 train_time:129170ms step_avg:144.65ms step:904/3200 train_loss:3.6758 train_time:129313ms step_avg:144.64ms step:905/3200 train_loss:3.6461 train_time:129456ms step_avg:144.64ms step:906/3200 train_loss:3.6345 train_time:129600ms step_avg:144.64ms step:907/3200 train_loss:3.8596 train_time:129745ms step_avg:144.64ms step:908/3200 train_loss:3.6622 train_time:129888ms step_avg:144.64ms step:909/3200 train_loss:3.6939 train_time:130030ms step_avg:144.64ms step:910/3200 train_loss:3.6052 train_time:130172ms step_avg:144.64ms step:911/3200 train_loss:3.6920 train_time:130315ms step_avg:144.63ms step:912/3200 train_loss:3.7634 train_time:130459ms step_avg:144.63ms step:913/3200 train_loss:3.7506 train_time:130603ms step_avg:144.63ms step:914/3200 train_loss:3.6291 train_time:130747ms step_avg:144.63ms step:915/3200 train_loss:3.8784 train_time:130890ms step_avg:144.63ms step:916/3200 train_loss:3.6827 train_time:131034ms step_avg:144.63ms step:917/3200 train_loss:3.7679 train_time:131176ms step_avg:144.63ms step:918/3200 train_loss:3.7449 train_time:131319ms step_avg:144.62ms step:919/3200 train_loss:4.9638 train_time:131463ms step_avg:144.62ms step:920/3200 train_loss:3.6604 train_time:131606ms step_avg:144.62ms step:921/3200 train_loss:3.7231 train_time:131750ms step_avg:144.62ms step:922/3200 train_loss:3.6810 train_time:131894ms step_avg:144.62ms step:923/3200 train_loss:3.7269 train_time:132039ms step_avg:144.62ms step:924/3200 train_loss:3.7401 train_time:132182ms step_avg:144.62ms step:925/3200 train_loss:3.8308 train_time:132325ms step_avg:144.62ms step:926/3200 train_loss:3.8039 train_time:132468ms step_avg:144.62ms step:927/3200 train_loss:3.7024 train_time:132610ms step_avg:144.61ms step:928/3200 train_loss:3.6881 train_time:132754ms step_avg:144.61ms step:929/3200 train_loss:3.9202 train_time:132899ms step_avg:144.61ms step:930/3200 train_loss:3.7557 train_time:133044ms step_avg:144.61ms step:931/3200 train_loss:3.5484 train_time:133188ms step_avg:144.61ms step:932/3200 train_loss:3.6386 train_time:133330ms step_avg:144.61ms step:933/3200 train_loss:3.8180 train_time:133473ms step_avg:144.61ms step:934/3200 train_loss:3.5387 train_time:133616ms step_avg:144.61ms step:935/3200 train_loss:3.7141 train_time:133760ms step_avg:144.61ms step:936/3200 train_loss:3.5889 train_time:133905ms step_avg:144.61ms step:937/3200 train_loss:3.6595 train_time:134048ms step_avg:144.60ms step:938/3200 train_loss:3.7582 train_time:134191ms step_avg:144.60ms step:939/3200 train_loss:3.6808 train_time:134334ms step_avg:144.60ms step:940/3200 train_loss:3.8466 train_time:134477ms step_avg:144.60ms step:941/3200 train_loss:3.6292 train_time:134620ms step_avg:144.60ms step:942/3200 train_loss:3.6917 train_time:134763ms step_avg:144.60ms step:943/3200 train_loss:3.4952 train_time:134908ms step_avg:144.60ms step:944/3200 train_loss:3.8488 train_time:135051ms step_avg:144.59ms step:945/3200 train_loss:3.5558 train_time:135349ms step_avg:144.76ms step:946/3200 train_loss:3.5712 train_time:135491ms step_avg:144.76ms step:947/3200 train_loss:5.1972 train_time:135633ms step_avg:144.75ms step:948/3200 train_loss:3.7430 train_time:135774ms step_avg:144.75ms step:949/3200 train_loss:3.6381 train_time:135915ms step_avg:144.74ms step:950/3200 train_loss:3.5341 train_time:136224ms step_avg:144.92ms step:951/3200 train_loss:3.5944 train_time:136366ms step_avg:144.92ms step:952/3200 train_loss:3.5503 train_time:136507ms step_avg:144.91ms step:953/3200 train_loss:3.6238 train_time:136649ms step_avg:144.91ms step:954/3200 train_loss:3.7017 train_time:136790ms step_avg:144.91ms step:955/3200 train_loss:3.5815 train_time:136932ms step_avg:144.90ms step:956/3200 train_loss:3.6204 train_time:137074ms step_avg:144.90ms step:957/3200 train_loss:3.5889 train_time:137222ms step_avg:144.90ms step:958/3200 train_loss:3.6445 train_time:137366ms step_avg:144.90ms step:959/3200 train_loss:3.6398 train_time:137510ms step_avg:144.90ms step:960/3200 train_loss:3.6550 train_time:137652ms step_avg:144.90ms step:961/3200 train_loss:3.5431 train_time:137794ms step_avg:144.89ms step:962/3200 train_loss:3.7925 train_time:137935ms step_avg:144.89ms step:963/3200 train_loss:3.7456 train_time:138078ms step_avg:144.89ms step:964/3200 train_loss:3.6806 train_time:138224ms step_avg:144.89ms step:965/3200 train_loss:3.5925 train_time:138369ms step_avg:144.89ms step:966/3200 train_loss:3.6276 train_time:138513ms step_avg:144.89ms step:967/3200 train_loss:3.8501 train_time:138655ms step_avg:144.89ms step:968/3200 train_loss:3.6717 train_time:138798ms step_avg:144.88ms step:969/3200 train_loss:3.6617 train_time:138941ms step_avg:144.88ms step:970/3200 train_loss:3.7214 train_time:139085ms step_avg:144.88ms step:971/3200 train_loss:3.5386 train_time:139227ms step_avg:144.88ms step:972/3200 train_loss:3.6926 train_time:139372ms step_avg:144.88ms step:973/3200 train_loss:3.6429 train_time:139515ms step_avg:144.88ms step:974/3200 train_loss:3.6830 train_time:139658ms step_avg:144.87ms step:975/3200 train_loss:3.7581 train_time:139801ms step_avg:144.87ms step:976/3200 train_loss:3.6256 train_time:139944ms step_avg:144.87ms step:977/3200 train_loss:3.8292 train_time:140087ms step_avg:144.87ms step:978/3200 train_loss:3.7109 train_time:140231ms step_avg:144.87ms step:979/3200 train_loss:3.5398 train_time:140375ms step_avg:144.87ms step:980/3200 train_loss:3.8343 train_time:140518ms step_avg:144.86ms step:981/3200 train_loss:3.5648 train_time:140662ms step_avg:144.86ms step:982/3200 train_loss:3.7298 train_time:140805ms step_avg:144.86ms step:983/3200 train_loss:3.7108 train_time:140948ms step_avg:144.86ms step:984/3200 train_loss:3.7115 train_time:141091ms step_avg:144.86ms step:985/3200 train_loss:3.6509 train_time:141234ms step_avg:144.86ms step:986/3200 train_loss:3.7460 train_time:141377ms step_avg:144.85ms step:987/3200 train_loss:3.5649 train_time:141521ms step_avg:144.85ms step:988/3200 train_loss:3.6398 train_time:141667ms step_avg:144.85ms step:989/3200 train_loss:3.6484 train_time:141810ms step_avg:144.85ms step:990/3200 train_loss:3.5781 train_time:141952ms step_avg:144.85ms step:991/3200 train_loss:3.8021 train_time:142095ms step_avg:144.85ms step:992/3200 train_loss:3.6178 train_time:142239ms step_avg:144.85ms step:993/3200 train_loss:3.5862 train_time:142381ms step_avg:144.84ms step:994/3200 train_loss:3.6576 train_time:142526ms step_avg:144.84ms step:995/3200 train_loss:3.7457 train_time:142669ms step_avg:144.84ms step:996/3200 train_loss:3.6948 train_time:142812ms step_avg:144.84ms step:997/3200 train_loss:3.6013 train_time:142955ms step_avg:144.84ms step:998/3200 train_loss:3.9511 train_time:143098ms step_avg:144.84ms step:999/3200 train_loss:3.6070 train_time:143242ms step_avg:144.84ms step:1000/3200 train_loss:3.7364 train_time:143385ms step_avg:144.83ms step:1000/3200 val_loss:3.6323 train_time:143440ms step_avg:144.89ms step:1001/3200 train_loss:3.6064 train_time:143538ms step_avg:144.84ms step:1002/3200 train_loss:3.6585 train_time:143686ms step_avg:144.85ms step:1003/3200 train_loss:3.5395 train_time:143828ms step_avg:144.84ms step:1004/3200 train_loss:3.7269 train_time:143970ms step_avg:144.84ms step:1005/3200 train_loss:3.7737 train_time:144112ms step_avg:144.84ms step:1006/3200 train_loss:3.5493 train_time:144254ms step_avg:144.83ms step:1007/3200 train_loss:3.6313 train_time:144396ms step_avg:144.83ms step:1008/3200 train_loss:3.5981 train_time:144542ms step_avg:144.83ms step:1009/3200 train_loss:3.7203 train_time:144689ms step_avg:144.83ms step:1010/3200 train_loss:3.8193 train_time:144833ms step_avg:144.83ms step:1011/3200 train_loss:3.7216 train_time:144976ms step_avg:144.83ms step:1012/3200 train_loss:3.6838 train_time:145117ms step_avg:144.83ms step:1013/3200 train_loss:3.5425 train_time:145260ms step_avg:144.83ms step:1014/3200 train_loss:3.6831 train_time:145402ms step_avg:144.82ms step:1015/3200 train_loss:3.7934 train_time:145545ms step_avg:144.82ms step:1016/3200 train_loss:3.5040 train_time:145690ms step_avg:144.82ms step:1017/3200 train_loss:3.5968 train_time:145834ms step_avg:144.82ms step:1018/3200 train_loss:3.5877 train_time:145978ms step_avg:144.82ms step:1019/3200 train_loss:3.5414 train_time:146120ms step_avg:144.82ms step:1020/3200 train_loss:3.6868 train_time:146263ms step_avg:144.82ms step:1021/3200 train_loss:3.5887 train_time:146406ms step_avg:144.81ms step:1022/3200 train_loss:3.5259 train_time:146549ms step_avg:144.81ms step:1023/3200 train_loss:3.6319 train_time:146693ms step_avg:144.81ms step:1024/3200 train_loss:3.6604 train_time:146836ms step_avg:144.81ms step:1025/3200 train_loss:3.6385 train_time:146979ms step_avg:144.81ms step:1026/3200 train_loss:3.6478 train_time:147122ms step_avg:144.81ms step:1027/3200 train_loss:3.8131 train_time:147265ms step_avg:144.80ms step:1028/3200 train_loss:3.4916 train_time:147409ms step_avg:144.80ms step:1029/3200 train_loss:3.5614 train_time:147552ms step_avg:144.80ms step:1030/3200 train_loss:3.5114 train_time:147696ms step_avg:144.80ms step:1031/3200 train_loss:3.6799 train_time:147841ms step_avg:144.80ms step:1032/3200 train_loss:3.6623 train_time:147986ms step_avg:144.80ms step:1033/3200 train_loss:3.8400 train_time:148129ms step_avg:144.80ms step:1034/3200 train_loss:3.6563 train_time:148272ms step_avg:144.80ms step:1035/3200 train_loss:3.5852 train_time:148415ms step_avg:144.80ms step:1036/3200 train_loss:3.5977 train_time:148558ms step_avg:144.79ms step:1037/3200 train_loss:3.6584 train_time:148701ms step_avg:144.79ms step:1038/3200 train_loss:3.9614 train_time:148845ms step_avg:144.79ms step:1039/3200 train_loss:3.7902 train_time:148989ms step_avg:144.79ms step:1040/3200 train_loss:3.6829 train_time:149131ms step_avg:144.79ms step:1041/3200 train_loss:3.5740 train_time:149274ms step_avg:144.79ms step:1042/3200 train_loss:3.6505 train_time:149416ms step_avg:144.78ms step:1043/3200 train_loss:3.6850 train_time:149560ms step_avg:144.78ms step:1044/3200 train_loss:3.6135 train_time:149703ms step_avg:144.78ms step:1045/3200 train_loss:3.6218 train_time:149847ms step_avg:144.78ms step:1046/3200 train_loss:3.7053 train_time:149990ms step_avg:144.78ms step:1047/3200 train_loss:3.6067 train_time:150133ms step_avg:144.78ms step:1048/3200 train_loss:3.8075 train_time:150276ms step_avg:144.77ms step:1049/3200 train_loss:3.6679 train_time:150419ms step_avg:144.77ms step:1050/3200 train_loss:3.5911 train_time:150563ms step_avg:144.77ms step:1051/3200 train_loss:3.5530 train_time:150708ms step_avg:144.77ms step:1052/3200 train_loss:3.6815 train_time:150850ms step_avg:144.77ms step:1053/3200 train_loss:3.5533 train_time:150994ms step_avg:144.77ms step:1054/3200 train_loss:3.8727 train_time:151136ms step_avg:144.77ms step:1055/3200 train_loss:3.7092 train_time:151279ms step_avg:144.76ms step:1056/3200 train_loss:3.5720 train_time:151423ms step_avg:144.76ms step:1057/3200 train_loss:3.6721 train_time:151567ms step_avg:144.76ms step:1058/3200 train_loss:3.7444 train_time:151711ms step_avg:144.76ms step:1059/3200 train_loss:3.4698 train_time:151853ms step_avg:144.76ms step:1060/3200 train_loss:3.5898 train_time:151998ms step_avg:144.76ms step:1061/3200 train_loss:3.6195 train_time:152140ms step_avg:144.76ms step:1062/3200 train_loss:3.5827 train_time:152284ms step_avg:144.76ms step:1063/3200 train_loss:3.5630 train_time:152429ms step_avg:144.76ms step:1064/3200 train_loss:3.6569 train_time:152572ms step_avg:144.76ms step:1065/3200 train_loss:3.5606 train_time:152715ms step_avg:144.75ms step:1066/3200 train_loss:3.5479 train_time:152859ms step_avg:144.75ms step:1067/3200 train_loss:3.5686 train_time:153002ms step_avg:144.75ms step:1068/3200 train_loss:3.4824 train_time:153145ms step_avg:144.75ms step:1069/3200 train_loss:3.5954 train_time:153288ms step_avg:144.75ms step:1070/3200 train_loss:3.4689 train_time:153431ms step_avg:144.75ms step:1071/3200 train_loss:3.7280 train_time:153574ms step_avg:144.75ms step:1072/3200 train_loss:3.6761 train_time:153718ms step_avg:144.74ms step:1073/3200 train_loss:3.6250 train_time:153861ms step_avg:144.74ms step:1074/3200 train_loss:3.6868 train_time:154005ms step_avg:144.74ms step:1075/3200 train_loss:3.6299 train_time:154149ms step_avg:144.74ms step:1076/3200 train_loss:3.5790 train_time:154291ms step_avg:144.74ms step:1077/3200 train_loss:3.9639 train_time:154435ms step_avg:144.74ms step:1078/3200 train_loss:3.6392 train_time:154578ms step_avg:144.74ms step:1079/3200 train_loss:3.3336 train_time:154722ms step_avg:144.74ms step:1080/3200 train_loss:3.7054 train_time:154866ms step_avg:144.73ms step:1081/3200 train_loss:3.6217 train_time:155010ms step_avg:144.73ms step:1082/3200 train_loss:3.6871 train_time:155153ms step_avg:144.73ms step:1083/3200 train_loss:3.7894 train_time:155296ms step_avg:144.73ms step:1084/3200 train_loss:3.6794 train_time:155439ms step_avg:144.73ms step:1085/3200 train_loss:3.6540 train_time:155581ms step_avg:144.73ms step:1086/3200 train_loss:3.6129 train_time:155725ms step_avg:144.73ms step:1087/3200 train_loss:3.8102 train_time:155868ms step_avg:144.72ms step:1088/3200 train_loss:3.6998 train_time:156011ms step_avg:144.72ms step:1089/3200 train_loss:3.5342 train_time:156153ms step_avg:144.72ms step:1090/3200 train_loss:3.5633 train_time:156297ms step_avg:144.72ms step:1091/3200 train_loss:3.6763 train_time:156440ms step_avg:144.72ms step:1092/3200 train_loss:3.4750 train_time:156583ms step_avg:144.72ms step:1093/3200 train_loss:3.6708 train_time:156728ms step_avg:144.72ms step:1094/3200 train_loss:3.7994 train_time:156872ms step_avg:144.72ms step:1095/3200 train_loss:3.6427 train_time:157014ms step_avg:144.71ms step:1096/3200 train_loss:3.5940 train_time:157157ms step_avg:144.71ms step:1097/3200 train_loss:3.6163 train_time:157300ms step_avg:144.71ms step:1098/3200 train_loss:3.6674 train_time:157447ms step_avg:144.71ms step:1099/3200 train_loss:3.7409 train_time:157585ms step_avg:144.71ms step:1100/3200 train_loss:3.6990 train_time:157729ms step_avg:144.71ms step:1101/3200 train_loss:3.6224 train_time:157873ms step_avg:144.70ms step:1102/3200 train_loss:3.4828 train_time:158016ms step_avg:144.70ms step:1103/3200 train_loss:3.5489 train_time:158159ms step_avg:144.70ms step:1104/3200 train_loss:3.6326 train_time:158302ms step_avg:144.70ms step:1105/3200 train_loss:3.5067 train_time:158445ms step_avg:144.70ms step:1106/3200 train_loss:4.2634 train_time:158589ms step_avg:144.70ms step:1107/3200 train_loss:3.4133 train_time:158732ms step_avg:144.70ms step:1108/3200 train_loss:3.7523 train_time:158875ms step_avg:144.69ms step:1109/3200 train_loss:3.5361 train_time:159019ms step_avg:144.69ms step:1110/3200 train_loss:3.6813 train_time:159162ms step_avg:144.69ms step:1111/3200 train_loss:3.6122 train_time:159306ms step_avg:144.69ms step:1112/3200 train_loss:3.6597 train_time:159449ms step_avg:144.69ms step:1113/3200 train_loss:3.7528 train_time:159592ms step_avg:144.69ms step:1114/3200 train_loss:3.6075 train_time:159734ms step_avg:144.69ms step:1115/3200 train_loss:3.5522 train_time:159877ms step_avg:144.69ms step:1116/3200 train_loss:3.4519 train_time:160020ms step_avg:144.68ms step:1117/3200 train_loss:3.6265 train_time:160165ms step_avg:144.68ms step:1118/3200 train_loss:3.7758 train_time:160309ms step_avg:144.68ms step:1119/3200 train_loss:3.8149 train_time:160451ms step_avg:144.68ms step:1120/3200 train_loss:3.6445 train_time:160595ms step_avg:144.68ms step:1121/3200 train_loss:3.6793 train_time:160737ms step_avg:144.68ms step:1122/3200 train_loss:3.5761 train_time:160880ms step_avg:144.68ms step:1123/3200 train_loss:3.6362 train_time:161024ms step_avg:144.68ms step:1124/3200 train_loss:3.7748 train_time:161169ms step_avg:144.68ms step:1125/3200 train_loss:3.5416 train_time:161312ms step_avg:144.67ms step:1125/3200 val_loss:3.6056 train_time:161367ms step_avg:144.72ms step:1126/3200 train_loss:3.4423 train_time:161463ms step_avg:144.68ms step:1127/3200 train_loss:3.6662 train_time:161611ms step_avg:144.68ms step:1128/3200 train_loss:3.8751 train_time:161754ms step_avg:144.68ms step:1129/3200 train_loss:3.4242 train_time:161896ms step_avg:144.68ms step:1130/3200 train_loss:3.7403 train_time:162039ms step_avg:144.68ms step:1131/3200 train_loss:3.5758 train_time:162180ms step_avg:144.67ms step:1132/3200 train_loss:3.6030 train_time:162322ms step_avg:144.67ms step:1133/3200 train_loss:3.5616 train_time:162468ms step_avg:144.67ms step:1134/3200 train_loss:3.7175 train_time:162771ms step_avg:144.81ms step:1135/3200 train_loss:3.6508 train_time:162913ms step_avg:144.81ms step:1136/3200 train_loss:3.6997 train_time:163055ms step_avg:144.81ms step:1137/3200 train_loss:3.7381 train_time:163198ms step_avg:144.81ms step:1138/3200 train_loss:3.6515 train_time:163341ms step_avg:144.81ms step:1139/3200 train_loss:3.5447 train_time:163482ms step_avg:144.80ms step:1140/3200 train_loss:3.8544 train_time:163790ms step_avg:144.95ms step:1141/3200 train_loss:3.6552 train_time:163933ms step_avg:144.95ms step:1142/3200 train_loss:3.7571 train_time:164075ms step_avg:144.94ms step:1143/3200 train_loss:3.6387 train_time:164217ms step_avg:144.94ms step:1144/3200 train_loss:3.5636 train_time:164358ms step_avg:144.94ms step:1145/3200 train_loss:3.6594 train_time:164500ms step_avg:144.93ms step:1146/3200 train_loss:3.7788 train_time:164642ms step_avg:144.93ms step:1147/3200 train_loss:3.7534 train_time:164791ms step_avg:144.94ms step:1148/3200 train_loss:3.6727 train_time:164934ms step_avg:144.93ms step:1149/3200 train_loss:3.6872 train_time:165078ms step_avg:144.93ms step:1150/3200 train_loss:3.5448 train_time:165220ms step_avg:144.93ms step:1151/3200 train_loss:3.5606 train_time:165363ms step_avg:144.93ms step:1152/3200 train_loss:3.5239 train_time:165504ms step_avg:144.93ms step:1153/3200 train_loss:3.6743 train_time:165648ms step_avg:144.92ms step:1154/3200 train_loss:3.6374 train_time:165795ms step_avg:144.93ms step:1155/3200 train_loss:3.7109 train_time:165938ms step_avg:144.92ms step:1156/3200 train_loss:3.5565 train_time:166082ms step_avg:144.92ms step:1157/3200 train_loss:3.7250 train_time:166224ms step_avg:144.92ms step:1158/3200 train_loss:3.6838 train_time:166366ms step_avg:144.92ms step:1159/3200 train_loss:3.4914 train_time:166509ms step_avg:144.92ms step:1160/3200 train_loss:3.5361 train_time:166653ms step_avg:144.92ms step:1161/3200 train_loss:3.5218 train_time:166797ms step_avg:144.91ms step:1162/3200 train_loss:3.3351 train_time:166941ms step_avg:144.91ms step:1163/3200 train_loss:3.6425 train_time:167086ms step_avg:144.91ms step:1164/3200 train_loss:3.5989 train_time:167230ms step_avg:144.91ms step:1165/3200 train_loss:3.4735 train_time:167373ms step_avg:144.91ms step:1166/3200 train_loss:3.4649 train_time:167515ms step_avg:144.91ms step:1167/3200 train_loss:3.5790 train_time:167658ms step_avg:144.91ms step:1168/3200 train_loss:3.5832 train_time:167801ms step_avg:144.91ms step:1169/3200 train_loss:3.9077 train_time:167944ms step_avg:144.90ms step:1170/3200 train_loss:3.5875 train_time:168089ms step_avg:144.90ms step:1171/3200 train_loss:3.5982 train_time:168232ms step_avg:144.90ms step:1172/3200 train_loss:3.5079 train_time:168375ms step_avg:144.90ms step:1173/3200 train_loss:3.5980 train_time:168517ms step_avg:144.90ms step:1174/3200 train_loss:3.7372 train_time:168662ms step_avg:144.90ms step:1175/3200 train_loss:3.5773 train_time:168806ms step_avg:144.90ms step:1176/3200 train_loss:3.5946 train_time:168949ms step_avg:144.90ms step:1177/3200 train_loss:3.6484 train_time:169095ms step_avg:144.90ms step:1178/3200 train_loss:3.6371 train_time:169237ms step_avg:144.89ms step:1179/3200 train_loss:3.6923 train_time:169380ms step_avg:144.89ms step:1180/3200 train_loss:3.5934 train_time:169523ms step_avg:144.89ms step:1181/3200 train_loss:3.5908 train_time:169665ms step_avg:144.89ms step:1182/3200 train_loss:3.5485 train_time:169808ms step_avg:144.89ms step:1183/3200 train_loss:3.6038 train_time:169952ms step_avg:144.89ms step:1184/3200 train_loss:3.5264 train_time:170095ms step_avg:144.89ms step:1185/3200 train_loss:3.7009 train_time:170238ms step_avg:144.88ms step:1186/3200 train_loss:3.7634 train_time:170381ms step_avg:144.88ms step:1187/3200 train_loss:3.5548 train_time:170525ms step_avg:144.88ms step:1188/3200 train_loss:3.6196 train_time:170669ms step_avg:144.88ms step:1189/3200 train_loss:3.6366 train_time:170812ms step_avg:144.88ms step:1190/3200 train_loss:3.4797 train_time:170955ms step_avg:144.88ms step:1191/3200 train_loss:3.6500 train_time:171097ms step_avg:144.88ms step:1192/3200 train_loss:3.7969 train_time:171242ms step_avg:144.87ms step:1193/3200 train_loss:3.5933 train_time:171383ms step_avg:144.87ms step:1194/3200 train_loss:3.4814 train_time:171527ms step_avg:144.87ms step:1195/3200 train_loss:3.7689 train_time:171672ms step_avg:144.87ms step:1196/3200 train_loss:3.5768 train_time:171813ms step_avg:144.87ms step:1197/3200 train_loss:3.5831 train_time:171956ms step_avg:144.87ms step:1198/3200 train_loss:3.4814 train_time:172099ms step_avg:144.86ms step:1199/3200 train_loss:3.4954 train_time:172242ms step_avg:144.86ms step:1200/3200 train_loss:3.5494 train_time:172386ms step_avg:144.86ms step:1201/3200 train_loss:3.6349 train_time:172529ms step_avg:144.86ms step:1202/3200 train_loss:3.7002 train_time:172673ms step_avg:144.86ms step:1203/3200 train_loss:3.7666 train_time:172814ms step_avg:144.86ms step:1204/3200 train_loss:3.6164 train_time:172958ms step_avg:144.86ms step:1205/3200 train_loss:3.5312 train_time:173100ms step_avg:144.85ms step:1206/3200 train_loss:3.6302 train_time:173243ms step_avg:144.85ms step:1207/3200 train_loss:3.6795 train_time:173387ms step_avg:144.85ms step:1208/3200 train_loss:3.7198 train_time:173530ms step_avg:144.85ms step:1209/3200 train_loss:3.5969 train_time:173674ms step_avg:144.85ms step:1210/3200 train_loss:3.4637 train_time:173815ms step_avg:144.85ms step:1211/3200 train_loss:3.5093 train_time:173959ms step_avg:144.84ms step:1212/3200 train_loss:3.6060 train_time:174101ms step_avg:144.84ms step:1213/3200 train_loss:3.6165 train_time:174244ms step_avg:144.84ms step:1214/3200 train_loss:3.6503 train_time:174389ms step_avg:144.84ms step:1215/3200 train_loss:3.5380 train_time:174532ms step_avg:144.84ms step:1216/3200 train_loss:3.6003 train_time:174675ms step_avg:144.84ms step:1217/3200 train_loss:3.5496 train_time:174817ms step_avg:144.84ms step:1218/3200 train_loss:3.5382 train_time:174960ms step_avg:144.83ms step:1219/3200 train_loss:3.6301 train_time:175103ms step_avg:144.83ms step:1220/3200 train_loss:3.4800 train_time:175246ms step_avg:144.83ms step:1221/3200 train_loss:3.7021 train_time:175393ms step_avg:144.83ms step:1222/3200 train_loss:3.7249 train_time:175535ms step_avg:144.83ms step:1223/3200 train_loss:3.6466 train_time:175678ms step_avg:144.83ms step:1224/3200 train_loss:3.5034 train_time:175821ms step_avg:144.83ms step:1225/3200 train_loss:3.4923 train_time:175964ms step_avg:144.83ms step:1226/3200 train_loss:3.5746 train_time:176106ms step_avg:144.82ms step:1227/3200 train_loss:3.5606 train_time:176250ms step_avg:144.82ms step:1228/3200 train_loss:3.4952 train_time:176394ms step_avg:144.82ms step:1229/3200 train_loss:3.6643 train_time:176537ms step_avg:144.82ms step:1230/3200 train_loss:3.5840 train_time:176680ms step_avg:144.82ms step:1231/3200 train_loss:3.6431 train_time:176823ms step_avg:144.82ms step:1232/3200 train_loss:3.8001 train_time:176966ms step_avg:144.82ms step:1233/3200 train_loss:3.7014 train_time:177110ms step_avg:144.82ms step:1234/3200 train_loss:3.6437 train_time:177253ms step_avg:144.81ms step:1235/3200 train_loss:3.7916 train_time:177396ms step_avg:144.81ms step:1236/3200 train_loss:3.5526 train_time:177538ms step_avg:144.81ms step:1237/3200 train_loss:3.5171 train_time:177681ms step_avg:144.81ms step:1238/3200 train_loss:3.4674 train_time:177825ms step_avg:144.81ms step:1239/3200 train_loss:3.5478 train_time:177969ms step_avg:144.81ms step:1240/3200 train_loss:3.5496 train_time:178111ms step_avg:144.81ms step:1241/3200 train_loss:3.5966 train_time:178254ms step_avg:144.80ms step:1242/3200 train_loss:3.6435 train_time:178397ms step_avg:144.80ms step:1243/3200 train_loss:3.5175 train_time:178540ms step_avg:144.80ms step:1244/3200 train_loss:3.6172 train_time:178683ms step_avg:144.80ms step:1245/3200 train_loss:3.6328 train_time:178827ms step_avg:144.80ms step:1246/3200 train_loss:3.6290 train_time:178973ms step_avg:144.80ms step:1247/3200 train_loss:3.4574 train_time:179116ms step_avg:144.80ms step:1248/3200 train_loss:3.6023 train_time:179258ms step_avg:144.80ms step:1249/3200 train_loss:3.6569 train_time:179401ms step_avg:144.80ms step:1250/3200 train_loss:3.6282 train_time:179545ms step_avg:144.79ms step:1250/3200 val_loss:3.5786 train_time:179602ms step_avg:144.84ms step:1251/3200 train_loss:3.5237 train_time:179697ms step_avg:144.80ms step:1252/3200 train_loss:3.7342 train_time:179846ms step_avg:144.80ms step:1253/3200 train_loss:3.6008 train_time:179989ms step_avg:144.80ms step:1254/3200 train_loss:3.5338 train_time:180131ms step_avg:144.80ms step:1255/3200 train_loss:3.6587 train_time:180273ms step_avg:144.80ms step:1256/3200 train_loss:3.7237 train_time:180414ms step_avg:144.79ms step:1257/3200 train_loss:3.5353 train_time:180556ms step_avg:144.79ms step:1258/3200 train_loss:3.5694 train_time:180702ms step_avg:144.79ms step:1259/3200 train_loss:3.5930 train_time:180847ms step_avg:144.79ms step:1260/3200 train_loss:3.5595 train_time:180991ms step_avg:144.79ms step:1261/3200 train_loss:3.4222 train_time:181133ms step_avg:144.79ms step:1262/3200 train_loss:3.5248 train_time:181275ms step_avg:144.79ms step:1263/3200 train_loss:3.5871 train_time:181417ms step_avg:144.79ms step:1264/3200 train_loss:3.4410 train_time:181559ms step_avg:144.78ms step:1265/3200 train_loss:3.6583 train_time:181703ms step_avg:144.78ms step:1266/3200 train_loss:3.6441 train_time:181850ms step_avg:144.78ms step:1267/3200 train_loss:3.6458 train_time:181993ms step_avg:144.78ms step:1268/3200 train_loss:3.5874 train_time:182137ms step_avg:144.78ms step:1269/3200 train_loss:3.6221 train_time:182279ms step_avg:144.78ms step:1270/3200 train_loss:3.4812 train_time:182420ms step_avg:144.78ms step:1271/3200 train_loss:3.3300 train_time:182563ms step_avg:144.78ms step:1272/3200 train_loss:3.6110 train_time:182705ms step_avg:144.77ms step:1273/3200 train_loss:3.5726 train_time:182850ms step_avg:144.77ms step:1274/3200 train_loss:3.6249 train_time:182993ms step_avg:144.77ms step:1275/3200 train_loss:3.5713 train_time:183137ms step_avg:144.77ms step:1276/3200 train_loss:3.6623 train_time:183279ms step_avg:144.77ms step:1277/3200 train_loss:3.6936 train_time:183421ms step_avg:144.77ms step:1278/3200 train_loss:3.6438 train_time:183564ms step_avg:144.77ms step:1279/3200 train_loss:3.6428 train_time:183707ms step_avg:144.76ms step:1280/3200 train_loss:3.4732 train_time:183852ms step_avg:144.77ms step:1281/3200 train_loss:3.5939 train_time:183995ms step_avg:144.76ms step:1282/3200 train_loss:3.6539 train_time:184139ms step_avg:144.76ms step:1283/3200 train_loss:3.6896 train_time:184281ms step_avg:144.76ms step:1284/3200 train_loss:3.5764 train_time:184423ms step_avg:144.76ms step:1285/3200 train_loss:3.6013 train_time:184566ms step_avg:144.76ms step:1286/3200 train_loss:3.5845 train_time:184709ms step_avg:144.76ms step:1287/3200 train_loss:3.5582 train_time:184852ms step_avg:144.75ms step:1288/3200 train_loss:3.6978 train_time:184995ms step_avg:144.75ms step:1289/3200 train_loss:3.5267 train_time:185139ms step_avg:144.75ms step:1290/3200 train_loss:3.6151 train_time:185284ms step_avg:144.75ms step:1291/3200 train_loss:3.6867 train_time:185427ms step_avg:144.75ms step:1292/3200 train_loss:3.6158 train_time:185570ms step_avg:144.75ms step:1293/3200 train_loss:3.7091 train_time:185712ms step_avg:144.75ms step:1294/3200 train_loss:3.7324 train_time:185854ms step_avg:144.75ms step:1295/3200 train_loss:3.6977 train_time:185998ms step_avg:144.75ms step:1296/3200 train_loss:3.5105 train_time:186142ms step_avg:144.74ms step:1297/3200 train_loss:3.5840 train_time:186285ms step_avg:144.74ms step:1298/3200 train_loss:3.4848 train_time:186430ms step_avg:144.74ms step:1299/3200 train_loss:3.5591 train_time:186574ms step_avg:144.74ms step:1300/3200 train_loss:3.6292 train_time:186716ms step_avg:144.74ms step:1301/3200 train_loss:3.6303 train_time:186859ms step_avg:144.74ms step:1302/3200 train_loss:3.6316 train_time:187002ms step_avg:144.74ms step:1303/3200 train_loss:3.7890 train_time:187147ms step_avg:144.74ms step:1304/3200 train_loss:3.5608 train_time:187291ms step_avg:144.74ms step:1305/3200 train_loss:3.7669 train_time:187434ms step_avg:144.74ms step:1306/3200 train_loss:3.4944 train_time:187576ms step_avg:144.73ms step:1307/3200 train_loss:3.6855 train_time:187719ms step_avg:144.73ms step:1308/3200 train_loss:3.6854 train_time:187862ms step_avg:144.73ms step:1309/3200 train_loss:3.5426 train_time:188006ms step_avg:144.73ms step:1310/3200 train_loss:3.5205 train_time:188150ms step_avg:144.73ms step:1311/3200 train_loss:3.5554 train_time:188293ms step_avg:144.73ms step:1312/3200 train_loss:3.5151 train_time:188435ms step_avg:144.73ms step:1313/3200 train_loss:3.6319 train_time:188578ms step_avg:144.73ms step:1314/3200 train_loss:3.5777 train_time:188723ms step_avg:144.73ms step:1315/3200 train_loss:3.2978 train_time:188866ms step_avg:144.72ms step:1316/3200 train_loss:3.5259 train_time:189010ms step_avg:144.72ms step:1317/3200 train_loss:3.6081 train_time:189152ms step_avg:144.72ms step:1318/3200 train_loss:3.6383 train_time:189296ms step_avg:144.72ms step:1319/3200 train_loss:3.5092 train_time:189440ms step_avg:144.72ms step:1320/3200 train_loss:3.6482 train_time:189585ms step_avg:144.72ms step:1321/3200 train_loss:3.7064 train_time:189730ms step_avg:144.72ms step:1322/3200 train_loss:3.5932 train_time:189874ms step_avg:144.72ms step:1323/3200 train_loss:3.5407 train_time:190177ms step_avg:144.84ms step:1324/3200 train_loss:3.5668 train_time:190321ms step_avg:144.84ms step:1325/3200 train_loss:3.6585 train_time:190461ms step_avg:144.84ms step:1326/3200 train_loss:3.7172 train_time:190602ms step_avg:144.83ms step:1327/3200 train_loss:3.4733 train_time:190744ms step_avg:144.83ms step:1328/3200 train_loss:3.3991 train_time:190887ms step_avg:144.83ms step:1329/3200 train_loss:3.7070 train_time:191029ms step_avg:144.83ms step:1330/3200 train_loss:3.5469 train_time:191335ms step_avg:144.95ms step:1331/3200 train_loss:3.6802 train_time:191473ms step_avg:144.95ms step:1332/3200 train_loss:3.5781 train_time:191615ms step_avg:144.94ms step:1333/3200 train_loss:3.9820 train_time:191756ms step_avg:144.94ms step:1334/3200 train_loss:3.6854 train_time:191898ms step_avg:144.94ms step:1335/3200 train_loss:3.5936 train_time:192039ms step_avg:144.94ms step:1336/3200 train_loss:3.5370 train_time:192181ms step_avg:144.93ms step:1337/3200 train_loss:3.5294 train_time:192330ms step_avg:144.94ms step:1338/3200 train_loss:3.7857 train_time:192474ms step_avg:144.94ms step:1339/3200 train_loss:3.7299 train_time:192617ms step_avg:144.93ms step:1340/3200 train_loss:3.5675 train_time:192759ms step_avg:144.93ms step:1341/3200 train_loss:3.5304 train_time:192902ms step_avg:144.93ms step:1342/3200 train_loss:3.8319 train_time:193045ms step_avg:144.93ms step:1343/3200 train_loss:3.6032 train_time:193188ms step_avg:144.93ms step:1344/3200 train_loss:3.6013 train_time:193334ms step_avg:144.93ms step:1345/3200 train_loss:3.6568 train_time:193477ms step_avg:144.93ms step:1346/3200 train_loss:3.6223 train_time:193621ms step_avg:144.93ms step:1347/3200 train_loss:3.5268 train_time:193765ms step_avg:144.93ms step:1348/3200 train_loss:3.4766 train_time:193909ms step_avg:144.92ms step:1349/3200 train_loss:3.5723 train_time:194052ms step_avg:144.92ms step:1350/3200 train_loss:3.4996 train_time:194194ms step_avg:144.92ms step:1351/3200 train_loss:3.6332 train_time:194337ms step_avg:144.92ms step:1352/3200 train_loss:3.4835 train_time:194481ms step_avg:144.92ms step:1353/3200 train_loss:3.5464 train_time:194625ms step_avg:144.92ms step:1354/3200 train_loss:3.6521 train_time:194770ms step_avg:144.92ms step:1355/3200 train_loss:3.4913 train_time:194912ms step_avg:144.92ms step:1356/3200 train_loss:3.4195 train_time:195055ms step_avg:144.91ms step:1357/3200 train_loss:3.7591 train_time:195198ms step_avg:144.91ms step:1358/3200 train_loss:3.6903 train_time:195341ms step_avg:144.91ms step:1359/3200 train_loss:3.4120 train_time:195485ms step_avg:144.91ms step:1360/3200 train_loss:3.6906 train_time:195629ms step_avg:144.91ms step:1361/3200 train_loss:3.5740 train_time:195772ms step_avg:144.91ms step:1362/3200 train_loss:3.4328 train_time:195915ms step_avg:144.91ms step:1363/3200 train_loss:3.6210 train_time:196058ms step_avg:144.91ms step:1364/3200 train_loss:3.5135 train_time:196200ms step_avg:144.90ms step:1365/3200 train_loss:3.5292 train_time:196344ms step_avg:144.90ms step:1366/3200 train_loss:3.5535 train_time:196488ms step_avg:144.90ms step:1367/3200 train_loss:3.6525 train_time:196631ms step_avg:144.90ms step:1368/3200 train_loss:3.6393 train_time:196775ms step_avg:144.90ms step:1369/3200 train_loss:3.5934 train_time:196918ms step_avg:144.90ms step:1370/3200 train_loss:3.5022 train_time:197060ms step_avg:144.90ms step:1371/3200 train_loss:3.8237 train_time:197203ms step_avg:144.90ms step:1372/3200 train_loss:3.5684 train_time:197348ms step_avg:144.90ms step:1373/3200 train_loss:3.6028 train_time:197493ms step_avg:144.90ms step:1374/3200 train_loss:3.5974 train_time:197635ms step_avg:144.89ms step:1375/3200 train_loss:3.3973 train_time:197779ms step_avg:144.89ms step:1375/3200 val_loss:3.5582 train_time:197835ms step_avg:144.93ms step:1376/3200 train_loss:3.7988 train_time:197932ms step_avg:144.90ms step:1377/3200 train_loss:3.5781 train_time:198080ms step_avg:144.90ms step:1378/3200 train_loss:3.7236 train_time:198222ms step_avg:144.90ms step:1379/3200 train_loss:3.7636 train_time:198364ms step_avg:144.90ms step:1380/3200 train_loss:3.4138 train_time:198507ms step_avg:144.90ms step:1381/3200 train_loss:3.5673 train_time:198648ms step_avg:144.89ms step:1382/3200 train_loss:4.0278 train_time:198790ms step_avg:144.89ms step:1383/3200 train_loss:3.4776 train_time:198937ms step_avg:144.89ms step:1384/3200 train_loss:3.6323 train_time:199085ms step_avg:144.89ms step:1385/3200 train_loss:3.7079 train_time:199229ms step_avg:144.89ms step:1386/3200 train_loss:3.6233 train_time:199371ms step_avg:144.89ms step:1387/3200 train_loss:3.6240 train_time:199512ms step_avg:144.89ms step:1388/3200 train_loss:3.4445 train_time:199655ms step_avg:144.89ms step:1389/3200 train_loss:3.5861 train_time:199797ms step_avg:144.89ms step:1390/3200 train_loss:3.5569 train_time:199941ms step_avg:144.88ms step:1391/3200 train_loss:3.8203 train_time:200086ms step_avg:144.88ms step:1392/3200 train_loss:3.5384 train_time:200230ms step_avg:144.88ms step:1393/3200 train_loss:3.5246 train_time:200372ms step_avg:144.88ms step:1394/3200 train_loss:3.4901 train_time:200515ms step_avg:144.88ms step:1395/3200 train_loss:3.7694 train_time:200657ms step_avg:144.88ms step:1396/3200 train_loss:3.6675 train_time:200800ms step_avg:144.88ms step:1397/3200 train_loss:3.6705 train_time:200946ms step_avg:144.88ms step:1398/3200 train_loss:3.5388 train_time:201090ms step_avg:144.88ms step:1399/3200 train_loss:3.5178 train_time:201235ms step_avg:144.88ms step:1400/3200 train_loss:3.5726 train_time:201379ms step_avg:144.88ms step:1401/3200 train_loss:3.5518 train_time:201521ms step_avg:144.87ms step:1402/3200 train_loss:3.5791 train_time:201664ms step_avg:144.87ms step:1403/3200 train_loss:3.5350 train_time:201807ms step_avg:144.87ms step:1404/3200 train_loss:3.7694 train_time:201950ms step_avg:144.87ms step:1405/3200 train_loss:3.5189 train_time:202093ms step_avg:144.87ms step:1406/3200 train_loss:3.5662 train_time:202236ms step_avg:144.87ms step:1407/3200 train_loss:3.5549 train_time:202380ms step_avg:144.87ms step:1408/3200 train_loss:3.4273 train_time:202523ms step_avg:144.87ms step:1409/3200 train_loss:3.5447 train_time:202667ms step_avg:144.87ms step:1410/3200 train_loss:3.5257 train_time:202809ms step_avg:144.86ms step:1411/3200 train_loss:3.5240 train_time:202951ms step_avg:144.86ms step:1412/3200 train_loss:3.6074 train_time:203095ms step_avg:144.86ms step:1413/3200 train_loss:3.5537 train_time:203238ms step_avg:144.86ms step:1414/3200 train_loss:3.6008 train_time:203381ms step_avg:144.86ms step:1415/3200 train_loss:3.5859 train_time:203524ms step_avg:144.86ms step:1416/3200 train_loss:3.6652 train_time:203668ms step_avg:144.86ms step:1417/3200 train_loss:3.4744 train_time:203810ms step_avg:144.85ms step:1418/3200 train_loss:3.5340 train_time:203953ms step_avg:144.85ms step:1419/3200 train_loss:3.6245 train_time:204096ms step_avg:144.85ms step:1420/3200 train_loss:3.6515 train_time:204241ms step_avg:144.85ms step:1421/3200 train_loss:3.6359 train_time:204385ms step_avg:144.85ms step:1422/3200 train_loss:3.6157 train_time:204531ms step_avg:144.85ms step:1423/3200 train_loss:3.5988 train_time:204671ms step_avg:144.85ms step:1424/3200 train_loss:3.5841 train_time:204814ms step_avg:144.85ms step:1425/3200 train_loss:3.5847 train_time:204958ms step_avg:144.85ms step:1426/3200 train_loss:3.4591 train_time:205101ms step_avg:144.85ms step:1427/3200 train_loss:3.5642 train_time:205246ms step_avg:144.85ms step:1428/3200 train_loss:3.5167 train_time:205389ms step_avg:144.84ms step:1429/3200 train_loss:3.6248 train_time:205531ms step_avg:144.84ms step:1430/3200 train_loss:3.5838 train_time:205674ms step_avg:144.84ms step:1431/3200 train_loss:3.5149 train_time:205817ms step_avg:144.84ms step:1432/3200 train_loss:3.5642 train_time:205960ms step_avg:144.84ms step:1433/3200 train_loss:3.5998 train_time:206108ms step_avg:144.84ms step:1434/3200 train_loss:3.4707 train_time:206249ms step_avg:144.84ms step:1435/3200 train_loss:3.5748 train_time:206391ms step_avg:144.84ms step:1436/3200 train_loss:3.3979 train_time:206535ms step_avg:144.84ms step:1437/3200 train_loss:3.4640 train_time:206680ms step_avg:144.84ms step:1438/3200 train_loss:3.6610 train_time:206824ms step_avg:144.83ms step:1439/3200 train_loss:3.6176 train_time:206967ms step_avg:144.83ms step:1440/3200 train_loss:3.5646 train_time:207110ms step_avg:144.83ms step:1441/3200 train_loss:3.4264 train_time:207252ms step_avg:144.83ms step:1442/3200 train_loss:3.5887 train_time:207395ms step_avg:144.83ms step:1443/3200 train_loss:3.6570 train_time:207540ms step_avg:144.83ms step:1444/3200 train_loss:3.7295 train_time:207684ms step_avg:144.83ms step:1445/3200 train_loss:3.6904 train_time:207828ms step_avg:144.83ms step:1446/3200 train_loss:3.5825 train_time:207970ms step_avg:144.83ms step:1447/3200 train_loss:3.4472 train_time:208114ms step_avg:144.83ms step:1448/3200 train_loss:3.5286 train_time:208256ms step_avg:144.82ms step:1449/3200 train_loss:3.5461 train_time:208401ms step_avg:144.82ms step:1450/3200 train_loss:3.6665 train_time:208545ms step_avg:144.82ms step:1451/3200 train_loss:3.6507 train_time:208689ms step_avg:144.82ms step:1452/3200 train_loss:3.4698 train_time:208832ms step_avg:144.82ms step:1453/3200 train_loss:3.5828 train_time:208975ms step_avg:144.82ms step:1454/3200 train_loss:3.4975 train_time:209119ms step_avg:144.82ms step:1455/3200 train_loss:3.5311 train_time:209262ms step_avg:144.82ms step:1456/3200 train_loss:3.5752 train_time:209406ms step_avg:144.82ms step:1457/3200 train_loss:3.5105 train_time:209549ms step_avg:144.82ms step:1458/3200 train_loss:3.4041 train_time:209692ms step_avg:144.82ms step:1459/3200 train_loss:3.6493 train_time:209836ms step_avg:144.81ms step:1460/3200 train_loss:3.5156 train_time:209981ms step_avg:144.81ms step:1461/3200 train_loss:3.5685 train_time:210125ms step_avg:144.81ms step:1462/3200 train_loss:3.6970 train_time:210269ms step_avg:144.81ms step:1463/3200 train_loss:3.5129 train_time:210411ms step_avg:144.81ms step:1464/3200 train_loss:3.7078 train_time:210554ms step_avg:144.81ms step:1465/3200 train_loss:3.6017 train_time:210697ms step_avg:144.81ms step:1466/3200 train_loss:3.6103 train_time:210841ms step_avg:144.81ms step:1467/3200 train_loss:3.5249 train_time:210985ms step_avg:144.81ms step:1468/3200 train_loss:3.6791 train_time:211129ms step_avg:144.81ms step:1469/3200 train_loss:3.5455 train_time:211272ms step_avg:144.81ms step:1470/3200 train_loss:3.5162 train_time:211414ms step_avg:144.80ms step:1471/3200 train_loss:3.5716 train_time:211557ms step_avg:144.80ms step:1472/3200 train_loss:3.4944 train_time:211701ms step_avg:144.80ms step:1473/3200 train_loss:3.5828 train_time:211845ms step_avg:144.80ms step:1474/3200 train_loss:3.6810 train_time:211988ms step_avg:144.80ms step:1475/3200 train_loss:3.5510 train_time:212131ms step_avg:144.80ms step:1476/3200 train_loss:3.3847 train_time:212274ms step_avg:144.80ms step:1477/3200 train_loss:3.5017 train_time:212417ms step_avg:144.80ms step:1478/3200 train_loss:3.4749 train_time:212560ms step_avg:144.80ms step:1479/3200 train_loss:3.5614 train_time:212704ms step_avg:144.80ms step:1480/3200 train_loss:3.6441 train_time:212848ms step_avg:144.79ms step:1481/3200 train_loss:3.5176 train_time:212991ms step_avg:144.79ms step:1482/3200 train_loss:3.6943 train_time:213133ms step_avg:144.79ms step:1483/3200 train_loss:3.6153 train_time:213276ms step_avg:144.79ms step:1484/3200 train_loss:3.5216 train_time:213418ms step_avg:144.79ms step:1485/3200 train_loss:3.5137 train_time:213561ms step_avg:144.79ms step:1486/3200 train_loss:3.5097 train_time:213705ms step_avg:144.79ms step:1487/3200 train_loss:3.4864 train_time:213848ms step_avg:144.79ms step:1488/3200 train_loss:3.5743 train_time:213991ms step_avg:144.78ms step:1489/3200 train_loss:3.4827 train_time:214134ms step_avg:144.78ms step:1490/3200 train_loss:3.5720 train_time:214278ms step_avg:144.78ms step:1491/3200 train_loss:3.5107 train_time:214420ms step_avg:144.78ms step:1492/3200 train_loss:3.4331 train_time:214564ms step_avg:144.78ms step:1493/3200 train_loss:3.5119 train_time:214708ms step_avg:144.78ms step:1494/3200 train_loss:3.6814 train_time:214851ms step_avg:144.78ms step:1495/3200 train_loss:3.5358 train_time:214995ms step_avg:144.78ms step:1496/3200 train_loss:3.2911 train_time:215137ms step_avg:144.78ms step:1497/3200 train_loss:3.5970 train_time:215280ms step_avg:144.77ms step:1498/3200 train_loss:3.5573 train_time:215423ms step_avg:144.77ms step:1499/3200 train_loss:3.6054 train_time:215567ms step_avg:144.77ms step:1500/3200 train_loss:3.5576 train_time:215710ms step_avg:144.77ms step:1500/3200 val_loss:3.5387 train_time:215766ms step_avg:144.81ms step:1501/3200 train_loss:3.5385 train_time:215864ms step_avg:144.78ms step:1502/3200 train_loss:3.3339 train_time:216011ms step_avg:144.78ms step:1503/3200 train_loss:3.6122 train_time:216153ms step_avg:144.78ms step:1504/3200 train_loss:3.4937 train_time:216295ms step_avg:144.78ms step:1505/3200 train_loss:3.4980 train_time:216437ms step_avg:144.77ms step:1506/3200 train_loss:3.4539 train_time:216578ms step_avg:144.77ms step:1507/3200 train_loss:3.5364 train_time:216720ms step_avg:144.77ms step:1508/3200 train_loss:3.4615 train_time:216864ms step_avg:144.77ms step:1509/3200 train_loss:3.7663 train_time:217010ms step_avg:144.77ms step:1510/3200 train_loss:3.5123 train_time:217153ms step_avg:144.77ms step:1511/3200 train_loss:3.5159 train_time:217295ms step_avg:144.77ms step:1512/3200 train_loss:3.6460 train_time:217604ms step_avg:144.88ms step:1513/3200 train_loss:3.6753 train_time:217738ms step_avg:144.87ms step:1514/3200 train_loss:3.5310 train_time:217881ms step_avg:144.87ms step:1515/3200 train_loss:3.3530 train_time:218023ms step_avg:144.87ms step:1516/3200 train_loss:3.4882 train_time:218165ms step_avg:144.86ms step:1517/3200 train_loss:3.4940 train_time:218307ms step_avg:144.86ms step:1518/3200 train_loss:3.5739 train_time:218449ms step_avg:144.86ms step:1519/3200 train_loss:3.4618 train_time:218595ms step_avg:144.86ms step:1520/3200 train_loss:3.7570 train_time:218896ms step_avg:144.96ms step:1521/3200 train_loss:3.4165 train_time:219035ms step_avg:144.96ms step:1522/3200 train_loss:3.4789 train_time:219177ms step_avg:144.96ms step:1523/3200 train_loss:3.6224 train_time:219318ms step_avg:144.96ms step:1524/3200 train_loss:3.4833 train_time:219460ms step_avg:144.95ms step:1525/3200 train_loss:3.5787 train_time:219602ms step_avg:144.95ms step:1526/3200 train_loss:3.5696 train_time:219747ms step_avg:144.95ms step:1527/3200 train_loss:3.5310 train_time:219893ms step_avg:144.95ms step:1528/3200 train_loss:3.5308 train_time:220038ms step_avg:144.95ms step:1529/3200 train_loss:3.6820 train_time:220180ms step_avg:144.95ms step:1530/3200 train_loss:3.6552 train_time:220322ms step_avg:144.95ms step:1531/3200 train_loss:3.4911 train_time:220464ms step_avg:144.95ms step:1532/3200 train_loss:3.4489 train_time:220607ms step_avg:144.95ms step:1533/3200 train_loss:3.6038 train_time:220750ms step_avg:144.94ms step:1534/3200 train_loss:3.5537 train_time:220894ms step_avg:144.94ms step:1535/3200 train_loss:3.5426 train_time:221037ms step_avg:144.94ms step:1536/3200 train_loss:3.5353 train_time:221180ms step_avg:144.94ms step:1537/3200 train_loss:3.4764 train_time:221323ms step_avg:144.94ms step:1538/3200 train_loss:3.5296 train_time:221466ms step_avg:144.94ms step:1539/3200 train_loss:3.7033 train_time:221607ms step_avg:144.94ms step:1540/3200 train_loss:3.6386 train_time:221753ms step_avg:144.94ms step:1541/3200 train_loss:3.5432 train_time:221895ms step_avg:144.93ms step:1542/3200 train_loss:3.5051 train_time:222040ms step_avg:144.93ms step:1543/3200 train_loss:3.5032 train_time:222183ms step_avg:144.93ms step:1544/3200 train_loss:3.4626 train_time:222327ms step_avg:144.93ms step:1545/3200 train_loss:3.5531 train_time:222472ms step_avg:144.93ms step:1546/3200 train_loss:3.5201 train_time:222614ms step_avg:144.93ms step:1547/3200 train_loss:3.4988 train_time:222756ms step_avg:144.93ms step:1548/3200 train_loss:3.4601 train_time:222899ms step_avg:144.93ms step:1549/3200 train_loss:3.4961 train_time:223042ms step_avg:144.93ms step:1550/3200 train_loss:3.6093 train_time:223187ms step_avg:144.93ms step:1551/3200 train_loss:3.5346 train_time:223331ms step_avg:144.93ms step:1552/3200 train_loss:3.4755 train_time:223473ms step_avg:144.92ms step:1553/3200 train_loss:3.4692 train_time:223615ms step_avg:144.92ms step:1554/3200 train_loss:3.4601 train_time:223758ms step_avg:144.92ms step:1555/3200 train_loss:3.5842 train_time:223902ms step_avg:144.92ms step:1556/3200 train_loss:3.5937 train_time:224046ms step_avg:144.92ms step:1557/3200 train_loss:3.5269 train_time:224189ms step_avg:144.92ms step:1558/3200 train_loss:3.5846 train_time:224332ms step_avg:144.92ms step:1559/3200 train_loss:3.5049 train_time:224475ms step_avg:144.92ms step:1560/3200 train_loss:3.4104 train_time:224618ms step_avg:144.91ms step:1561/3200 train_loss:3.6604 train_time:224762ms step_avg:144.91ms step:1562/3200 train_loss:3.4775 train_time:224905ms step_avg:144.91ms step:1563/3200 train_loss:3.4623 train_time:225049ms step_avg:144.91ms step:1564/3200 train_loss:3.5875 train_time:225193ms step_avg:144.91ms step:1565/3200 train_loss:3.4184 train_time:225336ms step_avg:144.91ms step:1566/3200 train_loss:3.4641 train_time:225479ms step_avg:144.91ms step:1567/3200 train_loss:3.6211 train_time:225622ms step_avg:144.91ms step:1568/3200 train_loss:3.4990 train_time:225765ms step_avg:144.91ms step:1569/3200 train_loss:3.4857 train_time:225910ms step_avg:144.91ms step:1570/3200 train_loss:3.5818 train_time:226053ms step_avg:144.91ms step:1571/3200 train_loss:3.5890 train_time:226198ms step_avg:144.91ms step:1572/3200 train_loss:3.4217 train_time:226341ms step_avg:144.90ms step:1573/3200 train_loss:3.4476 train_time:226484ms step_avg:144.90ms step:1574/3200 train_loss:3.5666 train_time:226627ms step_avg:144.90ms step:1575/3200 train_loss:3.4356 train_time:226770ms step_avg:144.90ms step:1576/3200 train_loss:3.5804 train_time:226914ms step_avg:144.90ms step:1577/3200 train_loss:3.4893 train_time:227057ms step_avg:144.90ms step:1578/3200 train_loss:3.5417 train_time:227201ms step_avg:144.90ms step:1579/3200 train_loss:3.5184 train_time:227344ms step_avg:144.90ms step:1580/3200 train_loss:3.4811 train_time:227488ms step_avg:144.90ms step:1581/3200 train_loss:3.4568 train_time:227631ms step_avg:144.90ms step:1582/3200 train_loss:3.6985 train_time:227774ms step_avg:144.89ms step:1583/3200 train_loss:3.4729 train_time:227916ms step_avg:144.89ms step:1584/3200 train_loss:3.6194 train_time:228060ms step_avg:144.89ms step:1585/3200 train_loss:3.4571 train_time:228202ms step_avg:144.89ms step:1586/3200 train_loss:3.6176 train_time:228347ms step_avg:144.89ms step:1587/3200 train_loss:3.4025 train_time:228490ms step_avg:144.89ms step:1588/3200 train_loss:3.5952 train_time:228633ms step_avg:144.89ms step:1589/3200 train_loss:3.5063 train_time:228776ms step_avg:144.89ms step:1590/3200 train_loss:3.6626 train_time:228919ms step_avg:144.89ms step:1591/3200 train_loss:3.4751 train_time:229063ms step_avg:144.88ms step:1592/3200 train_loss:3.4985 train_time:229206ms step_avg:144.88ms step:1593/3200 train_loss:3.5663 train_time:229351ms step_avg:144.88ms step:1594/3200 train_loss:3.5427 train_time:229493ms step_avg:144.88ms step:1595/3200 train_loss:3.5150 train_time:229636ms step_avg:144.88ms step:1596/3200 train_loss:3.6552 train_time:229780ms step_avg:144.88ms step:1597/3200 train_loss:3.3930 train_time:229923ms step_avg:144.88ms step:1598/3200 train_loss:3.5513 train_time:230067ms step_avg:144.88ms step:1599/3200 train_loss:3.5944 train_time:230211ms step_avg:144.88ms step:1600/3200 train_loss:3.6367 train_time:230353ms step_avg:144.88ms step:1601/3200 train_loss:3.4901 train_time:230496ms step_avg:144.88ms step:1602/3200 train_loss:3.7899 train_time:230640ms step_avg:144.87ms step:1603/3200 train_loss:3.6702 train_time:230783ms step_avg:144.87ms step:1604/3200 train_loss:3.4516 train_time:230927ms step_avg:144.87ms step:1605/3200 train_loss:3.4896 train_time:231071ms step_avg:144.87ms step:1606/3200 train_loss:3.3751 train_time:231214ms step_avg:144.87ms step:1607/3200 train_loss:3.6949 train_time:231356ms step_avg:144.87ms step:1608/3200 train_loss:3.5023 train_time:231500ms step_avg:144.87ms step:1609/3200 train_loss:3.5247 train_time:231645ms step_avg:144.87ms step:1610/3200 train_loss:3.4774 train_time:231789ms step_avg:144.87ms step:1611/3200 train_loss:4.0759 train_time:231931ms step_avg:144.87ms step:1612/3200 train_loss:3.7047 train_time:232074ms step_avg:144.87ms step:1613/3200 train_loss:3.6179 train_time:232216ms step_avg:144.86ms step:1614/3200 train_loss:3.4859 train_time:232358ms step_avg:144.86ms step:1615/3200 train_loss:3.5317 train_time:232502ms step_avg:144.86ms step:1616/3200 train_loss:3.5251 train_time:232645ms step_avg:144.86ms step:1617/3200 train_loss:3.4895 train_time:232789ms step_avg:144.86ms step:1618/3200 train_loss:3.5614 train_time:232931ms step_avg:144.86ms step:1619/3200 train_loss:3.5158 train_time:233075ms step_avg:144.86ms step:1620/3200 train_loss:3.4077 train_time:233218ms step_avg:144.86ms step:1621/3200 train_loss:3.6795 train_time:233361ms step_avg:144.85ms step:1622/3200 train_loss:3.5858 train_time:233504ms step_avg:144.85ms step:1623/3200 train_loss:3.3754 train_time:233649ms step_avg:144.85ms step:1624/3200 train_loss:3.4944 train_time:233793ms step_avg:144.85ms step:1625/3200 train_loss:3.4423 train_time:233936ms step_avg:144.85ms step:1625/3200 val_loss:3.5230 train_time:233992ms step_avg:144.89ms step:1626/3200 train_loss:3.5291 train_time:234091ms step_avg:144.86ms step:1627/3200 train_loss:3.4963 train_time:234237ms step_avg:144.86ms step:1628/3200 train_loss:3.4593 train_time:234380ms step_avg:144.86ms step:1629/3200 train_loss:3.5601 train_time:234522ms step_avg:144.86ms step:1630/3200 train_loss:3.4654 train_time:234663ms step_avg:144.85ms step:1631/3200 train_loss:3.5208 train_time:234807ms step_avg:144.85ms step:1632/3200 train_loss:3.3988 train_time:234950ms step_avg:144.85ms step:1633/3200 train_loss:3.3630 train_time:235094ms step_avg:144.85ms step:1634/3200 train_loss:3.5301 train_time:235240ms step_avg:144.85ms step:1635/3200 train_loss:3.5204 train_time:235383ms step_avg:144.85ms step:1636/3200 train_loss:3.4595 train_time:235526ms step_avg:144.85ms step:1637/3200 train_loss:3.5426 train_time:235669ms step_avg:144.85ms step:1638/3200 train_loss:3.5948 train_time:235811ms step_avg:144.85ms step:1639/3200 train_loss:3.6271 train_time:235953ms step_avg:144.85ms step:1640/3200 train_loss:3.7933 train_time:236097ms step_avg:144.85ms step:1641/3200 train_loss:3.6078 train_time:236242ms step_avg:144.84ms step:1642/3200 train_loss:3.5184 train_time:236387ms step_avg:144.84ms step:1643/3200 train_loss:3.6065 train_time:236531ms step_avg:144.84ms step:1644/3200 train_loss:3.5058 train_time:236673ms step_avg:144.84ms step:1645/3200 train_loss:3.5254 train_time:236816ms step_avg:144.84ms step:1646/3200 train_loss:3.5218 train_time:236959ms step_avg:144.84ms step:1647/3200 train_loss:3.3014 train_time:237104ms step_avg:144.84ms step:1648/3200 train_loss:3.5554 train_time:237248ms step_avg:144.84ms step:1649/3200 train_loss:3.4308 train_time:237392ms step_avg:144.84ms step:1650/3200 train_loss:3.5064 train_time:237535ms step_avg:144.84ms step:1651/3200 train_loss:3.4755 train_time:237678ms step_avg:144.84ms step:1652/3200 train_loss:3.5502 train_time:237821ms step_avg:144.84ms step:1653/3200 train_loss:3.4804 train_time:237963ms step_avg:144.83ms step:1654/3200 train_loss:3.6073 train_time:238106ms step_avg:144.83ms step:1655/3200 train_loss:3.5928 train_time:238251ms step_avg:144.83ms step:1656/3200 train_loss:3.4146 train_time:238393ms step_avg:144.83ms step:1657/3200 train_loss:3.5765 train_time:238535ms step_avg:144.83ms step:1658/3200 train_loss:3.4643 train_time:238679ms step_avg:144.83ms step:1659/3200 train_loss:3.4461 train_time:238821ms step_avg:144.83ms step:1660/3200 train_loss:3.5329 train_time:238966ms step_avg:144.83ms step:1661/3200 train_loss:3.5552 train_time:239107ms step_avg:144.83ms step:1662/3200 train_loss:3.4769 train_time:239252ms step_avg:144.83ms step:1663/3200 train_loss:3.5681 train_time:239396ms step_avg:144.83ms step:1664/3200 train_loss:3.5732 train_time:239539ms step_avg:144.82ms step:1665/3200 train_loss:3.6051 train_time:239683ms step_avg:144.82ms step:1666/3200 train_loss:3.5767 train_time:239827ms step_avg:144.82ms step:1667/3200 train_loss:3.7195 train_time:239971ms step_avg:144.82ms step:1668/3200 train_loss:3.4289 train_time:240114ms step_avg:144.82ms step:1669/3200 train_loss:3.5065 train_time:240257ms step_avg:144.82ms step:1670/3200 train_loss:3.4323 train_time:240400ms step_avg:144.82ms step:1671/3200 train_loss:3.4389 train_time:240543ms step_avg:144.82ms step:1672/3200 train_loss:3.5948 train_time:240687ms step_avg:144.82ms step:1673/3200 train_loss:3.7765 train_time:240832ms step_avg:144.82ms step:1674/3200 train_loss:3.4956 train_time:240975ms step_avg:144.82ms step:1675/3200 train_loss:3.4802 train_time:241117ms step_avg:144.81ms step:1676/3200 train_loss:3.3666 train_time:241260ms step_avg:144.81ms step:1677/3200 train_loss:3.5797 train_time:241404ms step_avg:144.81ms step:1678/3200 train_loss:3.4833 train_time:241547ms step_avg:144.81ms step:1679/3200 train_loss:3.5137 train_time:241691ms step_avg:144.81ms step:1680/3200 train_loss:3.5027 train_time:241835ms step_avg:144.81ms step:1681/3200 train_loss:3.3147 train_time:241978ms step_avg:144.81ms step:1682/3200 train_loss:3.5035 train_time:242121ms step_avg:144.81ms step:1683/3200 train_loss:3.5212 train_time:242266ms step_avg:144.81ms step:1684/3200 train_loss:3.5609 train_time:242410ms step_avg:144.81ms step:1685/3200 train_loss:3.5584 train_time:242553ms step_avg:144.81ms step:1686/3200 train_loss:3.4662 train_time:242696ms step_avg:144.81ms step:1687/3200 train_loss:3.5750 train_time:242840ms step_avg:144.81ms step:1688/3200 train_loss:3.4564 train_time:242983ms step_avg:144.81ms step:1689/3200 train_loss:3.5419 train_time:243127ms step_avg:144.80ms step:1690/3200 train_loss:3.4604 train_time:243271ms step_avg:144.80ms step:1691/3200 train_loss:3.3579 train_time:243413ms step_avg:144.80ms step:1692/3200 train_loss:3.5095 train_time:243556ms step_avg:144.80ms step:1693/3200 train_loss:3.5033 train_time:243699ms step_avg:144.80ms step:1694/3200 train_loss:3.4196 train_time:243840ms step_avg:144.80ms step:1695/3200 train_loss:3.8648 train_time:243984ms step_avg:144.80ms step:1696/3200 train_loss:3.5752 train_time:244128ms step_avg:144.80ms step:1697/3200 train_loss:3.5580 train_time:244272ms step_avg:144.80ms step:1698/3200 train_loss:3.4641 train_time:244415ms step_avg:144.80ms step:1699/3200 train_loss:3.3757 train_time:244558ms step_avg:144.79ms step:1700/3200 train_loss:3.4629 train_time:244700ms step_avg:144.79ms step:1701/3200 train_loss:3.4698 train_time:245009ms step_avg:144.89ms step:1702/3200 train_loss:3.5377 train_time:245146ms step_avg:144.89ms step:1703/3200 train_loss:3.4590 train_time:245288ms step_avg:144.88ms step:1704/3200 train_loss:3.6617 train_time:245430ms step_avg:144.88ms step:1705/3200 train_loss:3.4275 train_time:245572ms step_avg:144.88ms step:1706/3200 train_loss:3.6464 train_time:245715ms step_avg:144.88ms step:1707/3200 train_loss:3.4945 train_time:245857ms step_avg:144.88ms step:1708/3200 train_loss:3.2788 train_time:246005ms step_avg:144.88ms step:1709/3200 train_loss:3.6082 train_time:246152ms step_avg:144.88ms step:1710/3200 train_loss:3.5196 train_time:246454ms step_avg:144.97ms step:1711/3200 train_loss:3.5141 train_time:246596ms step_avg:144.97ms step:1712/3200 train_loss:3.5009 train_time:246738ms step_avg:144.97ms step:1713/3200 train_loss:3.5400 train_time:246879ms step_avg:144.97ms step:1714/3200 train_loss:3.5605 train_time:247025ms step_avg:144.97ms step:1715/3200 train_loss:3.4855 train_time:247164ms step_avg:144.96ms step:1716/3200 train_loss:3.4898 train_time:247306ms step_avg:144.96ms step:1717/3200 train_loss:3.3301 train_time:247452ms step_avg:144.96ms step:1718/3200 train_loss:3.4679 train_time:247597ms step_avg:144.96ms step:1719/3200 train_loss:3.4916 train_time:247740ms step_avg:144.96ms step:1720/3200 train_loss:3.4353 train_time:247882ms step_avg:144.96ms step:1721/3200 train_loss:3.5927 train_time:248026ms step_avg:144.96ms step:1722/3200 train_loss:3.3979 train_time:248169ms step_avg:144.96ms step:1723/3200 train_loss:3.5400 train_time:248312ms step_avg:144.96ms step:1724/3200 train_loss:3.6211 train_time:248455ms step_avg:144.96ms step:1725/3200 train_loss:3.4741 train_time:248600ms step_avg:144.96ms step:1726/3200 train_loss:3.6987 train_time:248743ms step_avg:144.96ms step:1727/3200 train_loss:3.4934 train_time:248886ms step_avg:144.95ms step:1728/3200 train_loss:3.5460 train_time:249031ms step_avg:144.95ms step:1729/3200 train_loss:3.5182 train_time:249173ms step_avg:144.95ms step:1730/3200 train_loss:3.5336 train_time:249317ms step_avg:144.95ms step:1731/3200 train_loss:3.8944 train_time:249460ms step_avg:144.95ms step:1732/3200 train_loss:3.5123 train_time:249604ms step_avg:144.95ms step:1733/3200 train_loss:3.6367 train_time:249748ms step_avg:144.95ms step:1734/3200 train_loss:3.4200 train_time:249893ms step_avg:144.95ms step:1735/3200 train_loss:3.4667 train_time:250036ms step_avg:144.95ms step:1736/3200 train_loss:3.4853 train_time:250179ms step_avg:144.95ms step:1737/3200 train_loss:3.4701 train_time:250321ms step_avg:144.95ms step:1738/3200 train_loss:3.6124 train_time:250465ms step_avg:144.94ms step:1739/3200 train_loss:3.4703 train_time:250610ms step_avg:144.94ms step:1740/3200 train_loss:3.5298 train_time:250752ms step_avg:144.94ms step:1741/3200 train_loss:3.5852 train_time:250896ms step_avg:144.94ms step:1742/3200 train_loss:3.3905 train_time:251037ms step_avg:144.94ms step:1743/3200 train_loss:3.2830 train_time:251181ms step_avg:144.94ms step:1744/3200 train_loss:3.2050 train_time:251324ms step_avg:144.94ms step:1745/3200 train_loss:3.5075 train_time:251467ms step_avg:144.94ms step:1746/3200 train_loss:3.5241 train_time:251611ms step_avg:144.94ms step:1747/3200 train_loss:3.4890 train_time:251754ms step_avg:144.94ms step:1748/3200 train_loss:3.5032 train_time:251897ms step_avg:144.94ms step:1749/3200 train_loss:3.7424 train_time:252041ms step_avg:144.93ms step:1750/3200 train_loss:3.4475 train_time:252183ms step_avg:144.93ms step:1750/3200 val_loss:3.5049 train_time:252238ms step_avg:144.96ms step:1751/3200 train_loss:3.5204 train_time:252337ms step_avg:144.94ms step:1752/3200 train_loss:3.5104 train_time:252487ms step_avg:144.94ms step:1753/3200 train_loss:3.1555 train_time:252630ms step_avg:144.94ms step:1754/3200 train_loss:3.2726 train_time:252772ms step_avg:144.94ms step:1755/3200 train_loss:3.3613 train_time:252913ms step_avg:144.94ms step:1756/3200 train_loss:3.3211 train_time:253056ms step_avg:144.93ms step:1757/3200 train_loss:3.4796 train_time:253197ms step_avg:144.93ms step:1758/3200 train_loss:3.3602 train_time:253343ms step_avg:144.93ms step:1759/3200 train_loss:3.3543 train_time:253492ms step_avg:144.94ms step:1760/3200 train_loss:4.4146 train_time:253637ms step_avg:144.94ms step:1761/3200 train_loss:3.4912 train_time:253780ms step_avg:144.93ms step:1762/3200 train_loss:3.5270 train_time:253922ms step_avg:144.93ms step:1763/3200 train_loss:3.5231 train_time:254064ms step_avg:144.93ms step:1764/3200 train_loss:3.5412 train_time:254207ms step_avg:144.93ms step:1765/3200 train_loss:3.4570 train_time:254352ms step_avg:144.93ms step:1766/3200 train_loss:3.4995 train_time:254496ms step_avg:144.93ms step:1767/3200 train_loss:3.5126 train_time:254640ms step_avg:144.93ms step:1768/3200 train_loss:3.7638 train_time:254785ms step_avg:144.93ms step:1769/3200 train_loss:3.4873 train_time:254928ms step_avg:144.93ms step:1770/3200 train_loss:3.5556 train_time:255071ms step_avg:144.93ms step:1771/3200 train_loss:3.8495 train_time:255213ms step_avg:144.93ms step:1772/3200 train_loss:3.4886 train_time:255356ms step_avg:144.92ms step:1773/3200 train_loss:3.3930 train_time:255498ms step_avg:144.92ms step:1774/3200 train_loss:3.6493 train_time:255643ms step_avg:144.92ms step:1775/3200 train_loss:3.3962 train_time:255787ms step_avg:144.92ms step:1776/3200 train_loss:3.5537 train_time:255931ms step_avg:144.92ms step:1777/3200 train_loss:3.6019 train_time:256073ms step_avg:144.92ms step:1778/3200 train_loss:3.6931 train_time:256216ms step_avg:144.92ms step:1779/3200 train_loss:3.4972 train_time:256359ms step_avg:144.92ms step:1780/3200 train_loss:3.7988 train_time:256504ms step_avg:144.92ms step:1781/3200 train_loss:3.5696 train_time:256649ms step_avg:144.92ms step:1782/3200 train_loss:3.5854 train_time:256793ms step_avg:144.92ms step:1783/3200 train_loss:3.3726 train_time:256936ms step_avg:144.92ms step:1784/3200 train_loss:3.4560 train_time:257079ms step_avg:144.91ms step:1785/3200 train_loss:3.5992 train_time:257221ms step_avg:144.91ms step:1786/3200 train_loss:3.4885 train_time:257364ms step_avg:144.91ms step:1787/3200 train_loss:3.6608 train_time:257508ms step_avg:144.91ms step:1788/3200 train_loss:3.4732 train_time:257653ms step_avg:144.91ms step:1789/3200 train_loss:3.4437 train_time:257796ms step_avg:144.91ms step:1790/3200 train_loss:3.5936 train_time:257939ms step_avg:144.91ms step:1791/3200 train_loss:3.4908 train_time:258082ms step_avg:144.91ms step:1792/3200 train_loss:3.4414 train_time:258225ms step_avg:144.91ms step:1793/3200 train_loss:3.5714 train_time:258370ms step_avg:144.91ms step:1794/3200 train_loss:3.4454 train_time:258514ms step_avg:144.91ms step:1795/3200 train_loss:3.4386 train_time:258657ms step_avg:144.91ms step:1796/3200 train_loss:3.4960 train_time:258799ms step_avg:144.90ms step:1797/3200 train_loss:3.4572 train_time:258944ms step_avg:144.90ms step:1798/3200 train_loss:3.5968 train_time:259087ms step_avg:144.90ms step:1799/3200 train_loss:3.4787 train_time:259230ms step_avg:144.90ms step:1800/3200 train_loss:3.5619 train_time:259374ms step_avg:144.90ms step:1801/3200 train_loss:3.4861 train_time:259517ms step_avg:144.90ms step:1802/3200 train_loss:3.5239 train_time:259660ms step_avg:144.90ms step:1803/3200 train_loss:3.4267 train_time:259804ms step_avg:144.90ms step:1804/3200 train_loss:3.3664 train_time:259947ms step_avg:144.90ms step:1805/3200 train_loss:3.6108 train_time:260092ms step_avg:144.90ms step:1806/3200 train_loss:3.5337 train_time:260234ms step_avg:144.90ms step:1807/3200 train_loss:3.5544 train_time:260377ms step_avg:144.90ms step:1808/3200 train_loss:3.6532 train_time:260520ms step_avg:144.89ms step:1809/3200 train_loss:3.4477 train_time:260665ms step_avg:144.89ms step:1810/3200 train_loss:3.5559 train_time:260808ms step_avg:144.89ms step:1811/3200 train_loss:3.6923 train_time:260952ms step_avg:144.89ms step:1812/3200 train_loss:3.5430 train_time:261095ms step_avg:144.89ms step:1813/3200 train_loss:3.5867 train_time:261238ms step_avg:144.89ms step:1814/3200 train_loss:3.6063 train_time:261381ms step_avg:144.89ms step:1815/3200 train_loss:3.5500 train_time:261525ms step_avg:144.89ms step:1816/3200 train_loss:3.5855 train_time:261668ms step_avg:144.89ms step:1817/3200 train_loss:3.5437 train_time:261811ms step_avg:144.89ms step:1818/3200 train_loss:3.5899 train_time:261954ms step_avg:144.89ms step:1819/3200 train_loss:3.5218 train_time:262096ms step_avg:144.88ms step:1820/3200 train_loss:3.5057 train_time:262239ms step_avg:144.88ms step:1821/3200 train_loss:3.4684 train_time:262383ms step_avg:144.88ms step:1822/3200 train_loss:3.4434 train_time:262526ms step_avg:144.88ms step:1823/3200 train_loss:3.3661 train_time:262670ms step_avg:144.88ms step:1824/3200 train_loss:3.5207 train_time:262814ms step_avg:144.88ms step:1825/3200 train_loss:3.6441 train_time:262956ms step_avg:144.88ms step:1826/3200 train_loss:3.5944 train_time:263100ms step_avg:144.88ms step:1827/3200 train_loss:3.5757 train_time:263241ms step_avg:144.88ms step:1828/3200 train_loss:3.4512 train_time:263385ms step_avg:144.88ms step:1829/3200 train_loss:3.4672 train_time:263528ms step_avg:144.88ms step:1830/3200 train_loss:3.6106 train_time:263673ms step_avg:144.88ms step:1831/3200 train_loss:3.3833 train_time:263815ms step_avg:144.87ms step:1832/3200 train_loss:3.5342 train_time:263958ms step_avg:144.87ms step:1833/3200 train_loss:3.4132 train_time:264101ms step_avg:144.87ms step:1834/3200 train_loss:3.7279 train_time:264245ms step_avg:144.87ms step:1835/3200 train_loss:3.5652 train_time:264389ms step_avg:144.87ms step:1836/3200 train_loss:3.5465 train_time:264534ms step_avg:144.87ms step:1837/3200 train_loss:3.6797 train_time:264677ms step_avg:144.87ms step:1838/3200 train_loss:3.5323 train_time:264820ms step_avg:144.87ms step:1839/3200 train_loss:3.4120 train_time:264964ms step_avg:144.87ms step:1840/3200 train_loss:3.5316 train_time:265107ms step_avg:144.87ms step:1841/3200 train_loss:3.4245 train_time:265250ms step_avg:144.87ms step:1842/3200 train_loss:3.5247 train_time:265394ms step_avg:144.87ms step:1843/3200 train_loss:3.5824 train_time:265537ms step_avg:144.86ms step:1844/3200 train_loss:3.3322 train_time:265680ms step_avg:144.86ms step:1845/3200 train_loss:3.4613 train_time:265823ms step_avg:144.86ms step:1846/3200 train_loss:3.5160 train_time:265965ms step_avg:144.86ms step:1847/3200 train_loss:3.4552 train_time:266108ms step_avg:144.86ms step:1848/3200 train_loss:3.3587 train_time:266252ms step_avg:144.86ms step:1849/3200 train_loss:3.6188 train_time:266396ms step_avg:144.86ms step:1850/3200 train_loss:3.3912 train_time:266538ms step_avg:144.86ms step:1851/3200 train_loss:3.4723 train_time:266682ms step_avg:144.86ms step:1852/3200 train_loss:3.4293 train_time:266826ms step_avg:144.86ms step:1853/3200 train_loss:3.6333 train_time:266969ms step_avg:144.86ms step:1854/3200 train_loss:3.6097 train_time:267114ms step_avg:144.86ms step:1855/3200 train_loss:3.4868 train_time:267256ms step_avg:144.85ms step:1856/3200 train_loss:3.4392 train_time:267399ms step_avg:144.85ms step:1857/3200 train_loss:3.4626 train_time:267542ms step_avg:144.85ms step:1858/3200 train_loss:3.7108 train_time:267686ms step_avg:144.85ms step:1859/3200 train_loss:3.5511 train_time:267828ms step_avg:144.85ms step:1860/3200 train_loss:3.4998 train_time:267972ms step_avg:144.85ms step:1861/3200 train_loss:3.5329 train_time:268117ms step_avg:144.85ms step:1862/3200 train_loss:3.4231 train_time:268260ms step_avg:144.85ms step:1863/3200 train_loss:3.4237 train_time:268403ms step_avg:144.85ms step:1864/3200 train_loss:3.4898 train_time:268546ms step_avg:144.85ms step:1865/3200 train_loss:3.5332 train_time:268691ms step_avg:144.85ms step:1866/3200 train_loss:3.2919 train_time:268834ms step_avg:144.85ms step:1867/3200 train_loss:3.4273 train_time:268977ms step_avg:144.84ms step:1868/3200 train_loss:3.3850 train_time:269121ms step_avg:144.84ms step:1869/3200 train_loss:3.3855 train_time:269264ms step_avg:144.84ms step:1870/3200 train_loss:3.5412 train_time:269406ms step_avg:144.84ms step:1871/3200 train_loss:3.5229 train_time:269549ms step_avg:144.84ms step:1872/3200 train_loss:3.4776 train_time:269695ms step_avg:144.84ms step:1873/3200 train_loss:3.4894 train_time:269838ms step_avg:144.84ms step:1874/3200 train_loss:3.4166 train_time:269981ms step_avg:144.84ms step:1875/3200 train_loss:3.5124 train_time:270126ms step_avg:144.84ms step:1875/3200 val_loss:3.4917 train_time:270181ms step_avg:144.87ms step:1876/3200 train_loss:3.5140 train_time:270279ms step_avg:144.84ms step:1877/3200 train_loss:3.4492 train_time:270428ms step_avg:144.85ms step:1878/3200 train_loss:3.4873 train_time:270572ms step_avg:144.85ms step:1879/3200 train_loss:3.6033 train_time:270714ms step_avg:144.84ms step:1880/3200 train_loss:3.4799 train_time:270856ms step_avg:144.84ms step:1881/3200 train_loss:3.5302 train_time:270997ms step_avg:144.84ms step:1882/3200 train_loss:3.4451 train_time:271139ms step_avg:144.84ms step:1883/3200 train_loss:3.5141 train_time:271287ms step_avg:144.84ms step:1884/3200 train_loss:3.5203 train_time:271434ms step_avg:144.84ms step:1885/3200 train_loss:3.2690 train_time:271577ms step_avg:144.84ms step:1886/3200 train_loss:3.6655 train_time:271720ms step_avg:144.84ms step:1887/3200 train_loss:3.3994 train_time:271862ms step_avg:144.84ms step:1888/3200 train_loss:3.4175 train_time:272004ms step_avg:144.84ms step:1889/3200 train_loss:3.4945 train_time:272146ms step_avg:144.84ms step:1890/3200 train_loss:3.5388 train_time:272451ms step_avg:144.92ms step:1891/3200 train_loss:3.3608 train_time:272593ms step_avg:144.92ms step:1892/3200 train_loss:3.6292 train_time:272736ms step_avg:144.92ms step:1893/3200 train_loss:3.3916 train_time:272877ms step_avg:144.92ms step:1894/3200 train_loss:3.5265 train_time:273019ms step_avg:144.91ms step:1895/3200 train_loss:3.5537 train_time:273161ms step_avg:144.91ms step:1896/3200 train_loss:3.3595 train_time:273303ms step_avg:144.91ms step:1897/3200 train_loss:3.5200 train_time:273451ms step_avg:144.91ms step:1898/3200 train_loss:3.4823 train_time:273597ms step_avg:144.91ms step:1899/3200 train_loss:3.5653 train_time:273740ms step_avg:144.91ms step:1900/3200 train_loss:3.3402 train_time:274049ms step_avg:145.00ms step:1901/3200 train_loss:3.5842 train_time:274190ms step_avg:145.00ms step:1902/3200 train_loss:3.4654 train_time:274333ms step_avg:145.00ms step:1903/3200 train_loss:3.6319 train_time:274475ms step_avg:144.99ms step:1904/3200 train_loss:3.4308 train_time:274617ms step_avg:144.99ms step:1905/3200 train_loss:3.7049 train_time:274758ms step_avg:144.99ms step:1906/3200 train_loss:3.4445 train_time:274902ms step_avg:144.99ms step:1907/3200 train_loss:3.4377 train_time:275052ms step_avg:144.99ms step:1908/3200 train_loss:3.5153 train_time:275195ms step_avg:144.99ms step:1909/3200 train_loss:3.3970 train_time:275338ms step_avg:144.99ms step:1910/3200 train_loss:3.4684 train_time:275480ms step_avg:144.99ms step:1911/3200 train_loss:3.5596 train_time:275622ms step_avg:144.99ms step:1912/3200 train_loss:3.4830 train_time:275764ms step_avg:144.99ms step:1913/3200 train_loss:3.3605 train_time:275908ms step_avg:144.99ms step:1914/3200 train_loss:3.2399 train_time:276055ms step_avg:144.99ms step:1915/3200 train_loss:3.4309 train_time:276198ms step_avg:144.99ms step:1916/3200 train_loss:3.6521 train_time:276341ms step_avg:144.98ms step:1917/3200 train_loss:3.6518 train_time:276483ms step_avg:144.98ms step:1918/3200 train_loss:3.5966 train_time:276626ms step_avg:144.98ms step:1919/3200 train_loss:3.4243 train_time:276769ms step_avg:144.98ms step:1920/3200 train_loss:3.6724 train_time:276914ms step_avg:144.98ms step:1921/3200 train_loss:3.4948 train_time:277057ms step_avg:144.98ms step:1922/3200 train_loss:3.4289 train_time:277201ms step_avg:144.98ms step:1923/3200 train_loss:3.6039 train_time:277346ms step_avg:144.98ms step:1924/3200 train_loss:3.5684 train_time:277490ms step_avg:144.98ms step:1925/3200 train_loss:3.4083 train_time:277633ms step_avg:144.98ms step:1926/3200 train_loss:3.4329 train_time:277776ms step_avg:144.98ms step:1927/3200 train_loss:3.3451 train_time:277918ms step_avg:144.98ms step:1928/3200 train_loss:3.4595 train_time:278061ms step_avg:144.97ms step:1929/3200 train_loss:3.3152 train_time:278205ms step_avg:144.97ms step:1930/3200 train_loss:3.4347 train_time:278349ms step_avg:144.97ms step:1931/3200 train_loss:3.5656 train_time:278494ms step_avg:144.97ms step:1932/3200 train_loss:3.4337 train_time:278636ms step_avg:144.97ms step:1933/3200 train_loss:3.5754 train_time:278779ms step_avg:144.97ms step:1934/3200 train_loss:3.4456 train_time:278922ms step_avg:144.97ms step:1935/3200 train_loss:3.4955 train_time:279065ms step_avg:144.97ms step:1936/3200 train_loss:3.5296 train_time:279208ms step_avg:144.97ms step:1937/3200 train_loss:3.4872 train_time:279353ms step_avg:144.97ms step:1938/3200 train_loss:3.5154 train_time:279496ms step_avg:144.97ms step:1939/3200 train_loss:3.4370 train_time:279639ms step_avg:144.97ms step:1940/3200 train_loss:3.5312 train_time:279782ms step_avg:144.96ms step:1941/3200 train_loss:3.5692 train_time:279924ms step_avg:144.96ms step:1942/3200 train_loss:3.4115 train_time:280067ms step_avg:144.96ms step:1943/3200 train_loss:3.4469 train_time:280211ms step_avg:144.96ms step:1944/3200 train_loss:3.5110 train_time:280356ms step_avg:144.96ms step:1945/3200 train_loss:3.3554 train_time:280498ms step_avg:144.96ms step:1946/3200 train_loss:3.6251 train_time:280642ms step_avg:144.96ms step:1947/3200 train_loss:3.5010 train_time:280785ms step_avg:144.96ms step:1948/3200 train_loss:3.4735 train_time:280929ms step_avg:144.96ms step:1949/3200 train_loss:3.4727 train_time:281074ms step_avg:144.96ms step:1950/3200 train_loss:3.3603 train_time:281217ms step_avg:144.96ms step:1951/3200 train_loss:3.4739 train_time:281360ms step_avg:144.96ms step:1952/3200 train_loss:3.3265 train_time:281504ms step_avg:144.96ms step:1953/3200 train_loss:3.5312 train_time:281648ms step_avg:144.95ms step:1954/3200 train_loss:3.5290 train_time:281791ms step_avg:144.95ms step:1955/3200 train_loss:3.4789 train_time:281935ms step_avg:144.95ms step:1956/3200 train_loss:3.3669 train_time:282078ms step_avg:144.95ms step:1957/3200 train_loss:3.4640 train_time:282219ms step_avg:144.95ms step:1958/3200 train_loss:3.6405 train_time:282363ms step_avg:144.95ms step:1959/3200 train_loss:3.5668 train_time:282506ms step_avg:144.95ms step:1960/3200 train_loss:3.5808 train_time:282652ms step_avg:144.95ms step:1961/3200 train_loss:3.3823 train_time:282796ms step_avg:144.95ms step:1962/3200 train_loss:3.5109 train_time:282938ms step_avg:144.95ms step:1963/3200 train_loss:3.5537 train_time:283083ms step_avg:144.95ms step:1964/3200 train_loss:3.5030 train_time:283227ms step_avg:144.95ms step:1965/3200 train_loss:3.4110 train_time:283372ms step_avg:144.95ms step:1966/3200 train_loss:3.8148 train_time:283514ms step_avg:144.95ms step:1967/3200 train_loss:3.4188 train_time:283658ms step_avg:144.95ms step:1968/3200 train_loss:3.4668 train_time:283801ms step_avg:144.94ms step:1969/3200 train_loss:3.5131 train_time:283945ms step_avg:144.94ms step:1970/3200 train_loss:3.4731 train_time:284091ms step_avg:144.94ms step:1971/3200 train_loss:3.3666 train_time:284234ms step_avg:144.94ms step:1972/3200 train_loss:3.3461 train_time:284376ms step_avg:144.94ms step:1973/3200 train_loss:3.4684 train_time:284518ms step_avg:144.94ms step:1974/3200 train_loss:3.4388 train_time:284661ms step_avg:144.94ms step:1975/3200 train_loss:3.4112 train_time:284804ms step_avg:144.94ms step:1976/3200 train_loss:3.5697 train_time:284949ms step_avg:144.94ms step:1977/3200 train_loss:3.4400 train_time:285093ms step_avg:144.94ms step:1978/3200 train_loss:3.8036 train_time:285236ms step_avg:144.94ms step:1979/3200 train_loss:3.4823 train_time:285379ms step_avg:144.94ms step:1980/3200 train_loss:3.4913 train_time:285523ms step_avg:144.94ms step:1981/3200 train_loss:3.4970 train_time:285665ms step_avg:144.93ms step:1982/3200 train_loss:3.5198 train_time:285807ms step_avg:144.93ms step:1983/3200 train_loss:3.4470 train_time:285953ms step_avg:144.93ms step:1984/3200 train_loss:3.4048 train_time:286096ms step_avg:144.93ms step:1985/3200 train_loss:3.4692 train_time:286239ms step_avg:144.93ms step:1986/3200 train_loss:3.5286 train_time:286383ms step_avg:144.93ms step:1987/3200 train_loss:3.4971 train_time:286527ms step_avg:144.93ms step:1988/3200 train_loss:3.4756 train_time:286669ms step_avg:144.93ms step:1989/3200 train_loss:3.5600 train_time:286812ms step_avg:144.93ms step:1990/3200 train_loss:3.5961 train_time:286955ms step_avg:144.93ms step:1991/3200 train_loss:3.3726 train_time:287097ms step_avg:144.93ms step:1992/3200 train_loss:3.3659 train_time:287241ms step_avg:144.92ms step:1993/3200 train_loss:3.5479 train_time:287387ms step_avg:144.93ms step:1994/3200 train_loss:3.3749 train_time:287531ms step_avg:144.92ms step:1995/3200 train_loss:3.4541 train_time:287676ms step_avg:144.92ms step:1996/3200 train_loss:3.5396 train_time:287819ms step_avg:144.92ms step:1997/3200 train_loss:3.3972 train_time:287962ms step_avg:144.92ms step:1998/3200 train_loss:3.5040 train_time:288105ms step_avg:144.92ms step:1999/3200 train_loss:3.5025 train_time:288248ms step_avg:144.92ms step:2000/3200 train_loss:3.4278 train_time:288392ms step_avg:144.92ms step:2000/3200 val_loss:3.4779 train_time:288446ms step_avg:144.95ms step:2001/3200 train_loss:3.5672 train_time:288542ms step_avg:144.92ms step:2002/3200 train_loss:3.5113 train_time:288692ms step_avg:144.93ms step:2003/3200 train_loss:3.5992 train_time:288834ms step_avg:144.92ms step:2004/3200 train_loss:3.5215 train_time:288976ms step_avg:144.92ms step:2005/3200 train_loss:3.5303 train_time:289118ms step_avg:144.92ms step:2006/3200 train_loss:3.4196 train_time:289260ms step_avg:144.92ms step:2007/3200 train_loss:3.4463 train_time:289402ms step_avg:144.92ms step:2008/3200 train_loss:3.4922 train_time:289548ms step_avg:144.92ms step:2009/3200 train_loss:3.5309 train_time:289695ms step_avg:144.92ms step:2010/3200 train_loss:3.4309 train_time:289838ms step_avg:144.92ms step:2011/3200 train_loss:3.5108 train_time:289980ms step_avg:144.92ms step:2012/3200 train_loss:3.4848 train_time:290122ms step_avg:144.92ms step:2013/3200 train_loss:3.4897 train_time:290264ms step_avg:144.91ms step:2014/3200 train_loss:3.4066 train_time:290407ms step_avg:144.91ms step:2015/3200 train_loss:3.4500 train_time:290551ms step_avg:144.91ms step:2016/3200 train_loss:3.4689 train_time:290696ms step_avg:144.91ms step:2017/3200 train_loss:3.6017 train_time:290840ms step_avg:144.91ms step:2018/3200 train_loss:3.4545 train_time:290983ms step_avg:144.91ms step:2019/3200 train_loss:3.6027 train_time:291125ms step_avg:144.91ms step:2020/3200 train_loss:3.6136 train_time:291267ms step_avg:144.91ms step:2021/3200 train_loss:3.3181 train_time:291410ms step_avg:144.91ms step:2022/3200 train_loss:3.5524 train_time:291554ms step_avg:144.91ms step:2023/3200 train_loss:3.4814 train_time:291697ms step_avg:144.91ms step:2024/3200 train_loss:3.5780 train_time:291840ms step_avg:144.91ms step:2025/3200 train_loss:3.6194 train_time:291984ms step_avg:144.91ms step:2026/3200 train_loss:3.4069 train_time:292126ms step_avg:144.90ms step:2027/3200 train_loss:3.4300 train_time:292269ms step_avg:144.90ms step:2028/3200 train_loss:3.3446 train_time:292413ms step_avg:144.90ms step:2029/3200 train_loss:3.4570 train_time:292555ms step_avg:144.90ms step:2030/3200 train_loss:3.3758 train_time:292698ms step_avg:144.90ms step:2031/3200 train_loss:3.4708 train_time:292842ms step_avg:144.90ms step:2032/3200 train_loss:3.4704 train_time:292986ms step_avg:144.90ms step:2033/3200 train_loss:3.4797 train_time:293129ms step_avg:144.90ms step:2034/3200 train_loss:3.3755 train_time:293273ms step_avg:144.90ms step:2035/3200 train_loss:3.5330 train_time:293416ms step_avg:144.90ms step:2036/3200 train_loss:3.5375 train_time:293558ms step_avg:144.90ms step:2037/3200 train_loss:3.5216 train_time:293701ms step_avg:144.89ms step:2038/3200 train_loss:3.3921 train_time:293844ms step_avg:144.89ms step:2039/3200 train_loss:3.6564 train_time:293987ms step_avg:144.89ms step:2040/3200 train_loss:3.4941 train_time:294130ms step_avg:144.89ms step:2041/3200 train_loss:3.5068 train_time:294275ms step_avg:144.89ms step:2042/3200 train_loss:3.4546 train_time:294418ms step_avg:144.89ms step:2043/3200 train_loss:3.3585 train_time:294562ms step_avg:144.89ms step:2044/3200 train_loss:3.4818 train_time:294705ms step_avg:144.89ms step:2045/3200 train_loss:3.4748 train_time:294849ms step_avg:144.89ms step:2046/3200 train_loss:3.3437 train_time:294993ms step_avg:144.89ms step:2047/3200 train_loss:3.4278 train_time:295136ms step_avg:144.89ms step:2048/3200 train_loss:3.4970 train_time:295279ms step_avg:144.89ms step:2049/3200 train_loss:3.4479 train_time:295423ms step_avg:144.89ms step:2050/3200 train_loss:3.4978 train_time:295566ms step_avg:144.89ms step:2051/3200 train_loss:3.6376 train_time:295709ms step_avg:144.88ms step:2052/3200 train_loss:3.5069 train_time:295853ms step_avg:144.88ms step:2053/3200 train_loss:3.4560 train_time:295996ms step_avg:144.88ms step:2054/3200 train_loss:3.4357 train_time:296140ms step_avg:144.88ms step:2055/3200 train_loss:3.3056 train_time:296283ms step_avg:144.88ms step:2056/3200 train_loss:3.4111 train_time:296426ms step_avg:144.88ms step:2057/3200 train_loss:3.5915 train_time:296570ms step_avg:144.88ms step:2058/3200 train_loss:3.6179 train_time:296714ms step_avg:144.88ms step:2059/3200 train_loss:3.4731 train_time:296856ms step_avg:144.88ms step:2060/3200 train_loss:3.5148 train_time:297003ms step_avg:144.88ms step:2061/3200 train_loss:3.4973 train_time:297146ms step_avg:144.88ms step:2062/3200 train_loss:3.4545 train_time:297290ms step_avg:144.88ms step:2063/3200 train_loss:3.3663 train_time:297435ms step_avg:144.88ms step:2064/3200 train_loss:3.6741 train_time:297577ms step_avg:144.88ms step:2065/3200 train_loss:3.5314 train_time:297719ms step_avg:144.88ms step:2066/3200 train_loss:3.4852 train_time:297862ms step_avg:144.87ms step:2067/3200 train_loss:3.5306 train_time:298006ms step_avg:144.87ms step:2068/3200 train_loss:3.4240 train_time:298150ms step_avg:144.87ms step:2069/3200 train_loss:3.4905 train_time:298298ms step_avg:144.88ms step:2070/3200 train_loss:3.6209 train_time:298439ms step_avg:144.87ms step:2071/3200 train_loss:3.6209 train_time:298582ms step_avg:144.87ms step:2072/3200 train_loss:3.4736 train_time:298727ms step_avg:144.87ms step:2073/3200 train_loss:3.5107 train_time:298870ms step_avg:144.87ms step:2074/3200 train_loss:3.3968 train_time:299015ms step_avg:144.87ms step:2075/3200 train_loss:3.9202 train_time:299158ms step_avg:144.87ms step:2076/3200 train_loss:3.3479 train_time:299301ms step_avg:144.87ms step:2077/3200 train_loss:3.5211 train_time:299443ms step_avg:144.87ms step:2078/3200 train_loss:3.4040 train_time:299587ms step_avg:144.87ms step:2079/3200 train_loss:3.3864 train_time:299889ms step_avg:144.94ms step:2080/3200 train_loss:3.4721 train_time:300033ms step_avg:144.94ms step:2081/3200 train_loss:3.7310 train_time:300175ms step_avg:144.94ms step:2082/3200 train_loss:3.3569 train_time:300318ms step_avg:144.94ms step:2083/3200 train_loss:3.6974 train_time:300458ms step_avg:144.94ms step:2084/3200 train_loss:3.4024 train_time:300600ms step_avg:144.94ms step:2085/3200 train_loss:3.3856 train_time:300741ms step_avg:144.94ms step:2086/3200 train_loss:3.6297 train_time:300890ms step_avg:144.94ms step:2087/3200 train_loss:3.5506 train_time:301036ms step_avg:144.94ms step:2088/3200 train_loss:3.5384 train_time:301178ms step_avg:144.94ms step:2089/3200 train_loss:3.5974 train_time:301321ms step_avg:144.94ms step:2090/3200 train_loss:3.5254 train_time:301627ms step_avg:145.01ms step:2091/3200 train_loss:3.5118 train_time:301769ms step_avg:145.01ms step:2092/3200 train_loss:3.4673 train_time:301912ms step_avg:145.01ms step:2093/3200 train_loss:3.5359 train_time:302053ms step_avg:145.01ms step:2094/3200 train_loss:3.4425 train_time:302196ms step_avg:145.01ms step:2095/3200 train_loss:3.2370 train_time:302338ms step_avg:145.01ms step:2096/3200 train_loss:3.4630 train_time:302480ms step_avg:145.01ms step:2097/3200 train_loss:3.6247 train_time:302628ms step_avg:145.01ms step:2098/3200 train_loss:3.4564 train_time:302773ms step_avg:145.01ms step:2099/3200 train_loss:3.3472 train_time:302916ms step_avg:145.01ms step:2100/3200 train_loss:3.4514 train_time:303057ms step_avg:145.00ms step:2101/3200 train_loss:3.4063 train_time:303199ms step_avg:145.00ms step:2102/3200 train_loss:3.5526 train_time:303341ms step_avg:145.00ms step:2103/3200 train_loss:3.3849 train_time:303486ms step_avg:145.00ms step:2104/3200 train_loss:3.3483 train_time:303631ms step_avg:145.00ms step:2105/3200 train_loss:3.6101 train_time:303776ms step_avg:145.00ms step:2106/3200 train_loss:3.3425 train_time:303918ms step_avg:145.00ms step:2107/3200 train_loss:3.7418 train_time:304061ms step_avg:145.00ms step:2108/3200 train_loss:3.5720 train_time:304204ms step_avg:145.00ms step:2109/3200 train_loss:3.4707 train_time:304346ms step_avg:145.00ms step:2110/3200 train_loss:3.5003 train_time:304490ms step_avg:145.00ms step:2111/3200 train_loss:3.3189 train_time:304634ms step_avg:144.99ms step:2112/3200 train_loss:3.7989 train_time:304778ms step_avg:144.99ms step:2113/3200 train_loss:3.4873 train_time:304920ms step_avg:144.99ms step:2114/3200 train_loss:3.4238 train_time:305063ms step_avg:144.99ms step:2115/3200 train_loss:3.5382 train_time:305206ms step_avg:144.99ms step:2116/3200 train_loss:3.4953 train_time:305349ms step_avg:144.99ms step:2117/3200 train_loss:3.4827 train_time:305493ms step_avg:144.99ms step:2118/3200 train_loss:3.5400 train_time:305637ms step_avg:144.99ms step:2119/3200 train_loss:3.3931 train_time:305780ms step_avg:144.99ms step:2120/3200 train_loss:3.4589 train_time:305923ms step_avg:144.99ms step:2121/3200 train_loss:3.1568 train_time:306065ms step_avg:144.99ms step:2122/3200 train_loss:3.3543 train_time:306210ms step_avg:144.99ms step:2123/3200 train_loss:3.5193 train_time:306354ms step_avg:144.99ms step:2124/3200 train_loss:3.4357 train_time:306497ms step_avg:144.98ms step:2125/3200 train_loss:3.5886 train_time:306640ms step_avg:144.98ms step:2125/3200 val_loss:3.4675 train_time:306696ms step_avg:145.01ms step:2126/3200 train_loss:3.4588 train_time:306796ms step_avg:144.99ms step:2127/3200 train_loss:3.5653 train_time:306942ms step_avg:144.99ms step:2128/3200 train_loss:3.5476 train_time:307084ms step_avg:144.99ms step:2129/3200 train_loss:3.4312 train_time:307226ms step_avg:144.99ms step:2130/3200 train_loss:3.4000 train_time:307367ms step_avg:144.98ms step:2131/3200 train_loss:3.4211 train_time:307510ms step_avg:144.98ms step:2132/3200 train_loss:3.5717 train_time:307652ms step_avg:144.98ms step:2133/3200 train_loss:3.4586 train_time:307796ms step_avg:144.98ms step:2134/3200 train_loss:3.3574 train_time:307941ms step_avg:144.98ms step:2135/3200 train_loss:3.4267 train_time:308085ms step_avg:144.98ms step:2136/3200 train_loss:3.5498 train_time:308227ms step_avg:144.98ms step:2137/3200 train_loss:3.5617 train_time:308370ms step_avg:144.98ms step:2138/3200 train_loss:3.5087 train_time:308513ms step_avg:144.98ms step:2139/3200 train_loss:3.4979 train_time:308656ms step_avg:144.98ms step:2140/3200 train_loss:3.4764 train_time:308800ms step_avg:144.98ms step:2141/3200 train_loss:3.5648 train_time:308944ms step_avg:144.98ms step:2142/3200 train_loss:3.8691 train_time:309087ms step_avg:144.98ms step:2143/3200 train_loss:3.3964 train_time:309230ms step_avg:144.97ms step:2144/3200 train_loss:3.4270 train_time:309374ms step_avg:144.97ms step:2145/3200 train_loss:3.4677 train_time:309517ms step_avg:144.97ms step:2146/3200 train_loss:3.5977 train_time:309660ms step_avg:144.97ms step:2147/3200 train_loss:3.5214 train_time:309803ms step_avg:144.97ms step:2148/3200 train_loss:3.9311 train_time:309949ms step_avg:144.97ms step:2149/3200 train_loss:3.4511 train_time:310092ms step_avg:144.97ms step:2150/3200 train_loss:3.4219 train_time:310236ms step_avg:144.97ms step:2151/3200 train_loss:3.4972 train_time:310379ms step_avg:144.97ms step:2152/3200 train_loss:3.5173 train_time:310521ms step_avg:144.97ms step:2153/3200 train_loss:3.4756 train_time:310663ms step_avg:144.97ms step:2154/3200 train_loss:3.4120 train_time:310806ms step_avg:144.97ms step:2155/3200 train_loss:3.6226 train_time:310950ms step_avg:144.97ms step:2156/3200 train_loss:3.2486 train_time:311094ms step_avg:144.96ms step:2157/3200 train_loss:3.4061 train_time:311238ms step_avg:144.96ms step:2158/3200 train_loss:3.5418 train_time:311380ms step_avg:144.96ms step:2159/3200 train_loss:3.4790 train_time:311523ms step_avg:144.96ms step:2160/3200 train_loss:3.6355 train_time:311665ms step_avg:144.96ms step:2161/3200 train_loss:3.5480 train_time:311810ms step_avg:144.96ms step:2162/3200 train_loss:3.4767 train_time:311955ms step_avg:144.96ms step:2163/3200 train_loss:3.4493 train_time:312099ms step_avg:144.96ms step:2164/3200 train_loss:3.4446 train_time:312242ms step_avg:144.96ms step:2165/3200 train_loss:3.5273 train_time:312385ms step_avg:144.96ms step:2166/3200 train_loss:3.5480 train_time:312528ms step_avg:144.96ms step:2167/3200 train_loss:3.4799 train_time:312673ms step_avg:144.96ms step:2168/3200 train_loss:3.3765 train_time:312817ms step_avg:144.96ms step:2169/3200 train_loss:3.4613 train_time:312961ms step_avg:144.96ms step:2170/3200 train_loss:3.5028 train_time:313103ms step_avg:144.96ms step:2171/3200 train_loss:3.6258 train_time:313246ms step_avg:144.95ms step:2172/3200 train_loss:3.4115 train_time:313390ms step_avg:144.95ms step:2173/3200 train_loss:3.4071 train_time:313534ms step_avg:144.95ms step:2174/3200 train_loss:3.4182 train_time:313677ms step_avg:144.95ms step:2175/3200 train_loss:3.4680 train_time:313821ms step_avg:144.95ms step:2176/3200 train_loss:3.4380 train_time:313964ms step_avg:144.95ms step:2177/3200 train_loss:3.4071 train_time:314108ms step_avg:144.95ms step:2178/3200 train_loss:3.6258 train_time:314251ms step_avg:144.95ms step:2179/3200 train_loss:3.4468 train_time:314395ms step_avg:144.95ms step:2180/3200 train_loss:3.4620 train_time:314539ms step_avg:144.95ms step:2181/3200 train_loss:3.5187 train_time:314682ms step_avg:144.95ms step:2182/3200 train_loss:3.4919 train_time:314824ms step_avg:144.95ms step:2183/3200 train_loss:3.4602 train_time:314970ms step_avg:144.95ms step:2184/3200 train_loss:3.3612 train_time:315114ms step_avg:144.95ms step:2185/3200 train_loss:3.5314 train_time:315258ms step_avg:144.95ms step:2186/3200 train_loss:3.6981 train_time:315402ms step_avg:144.95ms step:2187/3200 train_loss:3.3472 train_time:315545ms step_avg:144.94ms step:2188/3200 train_loss:3.3900 train_time:315688ms step_avg:144.94ms step:2189/3200 train_loss:3.2305 train_time:315832ms step_avg:144.94ms step:2190/3200 train_loss:3.3877 train_time:315977ms step_avg:144.94ms step:2191/3200 train_loss:3.5359 train_time:316120ms step_avg:144.94ms step:2192/3200 train_loss:3.4700 train_time:316264ms step_avg:144.94ms step:2193/3200 train_loss:3.6960 train_time:316407ms step_avg:144.94ms step:2194/3200 train_loss:3.4653 train_time:316550ms step_avg:144.94ms step:2195/3200 train_loss:3.5335 train_time:316694ms step_avg:144.94ms step:2196/3200 train_loss:3.4654 train_time:316838ms step_avg:144.94ms step:2197/3200 train_loss:3.3946 train_time:316981ms step_avg:144.94ms step:2198/3200 train_loss:3.4762 train_time:317123ms step_avg:144.94ms step:2199/3200 train_loss:3.4146 train_time:317267ms step_avg:144.94ms step:2200/3200 train_loss:3.4193 train_time:317411ms step_avg:144.94ms step:2201/3200 train_loss:3.4722 train_time:317555ms step_avg:144.94ms step:2202/3200 train_loss:3.4525 train_time:317699ms step_avg:144.94ms step:2203/3200 train_loss:3.4335 train_time:317842ms step_avg:144.93ms step:2204/3200 train_loss:3.9364 train_time:317984ms step_avg:144.93ms step:2205/3200 train_loss:3.3455 train_time:318128ms step_avg:144.93ms step:2206/3200 train_loss:3.4707 train_time:318273ms step_avg:144.93ms step:2207/3200 train_loss:3.4869 train_time:318417ms step_avg:144.93ms step:2208/3200 train_loss:3.4994 train_time:318560ms step_avg:144.93ms step:2209/3200 train_loss:3.4014 train_time:318702ms step_avg:144.93ms step:2210/3200 train_loss:3.4766 train_time:318846ms step_avg:144.93ms step:2211/3200 train_loss:3.4840 train_time:318990ms step_avg:144.93ms step:2212/3200 train_loss:3.4851 train_time:319134ms step_avg:144.93ms step:2213/3200 train_loss:3.5060 train_time:319278ms step_avg:144.93ms step:2214/3200 train_loss:3.3585 train_time:319421ms step_avg:144.93ms step:2215/3200 train_loss:3.4295 train_time:319564ms step_avg:144.93ms step:2216/3200 train_loss:3.5685 train_time:319706ms step_avg:144.93ms step:2217/3200 train_loss:3.5140 train_time:319850ms step_avg:144.93ms step:2218/3200 train_loss:3.4852 train_time:319993ms step_avg:144.92ms step:2219/3200 train_loss:3.4900 train_time:320137ms step_avg:144.92ms step:2220/3200 train_loss:3.3990 train_time:320280ms step_avg:144.92ms step:2221/3200 train_loss:3.6556 train_time:320422ms step_avg:144.92ms step:2222/3200 train_loss:3.5423 train_time:320566ms step_avg:144.92ms step:2223/3200 train_loss:3.5669 train_time:320711ms step_avg:144.92ms step:2224/3200 train_loss:3.4430 train_time:320855ms step_avg:144.92ms step:2225/3200 train_loss:3.5694 train_time:320998ms step_avg:144.92ms step:2226/3200 train_loss:3.3268 train_time:321141ms step_avg:144.92ms step:2227/3200 train_loss:3.5924 train_time:321283ms step_avg:144.92ms step:2228/3200 train_loss:3.5332 train_time:321427ms step_avg:144.92ms step:2229/3200 train_loss:3.3365 train_time:321570ms step_avg:144.92ms step:2230/3200 train_loss:3.6757 train_time:321715ms step_avg:144.92ms step:2231/3200 train_loss:3.3729 train_time:321859ms step_avg:144.92ms step:2232/3200 train_loss:3.8411 train_time:322002ms step_avg:144.92ms step:2233/3200 train_loss:3.5257 train_time:322146ms step_avg:144.91ms step:2234/3200 train_loss:3.4740 train_time:322288ms step_avg:144.91ms step:2235/3200 train_loss:3.5111 train_time:322432ms step_avg:144.91ms step:2236/3200 train_loss:3.2851 train_time:322577ms step_avg:144.91ms step:2237/3200 train_loss:3.2893 train_time:322720ms step_avg:144.91ms step:2238/3200 train_loss:3.5084 train_time:322863ms step_avg:144.91ms step:2239/3200 train_loss:3.6037 train_time:323006ms step_avg:144.91ms step:2240/3200 train_loss:3.3272 train_time:323151ms step_avg:144.91ms step:2241/3200 train_loss:3.3971 train_time:323294ms step_avg:144.91ms step:2242/3200 train_loss:3.5785 train_time:323437ms step_avg:144.91ms step:2243/3200 train_loss:3.5479 train_time:323580ms step_avg:144.91ms step:2244/3200 train_loss:3.4060 train_time:323723ms step_avg:144.91ms step:2245/3200 train_loss:3.4720 train_time:323867ms step_avg:144.91ms step:2246/3200 train_loss:3.4960 train_time:324013ms step_avg:144.91ms step:2247/3200 train_loss:3.3258 train_time:324157ms step_avg:144.91ms step:2248/3200 train_loss:3.3500 train_time:324300ms step_avg:144.91ms step:2249/3200 train_loss:3.6108 train_time:324442ms step_avg:144.90ms step:2250/3200 train_loss:3.3309 train_time:324586ms step_avg:144.90ms step:2250/3200 val_loss:3.4574 train_time:324640ms step_avg:144.93ms step:2251/3200 train_loss:3.3378 train_time:324739ms step_avg:144.91ms step:2252/3200 train_loss:3.4097 train_time:324886ms step_avg:144.91ms step:2253/3200 train_loss:3.3791 train_time:325029ms step_avg:144.91ms step:2254/3200 train_loss:3.4390 train_time:325171ms step_avg:144.91ms step:2255/3200 train_loss:3.4835 train_time:325313ms step_avg:144.91ms step:2256/3200 train_loss:3.3619 train_time:325456ms step_avg:144.90ms step:2257/3200 train_loss:3.6494 train_time:325598ms step_avg:144.90ms step:2258/3200 train_loss:3.5247 train_time:325743ms step_avg:144.90ms step:2259/3200 train_loss:3.8394 train_time:325888ms step_avg:144.90ms step:2260/3200 train_loss:3.5263 train_time:326031ms step_avg:144.90ms step:2261/3200 train_loss:3.5735 train_time:326176ms step_avg:144.90ms step:2262/3200 train_loss:3.4880 train_time:326318ms step_avg:144.90ms step:2263/3200 train_loss:3.4899 train_time:326462ms step_avg:144.90ms step:2264/3200 train_loss:3.2501 train_time:326605ms step_avg:144.90ms step:2265/3200 train_loss:3.3707 train_time:326749ms step_avg:144.90ms step:2266/3200 train_loss:3.5882 train_time:326893ms step_avg:144.90ms step:2267/3200 train_loss:3.3206 train_time:327037ms step_avg:144.90ms step:2268/3200 train_loss:3.3988 train_time:327331ms step_avg:144.97ms step:2269/3200 train_loss:3.3701 train_time:327474ms step_avg:144.96ms step:2270/3200 train_loss:3.3388 train_time:327616ms step_avg:144.96ms step:2271/3200 train_loss:3.7415 train_time:327756ms step_avg:144.96ms step:2272/3200 train_loss:3.3893 train_time:327898ms step_avg:144.96ms step:2273/3200 train_loss:3.3956 train_time:328041ms step_avg:144.96ms step:2274/3200 train_loss:3.4842 train_time:328183ms step_avg:144.96ms step:2275/3200 train_loss:3.4327 train_time:328330ms step_avg:144.96ms step:2276/3200 train_loss:3.4437 train_time:328475ms step_avg:144.96ms step:2277/3200 train_loss:3.3229 train_time:328618ms step_avg:144.96ms step:2278/3200 train_loss:3.4296 train_time:328762ms step_avg:144.96ms step:2279/3200 train_loss:3.5531 train_time:328904ms step_avg:144.96ms step:2280/3200 train_loss:3.3579 train_time:329204ms step_avg:145.02ms step:2281/3200 train_loss:3.4203 train_time:329346ms step_avg:145.02ms step:2282/3200 train_loss:3.4317 train_time:329487ms step_avg:145.02ms step:2283/3200 train_loss:3.5774 train_time:329628ms step_avg:145.02ms step:2284/3200 train_loss:3.4504 train_time:329770ms step_avg:145.02ms step:2285/3200 train_loss:3.4724 train_time:329914ms step_avg:145.02ms step:2286/3200 train_loss:3.4730 train_time:330055ms step_avg:145.02ms step:2287/3200 train_loss:3.4679 train_time:330204ms step_avg:145.02ms step:2288/3200 train_loss:3.4212 train_time:330349ms step_avg:145.02ms step:2289/3200 train_loss:3.5535 train_time:330492ms step_avg:145.02ms step:2290/3200 train_loss:3.5277 train_time:330635ms step_avg:145.02ms step:2291/3200 train_loss:3.4106 train_time:330780ms step_avg:145.02ms step:2292/3200 train_loss:3.7465 train_time:330922ms step_avg:145.01ms step:2293/3200 train_loss:3.4064 train_time:331066ms step_avg:145.01ms step:2294/3200 train_loss:3.3611 train_time:331209ms step_avg:145.01ms step:2295/3200 train_loss:3.5394 train_time:331353ms step_avg:145.01ms step:2296/3200 train_loss:3.4900 train_time:331497ms step_avg:145.01ms step:2297/3200 train_loss:3.4693 train_time:331640ms step_avg:145.01ms step:2298/3200 train_loss:3.8455 train_time:331783ms step_avg:145.01ms step:2299/3200 train_loss:3.3625 train_time:331925ms step_avg:145.01ms step:2300/3200 train_loss:3.3583 train_time:332067ms step_avg:145.01ms step:2301/3200 train_loss:3.6888 train_time:332212ms step_avg:145.01ms step:2302/3200 train_loss:3.4249 train_time:332354ms step_avg:145.01ms step:2303/3200 train_loss:3.4380 train_time:332498ms step_avg:145.01ms step:2304/3200 train_loss:3.4282 train_time:332642ms step_avg:145.01ms step:2305/3200 train_loss:3.3638 train_time:332786ms step_avg:145.00ms step:2306/3200 train_loss:3.5165 train_time:332929ms step_avg:145.00ms step:2307/3200 train_loss:3.3851 train_time:333071ms step_avg:145.00ms step:2308/3200 train_loss:3.4002 train_time:333216ms step_avg:145.00ms step:2309/3200 train_loss:3.5343 train_time:333361ms step_avg:145.00ms step:2310/3200 train_loss:3.4925 train_time:333505ms step_avg:145.00ms step:2311/3200 train_loss:3.3585 train_time:333647ms step_avg:145.00ms step:2312/3200 train_loss:3.4721 train_time:333790ms step_avg:145.00ms step:2313/3200 train_loss:3.5975 train_time:333933ms step_avg:145.00ms step:2314/3200 train_loss:3.4167 train_time:334077ms step_avg:145.00ms step:2315/3200 train_loss:3.3446 train_time:334222ms step_avg:145.00ms step:2316/3200 train_loss:3.4334 train_time:334367ms step_avg:145.00ms step:2317/3200 train_loss:3.3168 train_time:334509ms step_avg:145.00ms step:2318/3200 train_loss:3.4226 train_time:334653ms step_avg:145.00ms step:2319/3200 train_loss:3.4417 train_time:334795ms step_avg:145.00ms step:2320/3200 train_loss:3.2851 train_time:334938ms step_avg:144.99ms step:2321/3200 train_loss:3.4271 train_time:335082ms step_avg:144.99ms step:2322/3200 train_loss:3.4754 train_time:335225ms step_avg:144.99ms step:2323/3200 train_loss:3.3921 train_time:335369ms step_avg:144.99ms step:2324/3200 train_loss:3.4376 train_time:335513ms step_avg:144.99ms step:2325/3200 train_loss:3.3570 train_time:335657ms step_avg:144.99ms step:2326/3200 train_loss:3.4975 train_time:335800ms step_avg:144.99ms step:2327/3200 train_loss:3.5016 train_time:335945ms step_avg:144.99ms step:2328/3200 train_loss:3.2754 train_time:336087ms step_avg:144.99ms step:2329/3200 train_loss:3.3933 train_time:336230ms step_avg:144.99ms step:2330/3200 train_loss:3.4175 train_time:336373ms step_avg:144.99ms step:2331/3200 train_loss:3.3905 train_time:336517ms step_avg:144.99ms step:2332/3200 train_loss:3.5668 train_time:336662ms step_avg:144.99ms step:2333/3200 train_loss:3.4498 train_time:336806ms step_avg:144.99ms step:2334/3200 train_loss:3.4270 train_time:336948ms step_avg:144.99ms step:2335/3200 train_loss:3.5084 train_time:337090ms step_avg:144.99ms step:2336/3200 train_loss:3.3441 train_time:337234ms step_avg:144.98ms step:2337/3200 train_loss:3.4990 train_time:337377ms step_avg:144.98ms step:2338/3200 train_loss:3.4587 train_time:337522ms step_avg:144.98ms step:2339/3200 train_loss:3.4079 train_time:337666ms step_avg:144.98ms step:2340/3200 train_loss:3.4904 train_time:337809ms step_avg:144.98ms step:2341/3200 train_loss:3.5364 train_time:337950ms step_avg:144.98ms step:2342/3200 train_loss:3.4015 train_time:338094ms step_avg:144.98ms step:2343/3200 train_loss:3.4139 train_time:338237ms step_avg:144.98ms step:2344/3200 train_loss:3.4815 train_time:338380ms step_avg:144.98ms step:2345/3200 train_loss:3.4209 train_time:338525ms step_avg:144.98ms step:2346/3200 train_loss:3.5364 train_time:338667ms step_avg:144.98ms step:2347/3200 train_loss:3.4405 train_time:338810ms step_avg:144.98ms step:2348/3200 train_loss:3.5521 train_time:338952ms step_avg:144.98ms step:2349/3200 train_loss:3.5155 train_time:339096ms step_avg:144.97ms step:2350/3200 train_loss:3.5475 train_time:339240ms step_avg:144.97ms step:2351/3200 train_loss:3.2521 train_time:339385ms step_avg:144.97ms step:2352/3200 train_loss:3.3679 train_time:339528ms step_avg:144.97ms step:2353/3200 train_loss:3.3622 train_time:339671ms step_avg:144.97ms step:2354/3200 train_loss:3.5848 train_time:339814ms step_avg:144.97ms step:2355/3200 train_loss:3.3756 train_time:339959ms step_avg:144.97ms step:2356/3200 train_loss:3.3681 train_time:340102ms step_avg:144.97ms step:2357/3200 train_loss:3.5224 train_time:340246ms step_avg:144.97ms step:2358/3200 train_loss:3.3702 train_time:340388ms step_avg:144.97ms step:2359/3200 train_loss:3.4683 train_time:340532ms step_avg:144.97ms step:2360/3200 train_loss:3.3780 train_time:340677ms step_avg:144.97ms step:2361/3200 train_loss:3.3897 train_time:340821ms step_avg:144.97ms step:2362/3200 train_loss:3.4167 train_time:340965ms step_avg:144.97ms step:2363/3200 train_loss:3.4814 train_time:341108ms step_avg:144.97ms step:2364/3200 train_loss:3.4290 train_time:341250ms step_avg:144.97ms step:2365/3200 train_loss:3.8691 train_time:341393ms step_avg:144.97ms step:2366/3200 train_loss:3.4945 train_time:341537ms step_avg:144.96ms step:2367/3200 train_loss:3.6329 train_time:341681ms step_avg:144.96ms step:2368/3200 train_loss:3.4578 train_time:341825ms step_avg:144.96ms step:2369/3200 train_loss:3.4619 train_time:341970ms step_avg:144.96ms step:2370/3200 train_loss:3.4935 train_time:342113ms step_avg:144.96ms step:2371/3200 train_loss:3.3723 train_time:342257ms step_avg:144.96ms step:2372/3200 train_loss:3.6085 train_time:342401ms step_avg:144.96ms step:2373/3200 train_loss:3.4534 train_time:342546ms step_avg:144.96ms step:2374/3200 train_loss:4.0005 train_time:342689ms step_avg:144.96ms step:2375/3200 train_loss:3.4361 train_time:342832ms step_avg:144.96ms step:2375/3200 val_loss:3.4408 train_time:342887ms step_avg:144.98ms step:2376/3200 train_loss:3.3415 train_time:342987ms step_avg:144.96ms step:2377/3200 train_loss:3.5025 train_time:343134ms step_avg:144.97ms step:2378/3200 train_loss:3.4702 train_time:343278ms step_avg:144.97ms step:2379/3200 train_loss:3.4891 train_time:343419ms step_avg:144.96ms step:2380/3200 train_loss:3.4668 train_time:343560ms step_avg:144.96ms step:2381/3200 train_loss:3.3694 train_time:343702ms step_avg:144.96ms step:2382/3200 train_loss:3.4679 train_time:343845ms step_avg:144.96ms step:2383/3200 train_loss:3.4823 train_time:343991ms step_avg:144.96ms step:2384/3200 train_loss:3.4312 train_time:344136ms step_avg:144.96ms step:2385/3200 train_loss:3.3619 train_time:344281ms step_avg:144.96ms step:2386/3200 train_loss:3.4721 train_time:344424ms step_avg:144.96ms step:2387/3200 train_loss:3.4200 train_time:344567ms step_avg:144.96ms step:2388/3200 train_loss:3.4267 train_time:344710ms step_avg:144.96ms step:2389/3200 train_loss:3.4635 train_time:344853ms step_avg:144.96ms step:2390/3200 train_loss:3.4427 train_time:344996ms step_avg:144.96ms step:2391/3200 train_loss:3.4479 train_time:345140ms step_avg:144.96ms step:2392/3200 train_loss:3.3228 train_time:345284ms step_avg:144.96ms step:2393/3200 train_loss:3.5438 train_time:345427ms step_avg:144.95ms step:2394/3200 train_loss:3.3817 train_time:345569ms step_avg:144.95ms step:2395/3200 train_loss:3.4841 train_time:345712ms step_avg:144.95ms step:2396/3200 train_loss:3.5948 train_time:345854ms step_avg:144.95ms step:2397/3200 train_loss:3.6062 train_time:345996ms step_avg:144.95ms step:2398/3200 train_loss:3.5606 train_time:346142ms step_avg:144.95ms step:2399/3200 train_loss:3.5282 train_time:346288ms step_avg:144.95ms step:2400/3200 train_loss:3.4011 train_time:346432ms step_avg:144.95ms step:2401/3200 train_loss:3.4021 train_time:346575ms step_avg:144.95ms step:2402/3200 train_loss:3.5083 train_time:346716ms step_avg:144.95ms step:2403/3200 train_loss:3.3427 train_time:346859ms step_avg:144.95ms step:2404/3200 train_loss:3.4741 train_time:347004ms step_avg:144.95ms step:2405/3200 train_loss:3.6861 train_time:347149ms step_avg:144.95ms step:2406/3200 train_loss:3.4190 train_time:347292ms step_avg:144.95ms step:2407/3200 train_loss:3.5636 train_time:347436ms step_avg:144.95ms step:2408/3200 train_loss:3.4265 train_time:347579ms step_avg:144.95ms step:2409/3200 train_loss:3.3517 train_time:347722ms step_avg:144.94ms step:2410/3200 train_loss:3.4937 train_time:347864ms step_avg:144.94ms step:2411/3200 train_loss:3.2834 train_time:348008ms step_avg:144.94ms step:2412/3200 train_loss:3.7139 train_time:348152ms step_avg:144.94ms step:2413/3200 train_loss:3.4004 train_time:348295ms step_avg:144.94ms step:2414/3200 train_loss:3.4822 train_time:348438ms step_avg:144.94ms step:2415/3200 train_loss:3.3944 train_time:348582ms step_avg:144.94ms step:2416/3200 train_loss:3.4683 train_time:348724ms step_avg:144.94ms step:2417/3200 train_loss:3.2819 train_time:348867ms step_avg:144.94ms step:2418/3200 train_loss:3.2099 train_time:349012ms step_avg:144.94ms step:2419/3200 train_loss:3.5116 train_time:349155ms step_avg:144.94ms step:2420/3200 train_loss:3.3883 train_time:349298ms step_avg:144.94ms step:2421/3200 train_loss:3.4148 train_time:349443ms step_avg:144.94ms step:2422/3200 train_loss:3.5244 train_time:349587ms step_avg:144.94ms step:2423/3200 train_loss:3.5639 train_time:349731ms step_avg:144.94ms step:2424/3200 train_loss:3.3930 train_time:349874ms step_avg:144.94ms step:2425/3200 train_loss:3.4778 train_time:350016ms step_avg:144.93ms step:2426/3200 train_loss:3.4748 train_time:350160ms step_avg:144.93ms step:2427/3200 train_loss:3.4056 train_time:350304ms step_avg:144.93ms step:2428/3200 train_loss:3.3492 train_time:350449ms step_avg:144.93ms step:2429/3200 train_loss:3.4821 train_time:350591ms step_avg:144.93ms step:2430/3200 train_loss:3.3794 train_time:350734ms step_avg:144.93ms step:2431/3200 train_loss:3.4301 train_time:350877ms step_avg:144.93ms step:2432/3200 train_loss:3.4940 train_time:351020ms step_avg:144.93ms step:2433/3200 train_loss:3.4568 train_time:351163ms step_avg:144.93ms step:2434/3200 train_loss:3.3296 train_time:351308ms step_avg:144.93ms step:2435/3200 train_loss:3.2940 train_time:351450ms step_avg:144.93ms step:2436/3200 train_loss:3.4633 train_time:351593ms step_avg:144.93ms step:2437/3200 train_loss:3.3134 train_time:351736ms step_avg:144.93ms step:2438/3200 train_loss:3.3958 train_time:351880ms step_avg:144.93ms step:2439/3200 train_loss:3.4852 train_time:352022ms step_avg:144.92ms step:2440/3200 train_loss:3.4058 train_time:352166ms step_avg:144.92ms step:2441/3200 train_loss:3.4867 train_time:352311ms step_avg:144.92ms step:2442/3200 train_loss:3.3754 train_time:352454ms step_avg:144.92ms step:2443/3200 train_loss:3.4363 train_time:352597ms step_avg:144.92ms step:2444/3200 train_loss:3.3149 train_time:352740ms step_avg:144.92ms step:2445/3200 train_loss:3.3296 train_time:352885ms step_avg:144.92ms step:2446/3200 train_loss:3.4923 train_time:353030ms step_avg:144.92ms step:2447/3200 train_loss:3.3575 train_time:353173ms step_avg:144.92ms step:2448/3200 train_loss:3.4257 train_time:353316ms step_avg:144.92ms step:2449/3200 train_loss:3.5959 train_time:353460ms step_avg:144.92ms step:2450/3200 train_loss:3.4181 train_time:353604ms step_avg:144.92ms step:2451/3200 train_loss:3.4886 train_time:353748ms step_avg:144.92ms step:2452/3200 train_loss:3.3940 train_time:353891ms step_avg:144.92ms step:2453/3200 train_loss:3.4981 train_time:354034ms step_avg:144.92ms step:2454/3200 train_loss:3.3915 train_time:354177ms step_avg:144.92ms step:2455/3200 train_loss:3.5124 train_time:354320ms step_avg:144.92ms step:2456/3200 train_loss:3.4449 train_time:354462ms step_avg:144.91ms step:2457/3200 train_loss:3.3655 train_time:354759ms step_avg:144.98ms step:2458/3200 train_loss:3.2933 train_time:354902ms step_avg:144.98ms step:2459/3200 train_loss:3.4243 train_time:355045ms step_avg:144.98ms step:2460/3200 train_loss:4.0164 train_time:355187ms step_avg:144.97ms step:2461/3200 train_loss:3.4835 train_time:355329ms step_avg:144.97ms step:2462/3200 train_loss:3.3043 train_time:355471ms step_avg:144.97ms step:2463/3200 train_loss:3.4966 train_time:355614ms step_avg:144.97ms step:2464/3200 train_loss:3.4112 train_time:355763ms step_avg:144.97ms step:2465/3200 train_loss:3.6224 train_time:355909ms step_avg:144.97ms step:2466/3200 train_loss:3.7835 train_time:356051ms step_avg:144.97ms step:2467/3200 train_loss:3.5318 train_time:356193ms step_avg:144.97ms step:2468/3200 train_loss:3.3981 train_time:356335ms step_avg:144.97ms step:2469/3200 train_loss:3.5220 train_time:356477ms step_avg:144.97ms step:2470/3200 train_loss:3.5314 train_time:356777ms step_avg:145.03ms step:2471/3200 train_loss:3.3304 train_time:356920ms step_avg:145.03ms step:2472/3200 train_loss:3.4247 train_time:357061ms step_avg:145.03ms step:2473/3200 train_loss:3.4273 train_time:357205ms step_avg:145.03ms step:2474/3200 train_loss:3.5650 train_time:357347ms step_avg:145.03ms step:2475/3200 train_loss:3.6979 train_time:357489ms step_avg:145.03ms step:2476/3200 train_loss:3.2845 train_time:357632ms step_avg:145.03ms step:2477/3200 train_loss:3.4954 train_time:357779ms step_avg:145.03ms step:2478/3200 train_loss:3.4600 train_time:357925ms step_avg:145.03ms step:2479/3200 train_loss:3.2996 train_time:358069ms step_avg:145.03ms step:2480/3200 train_loss:3.2943 train_time:358212ms step_avg:145.02ms step:2481/3200 train_loss:3.4374 train_time:358353ms step_avg:145.02ms step:2482/3200 train_loss:3.4519 train_time:358495ms step_avg:145.02ms step:2483/3200 train_loss:3.4705 train_time:358638ms step_avg:145.02ms step:2484/3200 train_loss:3.4262 train_time:358783ms step_avg:145.02ms step:2485/3200 train_loss:3.4327 train_time:358928ms step_avg:145.02ms step:2486/3200 train_loss:3.3174 train_time:359073ms step_avg:145.02ms step:2487/3200 train_loss:3.5175 train_time:359215ms step_avg:145.02ms step:2488/3200 train_loss:3.4683 train_time:359357ms step_avg:145.02ms step:2489/3200 train_loss:3.3728 train_time:359498ms step_avg:145.02ms step:2490/3200 train_loss:3.4870 train_time:359644ms step_avg:145.02ms step:2491/3200 train_loss:3.5304 train_time:359789ms step_avg:145.02ms step:2492/3200 train_loss:3.6123 train_time:359933ms step_avg:145.02ms step:2493/3200 train_loss:3.4650 train_time:360077ms step_avg:145.02ms step:2494/3200 train_loss:3.3931 train_time:360220ms step_avg:145.02ms step:2495/3200 train_loss:3.5111 train_time:360362ms step_avg:145.01ms step:2496/3200 train_loss:3.4641 train_time:360504ms step_avg:145.01ms step:2497/3200 train_loss:3.3727 train_time:360647ms step_avg:145.01ms step:2498/3200 train_loss:3.4694 train_time:360792ms step_avg:145.01ms step:2499/3200 train_loss:3.5224 train_time:360935ms step_avg:145.01ms step:2500/3200 train_loss:3.5465 train_time:361078ms step_avg:145.01ms step:2500/3200 val_loss:3.4176 train_time:361134ms step_avg:145.03ms step:2501/3200 train_loss:3.4838 train_time:361233ms step_avg:145.02ms step:2502/3200 train_loss:3.4396 train_time:361377ms step_avg:145.01ms step:2503/3200 train_loss:3.4613 train_time:361520ms step_avg:145.01ms step:2504/3200 train_loss:3.3255 train_time:361664ms step_avg:145.01ms step:2505/3200 train_loss:3.5218 train_time:361807ms step_avg:145.01ms step:2506/3200 train_loss:3.4728 train_time:361949ms step_avg:145.01ms step:2507/3200 train_loss:3.4137 train_time:362091ms step_avg:145.01ms step:2508/3200 train_loss:3.4184 train_time:362236ms step_avg:145.01ms step:2509/3200 train_loss:3.3781 train_time:362382ms step_avg:145.01ms step:2510/3200 train_loss:3.5712 train_time:362527ms step_avg:145.01ms step:2511/3200 train_loss:3.3807 train_time:362672ms step_avg:145.01ms step:2512/3200 train_loss:3.3716 train_time:362814ms step_avg:145.01ms step:2513/3200 train_loss:3.4536 train_time:362956ms step_avg:145.01ms step:2514/3200 train_loss:3.4695 train_time:363098ms step_avg:145.01ms step:2515/3200 train_loss:3.3822 train_time:363244ms step_avg:145.01ms step:2516/3200 train_loss:3.4677 train_time:363388ms step_avg:145.01ms step:2517/3200 train_loss:3.4548 train_time:363534ms step_avg:145.01ms step:2518/3200 train_loss:3.3364 train_time:363676ms step_avg:145.01ms step:2519/3200 train_loss:3.3653 train_time:363819ms step_avg:145.01ms step:2520/3200 train_loss:3.4829 train_time:363963ms step_avg:145.01ms step:2521/3200 train_loss:3.4774 train_time:364106ms step_avg:145.00ms step:2522/3200 train_loss:3.3624 train_time:364251ms step_avg:145.00ms step:2523/3200 train_loss:3.3447 train_time:364394ms step_avg:145.00ms step:2524/3200 train_loss:3.4376 train_time:364537ms step_avg:145.00ms step:2525/3200 train_loss:3.2856 train_time:364680ms step_avg:145.00ms step:2526/3200 train_loss:3.5059 train_time:364823ms step_avg:145.00ms step:2527/3200 train_loss:3.4101 train_time:364967ms step_avg:145.00ms step:2528/3200 train_loss:3.4226 train_time:365110ms step_avg:145.00ms step:2529/3200 train_loss:3.4024 train_time:365252ms step_avg:145.00ms step:2530/3200 train_loss:3.4188 train_time:365396ms step_avg:145.00ms step:2531/3200 train_loss:3.4572 train_time:365539ms step_avg:145.00ms step:2532/3200 train_loss:3.2848 train_time:365684ms step_avg:145.00ms step:2533/3200 train_loss:3.4474 train_time:365828ms step_avg:145.00ms step:2534/3200 train_loss:3.3372 train_time:365972ms step_avg:145.00ms step:2535/3200 train_loss:3.3720 train_time:366114ms step_avg:145.00ms step:2536/3200 train_loss:3.4256 train_time:366258ms step_avg:145.00ms step:2537/3200 train_loss:3.4369 train_time:366400ms step_avg:144.99ms step:2538/3200 train_loss:3.2699 train_time:366545ms step_avg:144.99ms step:2539/3200 train_loss:3.5748 train_time:366690ms step_avg:144.99ms step:2540/3200 train_loss:3.2573 train_time:366833ms step_avg:144.99ms step:2541/3200 train_loss:3.4442 train_time:366975ms step_avg:144.99ms step:2542/3200 train_loss:3.2018 train_time:367118ms step_avg:144.99ms step:2543/3200 train_loss:3.6399 train_time:367264ms step_avg:144.99ms step:2544/3200 train_loss:3.4086 train_time:367408ms step_avg:144.99ms step:2545/3200 train_loss:3.5691 train_time:367552ms step_avg:144.99ms step:2546/3200 train_loss:3.4025 train_time:367696ms step_avg:144.99ms step:2547/3200 train_loss:3.3820 train_time:367839ms step_avg:144.99ms step:2548/3200 train_loss:3.3879 train_time:367982ms step_avg:144.99ms step:2549/3200 train_loss:3.5459 train_time:368125ms step_avg:144.99ms step:2550/3200 train_loss:3.4074 train_time:368269ms step_avg:144.99ms step:2551/3200 train_loss:3.4069 train_time:368411ms step_avg:144.99ms step:2552/3200 train_loss:3.4341 train_time:368554ms step_avg:144.99ms step:2553/3200 train_loss:3.4579 train_time:368698ms step_avg:144.99ms step:2554/3200 train_loss:3.3634 train_time:368841ms step_avg:144.98ms step:2555/3200 train_loss:3.4683 train_time:368984ms step_avg:144.98ms step:2556/3200 train_loss:3.5225 train_time:369127ms step_avg:144.98ms step:2557/3200 train_loss:3.5143 train_time:369270ms step_avg:144.98ms step:2558/3200 train_loss:3.3594 train_time:369413ms step_avg:144.98ms step:2559/3200 train_loss:3.3614 train_time:369556ms step_avg:144.98ms step:2560/3200 train_loss:3.3680 train_time:369699ms step_avg:144.98ms step:2561/3200 train_loss:3.4907 train_time:369843ms step_avg:144.98ms step:2562/3200 train_loss:3.5241 train_time:369987ms step_avg:144.98ms step:2563/3200 train_loss:3.4042 train_time:370130ms step_avg:144.98ms step:2564/3200 train_loss:3.4346 train_time:370274ms step_avg:144.98ms step:2565/3200 train_loss:3.3551 train_time:370416ms step_avg:144.98ms step:2566/3200 train_loss:3.3645 train_time:370560ms step_avg:144.98ms step:2567/3200 train_loss:3.3637 train_time:370704ms step_avg:144.98ms step:2568/3200 train_loss:3.4074 train_time:370849ms step_avg:144.98ms step:2569/3200 train_loss:3.5538 train_time:370992ms step_avg:144.98ms step:2570/3200 train_loss:3.4527 train_time:371136ms step_avg:144.97ms step:2571/3200 train_loss:3.5313 train_time:371278ms step_avg:144.97ms step:2572/3200 train_loss:3.2930 train_time:371421ms step_avg:144.97ms step:2573/3200 train_loss:3.3978 train_time:371564ms step_avg:144.97ms step:2574/3200 train_loss:3.0559 train_time:371710ms step_avg:144.97ms step:2575/3200 train_loss:3.3085 train_time:371853ms step_avg:144.97ms step:2576/3200 train_loss:3.2457 train_time:371995ms step_avg:144.97ms step:2577/3200 train_loss:3.3655 train_time:372139ms step_avg:144.97ms step:2578/3200 train_loss:3.4171 train_time:372282ms step_avg:144.97ms step:2579/3200 train_loss:3.3240 train_time:372425ms step_avg:144.97ms step:2580/3200 train_loss:3.3788 train_time:372570ms step_avg:144.97ms step:2581/3200 train_loss:3.3231 train_time:372712ms step_avg:144.97ms step:2582/3200 train_loss:3.4301 train_time:372856ms step_avg:144.97ms step:2583/3200 train_loss:3.3052 train_time:372998ms step_avg:144.97ms step:2584/3200 train_loss:3.5038 train_time:373142ms step_avg:144.97ms step:2585/3200 train_loss:3.4110 train_time:373287ms step_avg:144.97ms step:2586/3200 train_loss:3.4241 train_time:373430ms step_avg:144.97ms step:2587/3200 train_loss:3.5481 train_time:373574ms step_avg:144.96ms step:2588/3200 train_loss:3.4398 train_time:373716ms step_avg:144.96ms step:2589/3200 train_loss:3.2942 train_time:373861ms step_avg:144.96ms step:2590/3200 train_loss:3.4659 train_time:374004ms step_avg:144.96ms step:2591/3200 train_loss:3.3668 train_time:374147ms step_avg:144.96ms step:2592/3200 train_loss:3.5828 train_time:374291ms step_avg:144.96ms step:2593/3200 train_loss:3.4437 train_time:374433ms step_avg:144.96ms step:2594/3200 train_loss:3.2629 train_time:374575ms step_avg:144.96ms step:2595/3200 train_loss:3.3377 train_time:374719ms step_avg:144.96ms step:2596/3200 train_loss:3.7649 train_time:374863ms step_avg:144.96ms step:2597/3200 train_loss:3.4234 train_time:375007ms step_avg:144.96ms step:2598/3200 train_loss:3.4231 train_time:375151ms step_avg:144.96ms step:2599/3200 train_loss:3.2745 train_time:375293ms step_avg:144.96ms step:2600/3200 train_loss:3.5150 train_time:375436ms step_avg:144.96ms step:2601/3200 train_loss:3.6832 train_time:375579ms step_avg:144.96ms step:2602/3200 train_loss:3.2620 train_time:375723ms step_avg:144.95ms step:2603/3200 train_loss:3.4074 train_time:375867ms step_avg:144.95ms step:2604/3200 train_loss:3.2402 train_time:376011ms step_avg:144.95ms step:2605/3200 train_loss:3.5325 train_time:376155ms step_avg:144.95ms step:2606/3200 train_loss:3.4016 train_time:376298ms step_avg:144.95ms step:2607/3200 train_loss:3.2925 train_time:376441ms step_avg:144.95ms step:2608/3200 train_loss:3.2544 train_time:376585ms step_avg:144.95ms step:2609/3200 train_loss:3.3763 train_time:376729ms step_avg:144.95ms step:2610/3200 train_loss:3.5479 train_time:376873ms step_avg:144.95ms step:2611/3200 train_loss:3.4165 train_time:377016ms step_avg:144.95ms step:2612/3200 train_loss:3.2565 train_time:377160ms step_avg:144.95ms step:2613/3200 train_loss:3.3449 train_time:377303ms step_avg:144.95ms step:2614/3200 train_loss:3.4582 train_time:377448ms step_avg:144.95ms step:2615/3200 train_loss:3.3942 train_time:377591ms step_avg:144.95ms step:2616/3200 train_loss:3.3908 train_time:377733ms step_avg:144.95ms step:2617/3200 train_loss:3.4244 train_time:377876ms step_avg:144.95ms step:2618/3200 train_loss:3.4638 train_time:378020ms step_avg:144.95ms step:2619/3200 train_loss:3.3119 train_time:378163ms step_avg:144.95ms step:2620/3200 train_loss:3.4852 train_time:378307ms step_avg:144.95ms step:2621/3200 train_loss:3.4500 train_time:378451ms step_avg:144.94ms step:2622/3200 train_loss:3.5779 train_time:378593ms step_avg:144.94ms step:2623/3200 train_loss:3.4863 train_time:378736ms step_avg:144.94ms step:2624/3200 train_loss:3.4061 train_time:378878ms step_avg:144.94ms step:2625/3200 train_loss:3.3630 train_time:379023ms step_avg:144.94ms step:2625/3200 val_loss:3.3948 train_time:379077ms step_avg:144.96ms step:2626/3200 train_loss:3.3890 train_time:379176ms step_avg:144.94ms step:2627/3200 train_loss:3.4497 train_time:379323ms step_avg:144.95ms step:2628/3200 train_loss:3.2684 train_time:379465ms step_avg:144.94ms step:2629/3200 train_loss:3.5321 train_time:379608ms step_avg:144.94ms step:2630/3200 train_loss:3.4191 train_time:379750ms step_avg:144.94ms step:2631/3200 train_loss:3.4636 train_time:379893ms step_avg:144.94ms step:2632/3200 train_loss:3.6912 train_time:380035ms step_avg:144.94ms step:2633/3200 train_loss:3.4374 train_time:380182ms step_avg:144.94ms step:2634/3200 train_loss:3.3593 train_time:380330ms step_avg:144.94ms step:2635/3200 train_loss:3.3297 train_time:380473ms step_avg:144.94ms step:2636/3200 train_loss:3.3724 train_time:380615ms step_avg:144.94ms step:2637/3200 train_loss:3.1628 train_time:380757ms step_avg:144.94ms step:2638/3200 train_loss:3.4768 train_time:380902ms step_avg:144.94ms step:2639/3200 train_loss:3.4517 train_time:381045ms step_avg:144.94ms step:2640/3200 train_loss:3.3323 train_time:381191ms step_avg:144.94ms step:2641/3200 train_loss:3.4162 train_time:381335ms step_avg:144.94ms step:2642/3200 train_loss:3.4524 train_time:381479ms step_avg:144.94ms step:2643/3200 train_loss:3.2488 train_time:381623ms step_avg:144.94ms step:2644/3200 train_loss:3.3684 train_time:381767ms step_avg:144.94ms step:2645/3200 train_loss:3.4402 train_time:381909ms step_avg:144.94ms step:2646/3200 train_loss:3.4044 train_time:382219ms step_avg:145.00ms step:2647/3200 train_loss:3.2952 train_time:382362ms step_avg:145.00ms step:2648/3200 train_loss:3.5172 train_time:382505ms step_avg:145.00ms step:2649/3200 train_loss:3.7752 train_time:382647ms step_avg:145.00ms step:2650/3200 train_loss:3.4151 train_time:382788ms step_avg:145.00ms step:2651/3200 train_loss:3.3770 train_time:382930ms step_avg:144.99ms step:2652/3200 train_loss:3.5066 train_time:383072ms step_avg:144.99ms step:2653/3200 train_loss:3.3414 train_time:383219ms step_avg:144.99ms step:2654/3200 train_loss:3.3301 train_time:383365ms step_avg:144.99ms step:2655/3200 train_loss:3.4059 train_time:383509ms step_avg:144.99ms step:2656/3200 train_loss:3.3276 train_time:383653ms step_avg:144.99ms step:2657/3200 train_loss:3.3568 train_time:383794ms step_avg:144.99ms step:2658/3200 train_loss:3.3303 train_time:383936ms step_avg:144.99ms step:2659/3200 train_loss:3.4149 train_time:384078ms step_avg:144.99ms step:2660/3200 train_loss:3.5477 train_time:384386ms step_avg:145.05ms step:2661/3200 train_loss:3.3491 train_time:384528ms step_avg:145.05ms step:2662/3200 train_loss:3.4969 train_time:384670ms step_avg:145.05ms step:2663/3200 train_loss:3.3691 train_time:384813ms step_avg:145.05ms step:2664/3200 train_loss:3.3605 train_time:384955ms step_avg:145.05ms step:2665/3200 train_loss:3.2932 train_time:385098ms step_avg:145.05ms step:2666/3200 train_loss:3.3376 train_time:385241ms step_avg:145.05ms step:2667/3200 train_loss:3.3779 train_time:385390ms step_avg:145.05ms step:2668/3200 train_loss:3.4139 train_time:385533ms step_avg:145.05ms step:2669/3200 train_loss:3.3327 train_time:385675ms step_avg:145.05ms step:2670/3200 train_loss:3.3947 train_time:385818ms step_avg:145.04ms step:2671/3200 train_loss:3.2774 train_time:385961ms step_avg:145.04ms step:2672/3200 train_loss:3.3474 train_time:386104ms step_avg:145.04ms step:2673/3200 train_loss:3.3330 train_time:386247ms step_avg:145.04ms step:2674/3200 train_loss:3.3974 train_time:386392ms step_avg:145.04ms step:2675/3200 train_loss:3.4238 train_time:386535ms step_avg:145.04ms step:2676/3200 train_loss:3.3834 train_time:386678ms step_avg:145.04ms step:2677/3200 train_loss:3.3814 train_time:386820ms step_avg:145.04ms step:2678/3200 train_loss:3.4087 train_time:386965ms step_avg:145.04ms step:2679/3200 train_loss:3.4503 train_time:387108ms step_avg:145.04ms step:2680/3200 train_loss:3.3631 train_time:387251ms step_avg:145.04ms step:2681/3200 train_loss:3.2858 train_time:387394ms step_avg:145.04ms step:2682/3200 train_loss:3.3359 train_time:387537ms step_avg:145.04ms step:2683/3200 train_loss:3.8020 train_time:387681ms step_avg:145.04ms step:2684/3200 train_loss:3.3887 train_time:387825ms step_avg:145.04ms step:2685/3200 train_loss:3.4199 train_time:387969ms step_avg:145.04ms step:2686/3200 train_loss:3.4658 train_time:388111ms step_avg:145.03ms step:2687/3200 train_loss:3.3889 train_time:388253ms step_avg:145.03ms step:2688/3200 train_loss:3.4670 train_time:388398ms step_avg:145.03ms step:2689/3200 train_loss:3.3961 train_time:388541ms step_avg:145.03ms step:2690/3200 train_loss:3.3825 train_time:388684ms step_avg:145.03ms step:2691/3200 train_loss:3.4125 train_time:388828ms step_avg:145.03ms step:2692/3200 train_loss:3.4844 train_time:388971ms step_avg:145.03ms step:2693/3200 train_loss:3.2746 train_time:389114ms step_avg:145.03ms step:2694/3200 train_loss:3.6594 train_time:389256ms step_avg:145.03ms step:2695/3200 train_loss:3.4583 train_time:389401ms step_avg:145.03ms step:2696/3200 train_loss:3.2573 train_time:389546ms step_avg:145.03ms step:2697/3200 train_loss:3.4469 train_time:389689ms step_avg:145.03ms step:2698/3200 train_loss:3.4103 train_time:389834ms step_avg:145.03ms step:2699/3200 train_loss:3.3599 train_time:389976ms step_avg:145.03ms step:2700/3200 train_loss:3.4667 train_time:390119ms step_avg:145.03ms step:2701/3200 train_loss:3.4341 train_time:390263ms step_avg:145.03ms step:2702/3200 train_loss:3.3399 train_time:390407ms step_avg:145.02ms step:2703/3200 train_loss:3.3601 train_time:390550ms step_avg:145.02ms step:2704/3200 train_loss:3.3734 train_time:390694ms step_avg:145.02ms step:2705/3200 train_loss:3.3433 train_time:390836ms step_avg:145.02ms step:2706/3200 train_loss:3.5086 train_time:390980ms step_avg:145.02ms step:2707/3200 train_loss:3.4763 train_time:391123ms step_avg:145.02ms step:2708/3200 train_loss:3.3760 train_time:391269ms step_avg:145.02ms step:2709/3200 train_loss:3.3791 train_time:391412ms step_avg:145.02ms step:2710/3200 train_loss:3.4786 train_time:391554ms step_avg:145.02ms step:2711/3200 train_loss:3.3558 train_time:391698ms step_avg:145.02ms step:2712/3200 train_loss:3.4680 train_time:391843ms step_avg:145.02ms step:2713/3200 train_loss:3.2074 train_time:391986ms step_avg:145.02ms step:2714/3200 train_loss:3.4055 train_time:392130ms step_avg:145.02ms step:2715/3200 train_loss:3.2924 train_time:392274ms step_avg:145.02ms step:2716/3200 train_loss:3.3074 train_time:392418ms step_avg:145.02ms step:2717/3200 train_loss:3.4964 train_time:392561ms step_avg:145.02ms step:2718/3200 train_loss:3.3958 train_time:392703ms step_avg:145.02ms step:2719/3200 train_loss:3.6267 train_time:392847ms step_avg:145.02ms step:2720/3200 train_loss:3.3767 train_time:392989ms step_avg:145.01ms step:2721/3200 train_loss:3.3664 train_time:393132ms step_avg:145.01ms step:2722/3200 train_loss:3.5896 train_time:393275ms step_avg:145.01ms step:2723/3200 train_loss:3.3627 train_time:393419ms step_avg:145.01ms step:2724/3200 train_loss:3.5387 train_time:393562ms step_avg:145.01ms step:2725/3200 train_loss:3.4115 train_time:393707ms step_avg:145.01ms step:2726/3200 train_loss:3.3711 train_time:393851ms step_avg:145.01ms step:2727/3200 train_loss:3.3868 train_time:393993ms step_avg:145.01ms step:2728/3200 train_loss:3.7198 train_time:394136ms step_avg:145.01ms step:2729/3200 train_loss:3.4461 train_time:394279ms step_avg:145.01ms step:2730/3200 train_loss:3.3121 train_time:394424ms step_avg:145.01ms step:2731/3200 train_loss:3.4193 train_time:394568ms step_avg:145.01ms step:2732/3200 train_loss:3.3305 train_time:394711ms step_avg:145.01ms step:2733/3200 train_loss:3.2170 train_time:394853ms step_avg:145.01ms step:2734/3200 train_loss:3.3299 train_time:394996ms step_avg:145.01ms step:2735/3200 train_loss:3.4054 train_time:395141ms step_avg:145.01ms step:2736/3200 train_loss:3.2925 train_time:395285ms step_avg:145.01ms step:2737/3200 train_loss:3.6924 train_time:395430ms step_avg:145.01ms step:2738/3200 train_loss:3.4428 train_time:395573ms step_avg:145.00ms step:2739/3200 train_loss:3.6364 train_time:395715ms step_avg:145.00ms step:2740/3200 train_loss:3.3867 train_time:395858ms step_avg:145.00ms step:2741/3200 train_loss:3.3848 train_time:396002ms step_avg:145.00ms step:2742/3200 train_loss:3.3259 train_time:396145ms step_avg:145.00ms step:2743/3200 train_loss:3.3947 train_time:396289ms step_avg:145.00ms step:2744/3200 train_loss:3.4074 train_time:396432ms step_avg:145.00ms step:2745/3200 train_loss:3.5033 train_time:396576ms step_avg:145.00ms step:2746/3200 train_loss:3.2724 train_time:396719ms step_avg:145.00ms step:2747/3200 train_loss:3.3648 train_time:396864ms step_avg:145.00ms step:2748/3200 train_loss:3.4096 train_time:397007ms step_avg:145.00ms step:2749/3200 train_loss:3.5157 train_time:397149ms step_avg:145.00ms step:2750/3200 train_loss:3.3597 train_time:397292ms step_avg:145.00ms step:2750/3200 val_loss:3.3732 train_time:397347ms step_avg:145.02ms step:2751/3200 train_loss:3.4355 train_time:397444ms step_avg:145.00ms step:2752/3200 train_loss:3.4857 train_time:397593ms step_avg:145.00ms step:2753/3200 train_loss:3.3916 train_time:397735ms step_avg:145.00ms step:2754/3200 train_loss:3.3215 train_time:397876ms step_avg:145.00ms step:2755/3200 train_loss:3.3282 train_time:398018ms step_avg:145.00ms step:2756/3200 train_loss:3.4016 train_time:398162ms step_avg:145.00ms step:2757/3200 train_loss:3.3443 train_time:398304ms step_avg:145.00ms step:2758/3200 train_loss:3.2258 train_time:398450ms step_avg:145.00ms step:2759/3200 train_loss:3.6146 train_time:398596ms step_avg:145.00ms step:2760/3200 train_loss:3.4260 train_time:398741ms step_avg:145.00ms step:2761/3200 train_loss:3.3865 train_time:398886ms step_avg:145.00ms step:2762/3200 train_loss:3.3599 train_time:399029ms step_avg:145.00ms step:2763/3200 train_loss:3.2679 train_time:399172ms step_avg:145.00ms step:2764/3200 train_loss:3.4383 train_time:399314ms step_avg:144.99ms step:2765/3200 train_loss:3.3646 train_time:399457ms step_avg:144.99ms step:2766/3200 train_loss:3.2599 train_time:399601ms step_avg:144.99ms step:2767/3200 train_loss:3.3499 train_time:399746ms step_avg:144.99ms step:2768/3200 train_loss:3.4285 train_time:399890ms step_avg:144.99ms step:2769/3200 train_loss:3.3123 train_time:400032ms step_avg:144.99ms step:2770/3200 train_loss:3.3865 train_time:400174ms step_avg:144.99ms step:2771/3200 train_loss:3.3616 train_time:400316ms step_avg:144.99ms step:2772/3200 train_loss:3.7967 train_time:400458ms step_avg:144.99ms step:2773/3200 train_loss:3.2764 train_time:400602ms step_avg:144.99ms step:2774/3200 train_loss:3.4110 train_time:400747ms step_avg:144.99ms step:2775/3200 train_loss:3.4714 train_time:400891ms step_avg:144.99ms step:2776/3200 train_loss:3.4341 train_time:401033ms step_avg:144.99ms step:2777/3200 train_loss:3.5048 train_time:401177ms step_avg:144.99ms step:2778/3200 train_loss:3.5098 train_time:401319ms step_avg:144.99ms step:2779/3200 train_loss:3.3846 train_time:401462ms step_avg:144.98ms step:2780/3200 train_loss:3.2491 train_time:401608ms step_avg:144.98ms step:2781/3200 train_loss:3.3947 train_time:401752ms step_avg:144.98ms step:2782/3200 train_loss:3.4179 train_time:401895ms step_avg:144.98ms step:2783/3200 train_loss:3.2852 train_time:402039ms step_avg:144.98ms step:2784/3200 train_loss:3.3807 train_time:402182ms step_avg:144.98ms step:2785/3200 train_loss:3.4511 train_time:402325ms step_avg:144.98ms step:2786/3200 train_loss:3.3237 train_time:402469ms step_avg:144.98ms step:2787/3200 train_loss:3.4373 train_time:402612ms step_avg:144.98ms step:2788/3200 train_loss:3.4070 train_time:402755ms step_avg:144.98ms step:2789/3200 train_loss:3.3388 train_time:402897ms step_avg:144.98ms step:2790/3200 train_loss:3.4292 train_time:403042ms step_avg:144.98ms step:2791/3200 train_loss:3.3488 train_time:403185ms step_avg:144.98ms step:2792/3200 train_loss:3.2528 train_time:403329ms step_avg:144.98ms step:2793/3200 train_loss:3.3505 train_time:403472ms step_avg:144.98ms step:2794/3200 train_loss:3.3956 train_time:403615ms step_avg:144.98ms step:2795/3200 train_loss:3.3112 train_time:403759ms step_avg:144.98ms step:2796/3200 train_loss:3.3497 train_time:403901ms step_avg:144.98ms step:2797/3200 train_loss:3.2671 train_time:404045ms step_avg:144.98ms step:2798/3200 train_loss:3.3725 train_time:404190ms step_avg:144.97ms step:2799/3200 train_loss:3.3277 train_time:404333ms step_avg:144.97ms step:2800/3200 train_loss:3.4925 train_time:404476ms step_avg:144.97ms step:2801/3200 train_loss:3.4404 train_time:404620ms step_avg:144.97ms step:2802/3200 train_loss:3.4203 train_time:404764ms step_avg:144.97ms step:2803/3200 train_loss:3.3591 train_time:404908ms step_avg:144.97ms step:2804/3200 train_loss:3.5395 train_time:405051ms step_avg:144.97ms step:2805/3200 train_loss:3.5076 train_time:405194ms step_avg:144.97ms step:2806/3200 train_loss:3.2258 train_time:405337ms step_avg:144.97ms step:2807/3200 train_loss:3.6306 train_time:405481ms step_avg:144.97ms step:2808/3200 train_loss:3.3745 train_time:405626ms step_avg:144.97ms step:2809/3200 train_loss:3.3084 train_time:405771ms step_avg:144.97ms step:2810/3200 train_loss:3.3238 train_time:405913ms step_avg:144.97ms step:2811/3200 train_loss:3.4948 train_time:406057ms step_avg:144.97ms step:2812/3200 train_loss:3.4723 train_time:406199ms step_avg:144.97ms step:2813/3200 train_loss:3.2249 train_time:406342ms step_avg:144.97ms step:2814/3200 train_loss:3.4532 train_time:406487ms step_avg:144.97ms step:2815/3200 train_loss:3.5210 train_time:406631ms step_avg:144.97ms step:2816/3200 train_loss:3.3283 train_time:406773ms step_avg:144.97ms step:2817/3200 train_loss:2.9893 train_time:406917ms step_avg:144.96ms step:2818/3200 train_loss:3.3518 train_time:407061ms step_avg:144.96ms step:2819/3200 train_loss:3.3178 train_time:407205ms step_avg:144.96ms step:2820/3200 train_loss:3.5129 train_time:407349ms step_avg:144.96ms step:2821/3200 train_loss:3.3749 train_time:407492ms step_avg:144.96ms step:2822/3200 train_loss:3.4321 train_time:407635ms step_avg:144.96ms step:2823/3200 train_loss:3.3697 train_time:407779ms step_avg:144.96ms step:2824/3200 train_loss:3.3382 train_time:407923ms step_avg:144.96ms step:2825/3200 train_loss:3.2366 train_time:408066ms step_avg:144.96ms step:2826/3200 train_loss:3.4951 train_time:408210ms step_avg:144.96ms step:2827/3200 train_loss:3.3925 train_time:408352ms step_avg:144.96ms step:2828/3200 train_loss:3.2830 train_time:408495ms step_avg:144.96ms step:2829/3200 train_loss:3.4118 train_time:408639ms step_avg:144.96ms step:2830/3200 train_loss:3.4027 train_time:408782ms step_avg:144.96ms step:2831/3200 train_loss:3.3421 train_time:408927ms step_avg:144.96ms step:2832/3200 train_loss:3.4907 train_time:409070ms step_avg:144.96ms step:2833/3200 train_loss:3.4036 train_time:409213ms step_avg:144.96ms step:2834/3200 train_loss:3.3894 train_time:409355ms step_avg:144.96ms step:2835/3200 train_loss:3.2010 train_time:409652ms step_avg:145.01ms step:2836/3200 train_loss:3.4222 train_time:409794ms step_avg:145.01ms step:2837/3200 train_loss:3.3515 train_time:409935ms step_avg:145.01ms step:2838/3200 train_loss:3.6565 train_time:410077ms step_avg:145.01ms step:2839/3200 train_loss:3.3107 train_time:410221ms step_avg:145.01ms step:2840/3200 train_loss:3.3237 train_time:410364ms step_avg:145.00ms step:2841/3200 train_loss:3.3726 train_time:410506ms step_avg:145.00ms step:2842/3200 train_loss:3.3060 train_time:410652ms step_avg:145.00ms step:2843/3200 train_loss:3.3113 train_time:410797ms step_avg:145.00ms step:2844/3200 train_loss:3.4822 train_time:410941ms step_avg:145.00ms step:2845/3200 train_loss:3.3630 train_time:411083ms step_avg:145.00ms step:2846/3200 train_loss:3.3909 train_time:411226ms step_avg:145.00ms step:2847/3200 train_loss:3.3529 train_time:411368ms step_avg:145.00ms step:2848/3200 train_loss:3.6189 train_time:411510ms step_avg:145.00ms step:2849/3200 train_loss:3.2789 train_time:411654ms step_avg:145.00ms step:2850/3200 train_loss:3.3203 train_time:411960ms step_avg:145.06ms step:2851/3200 train_loss:3.4174 train_time:412104ms step_avg:145.06ms step:2852/3200 train_loss:3.3913 train_time:412245ms step_avg:145.05ms step:2853/3200 train_loss:3.3516 train_time:412388ms step_avg:145.05ms step:2854/3200 train_loss:3.4227 train_time:412530ms step_avg:145.05ms step:2855/3200 train_loss:3.2500 train_time:412672ms step_avg:145.05ms step:2856/3200 train_loss:3.2715 train_time:412814ms step_avg:145.05ms step:2857/3200 train_loss:3.3683 train_time:412964ms step_avg:145.05ms step:2858/3200 train_loss:3.3632 train_time:413108ms step_avg:145.05ms step:2859/3200 train_loss:3.2660 train_time:413251ms step_avg:145.05ms step:2860/3200 train_loss:3.3497 train_time:413393ms step_avg:145.05ms step:2861/3200 train_loss:3.3158 train_time:413534ms step_avg:145.05ms step:2862/3200 train_loss:3.3496 train_time:413677ms step_avg:145.05ms step:2863/3200 train_loss:3.3933 train_time:413820ms step_avg:145.05ms step:2864/3200 train_loss:3.6603 train_time:413966ms step_avg:145.05ms step:2865/3200 train_loss:3.4591 train_time:414111ms step_avg:145.05ms step:2866/3200 train_loss:3.3525 train_time:414254ms step_avg:145.05ms step:2867/3200 train_loss:3.2296 train_time:414395ms step_avg:145.05ms step:2868/3200 train_loss:3.4431 train_time:414539ms step_avg:145.05ms step:2869/3200 train_loss:3.4009 train_time:414681ms step_avg:145.04ms step:2870/3200 train_loss:3.3535 train_time:414826ms step_avg:145.04ms step:2871/3200 train_loss:3.4927 train_time:414970ms step_avg:145.04ms step:2872/3200 train_loss:3.2608 train_time:415113ms step_avg:145.04ms step:2873/3200 train_loss:3.3392 train_time:415256ms step_avg:145.04ms step:2874/3200 train_loss:3.2004 train_time:415397ms step_avg:145.04ms step:2875/3200 train_loss:3.3520 train_time:415541ms step_avg:145.04ms step:2875/3200 val_loss:3.3532 train_time:415596ms step_avg:145.06ms step:2876/3200 train_loss:3.2676 train_time:415693ms step_avg:145.04ms step:2877/3200 train_loss:3.2617 train_time:415840ms step_avg:145.04ms step:2878/3200 train_loss:3.3444 train_time:415982ms step_avg:145.04ms step:2879/3200 train_loss:3.4589 train_time:416125ms step_avg:145.04ms step:2880/3200 train_loss:3.4073 train_time:416268ms step_avg:145.04ms step:2881/3200 train_loss:3.3637 train_time:416409ms step_avg:145.04ms step:2882/3200 train_loss:3.3494 train_time:416551ms step_avg:145.04ms step:2883/3200 train_loss:3.4677 train_time:416695ms step_avg:145.04ms step:2884/3200 train_loss:3.2429 train_time:416842ms step_avg:145.04ms step:2885/3200 train_loss:3.2721 train_time:416986ms step_avg:145.04ms step:2886/3200 train_loss:3.3195 train_time:417129ms step_avg:145.04ms step:2887/3200 train_loss:3.3103 train_time:417272ms step_avg:145.04ms step:2888/3200 train_loss:3.3172 train_time:417415ms step_avg:145.04ms step:2889/3200 train_loss:3.3506 train_time:417557ms step_avg:145.04ms step:2890/3200 train_loss:3.5347 train_time:417701ms step_avg:145.04ms step:2891/3200 train_loss:3.3749 train_time:417847ms step_avg:145.04ms step:2892/3200 train_loss:3.2104 train_time:417990ms step_avg:145.03ms step:2893/3200 train_loss:3.1363 train_time:418133ms step_avg:145.03ms step:2894/3200 train_loss:3.2829 train_time:418276ms step_avg:145.03ms step:2895/3200 train_loss:3.1682 train_time:418420ms step_avg:145.03ms step:2896/3200 train_loss:3.3441 train_time:418561ms step_avg:145.03ms step:2897/3200 train_loss:3.4764 train_time:418705ms step_avg:145.03ms step:2898/3200 train_loss:3.2996 train_time:418850ms step_avg:145.03ms step:2899/3200 train_loss:3.3911 train_time:418993ms step_avg:145.03ms step:2900/3200 train_loss:3.2771 train_time:419137ms step_avg:145.03ms step:2901/3200 train_loss:3.4664 train_time:419280ms step_avg:145.03ms step:2902/3200 train_loss:3.4486 train_time:419423ms step_avg:145.03ms step:2903/3200 train_loss:3.4773 train_time:419567ms step_avg:145.03ms step:2904/3200 train_loss:3.2093 train_time:419710ms step_avg:145.03ms step:2905/3200 train_loss:3.3526 train_time:419852ms step_avg:145.03ms step:2906/3200 train_loss:3.3306 train_time:419994ms step_avg:145.03ms step:2907/3200 train_loss:3.3890 train_time:420139ms step_avg:145.03ms step:2908/3200 train_loss:3.3385 train_time:420283ms step_avg:145.03ms step:2909/3200 train_loss:3.3104 train_time:420427ms step_avg:145.02ms step:2910/3200 train_loss:3.6411 train_time:420570ms step_avg:145.02ms step:2911/3200 train_loss:3.3499 train_time:420714ms step_avg:145.02ms step:2912/3200 train_loss:3.2677 train_time:420856ms step_avg:145.02ms step:2913/3200 train_loss:3.2542 train_time:421000ms step_avg:145.02ms step:2914/3200 train_loss:3.7230 train_time:421145ms step_avg:145.02ms step:2915/3200 train_loss:3.3162 train_time:421289ms step_avg:145.02ms step:2916/3200 train_loss:3.2725 train_time:421431ms step_avg:145.02ms step:2917/3200 train_loss:3.2577 train_time:421574ms step_avg:145.02ms step:2918/3200 train_loss:3.5356 train_time:421717ms step_avg:145.02ms step:2919/3200 train_loss:3.0337 train_time:421860ms step_avg:145.02ms step:2920/3200 train_loss:3.2470 train_time:422005ms step_avg:145.02ms step:2921/3200 train_loss:3.2627 train_time:422149ms step_avg:145.02ms step:2922/3200 train_loss:3.3617 train_time:422292ms step_avg:145.02ms step:2923/3200 train_loss:3.3999 train_time:422435ms step_avg:145.02ms step:2924/3200 train_loss:3.4308 train_time:422577ms step_avg:145.02ms step:2925/3200 train_loss:3.4441 train_time:422722ms step_avg:145.02ms step:2926/3200 train_loss:3.3304 train_time:422867ms step_avg:145.02ms step:2927/3200 train_loss:3.3331 train_time:423011ms step_avg:145.02ms step:2928/3200 train_loss:3.3204 train_time:423154ms step_avg:145.01ms step:2929/3200 train_loss:3.3283 train_time:423298ms step_avg:145.01ms step:2930/3200 train_loss:3.2838 train_time:423441ms step_avg:145.01ms step:2931/3200 train_loss:3.3165 train_time:423585ms step_avg:145.01ms step:2932/3200 train_loss:3.4473 train_time:423729ms step_avg:145.01ms step:2933/3200 train_loss:3.4853 train_time:423872ms step_avg:145.01ms step:2934/3200 train_loss:3.4573 train_time:424015ms step_avg:145.01ms step:2935/3200 train_loss:3.2982 train_time:424158ms step_avg:145.01ms step:2936/3200 train_loss:3.3514 train_time:424302ms step_avg:145.01ms step:2937/3200 train_loss:3.2982 train_time:424447ms step_avg:145.01ms step:2938/3200 train_loss:3.3239 train_time:424590ms step_avg:145.01ms step:2939/3200 train_loss:3.3464 train_time:424733ms step_avg:145.01ms step:2940/3200 train_loss:3.3862 train_time:424876ms step_avg:145.01ms step:2941/3200 train_loss:3.4255 train_time:425021ms step_avg:145.01ms step:2942/3200 train_loss:3.4212 train_time:425166ms step_avg:145.01ms step:2943/3200 train_loss:3.3486 train_time:425309ms step_avg:145.01ms step:2944/3200 train_loss:3.2210 train_time:425452ms step_avg:145.01ms step:2945/3200 train_loss:3.1764 train_time:425594ms step_avg:145.01ms step:2946/3200 train_loss:3.3726 train_time:425737ms step_avg:145.01ms step:2947/3200 train_loss:3.4334 train_time:425882ms step_avg:145.01ms step:2948/3200 train_loss:3.3695 train_time:426026ms step_avg:145.01ms step:2949/3200 train_loss:3.5552 train_time:426170ms step_avg:145.01ms step:2950/3200 train_loss:3.3715 train_time:426313ms step_avg:145.00ms step:2951/3200 train_loss:3.3735 train_time:426455ms step_avg:145.00ms step:2952/3200 train_loss:3.7714 train_time:426598ms step_avg:145.00ms step:2953/3200 train_loss:3.4483 train_time:426742ms step_avg:145.00ms step:2954/3200 train_loss:3.3931 train_time:426887ms step_avg:145.00ms step:2955/3200 train_loss:3.4087 train_time:427031ms step_avg:145.00ms step:2956/3200 train_loss:3.3381 train_time:427173ms step_avg:145.00ms step:2957/3200 train_loss:3.3630 train_time:427317ms step_avg:145.00ms step:2958/3200 train_loss:3.2411 train_time:427461ms step_avg:145.00ms step:2959/3200 train_loss:3.3188 train_time:427605ms step_avg:145.00ms step:2960/3200 train_loss:3.4643 train_time:427750ms step_avg:145.00ms step:2961/3200 train_loss:3.2674 train_time:427892ms step_avg:145.00ms step:2962/3200 train_loss:3.3902 train_time:428035ms step_avg:145.00ms step:2963/3200 train_loss:3.2643 train_time:428178ms step_avg:145.00ms step:2964/3200 train_loss:3.3222 train_time:428320ms step_avg:145.00ms step:2965/3200 train_loss:3.3011 train_time:428465ms step_avg:145.00ms step:2966/3200 train_loss:3.4068 train_time:428608ms step_avg:145.00ms step:2967/3200 train_loss:3.2886 train_time:428750ms step_avg:145.00ms step:2968/3200 train_loss:3.5261 train_time:428893ms step_avg:144.99ms step:2969/3200 train_loss:3.3805 train_time:429037ms step_avg:144.99ms step:2970/3200 train_loss:3.3954 train_time:429182ms step_avg:144.99ms step:2971/3200 train_loss:3.3739 train_time:429326ms step_avg:144.99ms step:2972/3200 train_loss:3.4552 train_time:429468ms step_avg:144.99ms step:2973/3200 train_loss:3.2783 train_time:429611ms step_avg:144.99ms step:2974/3200 train_loss:3.2836 train_time:429753ms step_avg:144.99ms step:2975/3200 train_loss:3.2147 train_time:429897ms step_avg:144.99ms step:2976/3200 train_loss:3.2866 train_time:430041ms step_avg:144.99ms step:2977/3200 train_loss:3.2716 train_time:430186ms step_avg:144.99ms step:2978/3200 train_loss:3.2991 train_time:430330ms step_avg:144.99ms step:2979/3200 train_loss:3.5752 train_time:430472ms step_avg:144.99ms step:2980/3200 train_loss:3.3817 train_time:430615ms step_avg:144.99ms step:2981/3200 train_loss:3.4170 train_time:430759ms step_avg:144.99ms step:2982/3200 train_loss:3.4428 train_time:430903ms step_avg:144.99ms step:2983/3200 train_loss:3.5096 train_time:431048ms step_avg:144.99ms step:2984/3200 train_loss:3.3244 train_time:431191ms step_avg:144.99ms step:2985/3200 train_loss:3.4147 train_time:431333ms step_avg:144.99ms step:2986/3200 train_loss:3.4183 train_time:431476ms step_avg:144.99ms step:2987/3200 train_loss:3.3691 train_time:431621ms step_avg:144.99ms step:2988/3200 train_loss:3.4864 train_time:431766ms step_avg:144.99ms step:2989/3200 train_loss:3.0841 train_time:431909ms step_avg:144.98ms step:2990/3200 train_loss:3.4313 train_time:432052ms step_avg:144.98ms step:2991/3200 train_loss:3.3862 train_time:432194ms step_avg:144.98ms step:2992/3200 train_loss:3.3463 train_time:432338ms step_avg:144.98ms step:2993/3200 train_loss:3.2742 train_time:432482ms step_avg:144.98ms step:2994/3200 train_loss:3.4232 train_time:432627ms step_avg:144.98ms step:2995/3200 train_loss:3.2414 train_time:432771ms step_avg:144.98ms step:2996/3200 train_loss:3.2579 train_time:432914ms step_avg:144.98ms step:2997/3200 train_loss:3.3366 train_time:433056ms step_avg:144.98ms step:2998/3200 train_loss:3.2748 train_time:433201ms step_avg:144.98ms step:2999/3200 train_loss:3.3961 train_time:433345ms step_avg:144.98ms step:3000/3200 train_loss:3.2979 train_time:433489ms step_avg:144.98ms step:3000/3200 val_loss:3.3348 train_time:433544ms step_avg:145.00ms step:3001/3200 train_loss:3.2900 train_time:433642ms step_avg:144.98ms step:3002/3200 train_loss:3.2377 train_time:433791ms step_avg:144.98ms step:3003/3200 train_loss:3.2767 train_time:433934ms step_avg:144.98ms step:3004/3200 train_loss:3.4058 train_time:434077ms step_avg:144.98ms step:3005/3200 train_loss:3.7424 train_time:434218ms step_avg:144.98ms step:3006/3200 train_loss:3.3124 train_time:434360ms step_avg:144.98ms step:3007/3200 train_loss:3.3921 train_time:434502ms step_avg:144.98ms step:3008/3200 train_loss:3.1944 train_time:434648ms step_avg:144.98ms step:3009/3200 train_loss:3.4123 train_time:434794ms step_avg:144.98ms step:3010/3200 train_loss:3.3125 train_time:434937ms step_avg:144.98ms step:3011/3200 train_loss:3.3753 train_time:435082ms step_avg:144.98ms step:3012/3200 train_loss:3.3665 train_time:435226ms step_avg:144.98ms step:3013/3200 train_loss:3.2553 train_time:435369ms step_avg:144.98ms step:3014/3200 train_loss:3.4499 train_time:435511ms step_avg:144.98ms step:3015/3200 train_loss:3.4108 train_time:435654ms step_avg:144.98ms step:3016/3200 train_loss:3.2775 train_time:435798ms step_avg:144.98ms step:3017/3200 train_loss:3.3227 train_time:435942ms step_avg:144.98ms step:3018/3200 train_loss:3.3625 train_time:436086ms step_avg:144.98ms step:3019/3200 train_loss:3.4048 train_time:436231ms step_avg:144.98ms step:3020/3200 train_loss:3.1786 train_time:436375ms step_avg:144.97ms step:3021/3200 train_loss:3.4863 train_time:436517ms step_avg:144.97ms step:3022/3200 train_loss:3.3188 train_time:436660ms step_avg:144.97ms step:3023/3200 train_loss:3.2361 train_time:436804ms step_avg:144.97ms step:3024/3200 train_loss:3.3370 train_time:437094ms step_avg:145.02ms step:3025/3200 train_loss:3.3144 train_time:437234ms step_avg:145.02ms step:3026/3200 train_loss:3.3709 train_time:437378ms step_avg:145.02ms step:3027/3200 train_loss:3.3937 train_time:437519ms step_avg:145.02ms step:3028/3200 train_loss:3.2985 train_time:437662ms step_avg:145.02ms step:3029/3200 train_loss:3.1073 train_time:437803ms step_avg:145.02ms step:3030/3200 train_loss:3.4370 train_time:437946ms step_avg:145.02ms step:3031/3200 train_loss:3.2072 train_time:438093ms step_avg:145.02ms step:3032/3200 train_loss:3.1982 train_time:438238ms step_avg:145.02ms step:3033/3200 train_loss:3.5372 train_time:438381ms step_avg:145.02ms step:3034/3200 train_loss:3.5350 train_time:438524ms step_avg:145.01ms step:3035/3200 train_loss:3.3059 train_time:438667ms step_avg:145.01ms step:3036/3200 train_loss:3.3827 train_time:438810ms step_avg:145.01ms step:3037/3200 train_loss:3.3302 train_time:438953ms step_avg:145.01ms step:3038/3200 train_loss:3.2334 train_time:439097ms step_avg:145.01ms step:3039/3200 train_loss:3.2771 train_time:439241ms step_avg:145.01ms step:3040/3200 train_loss:3.3846 train_time:439542ms step_avg:145.06ms step:3041/3200 train_loss:3.3686 train_time:439684ms step_avg:145.06ms step:3042/3200 train_loss:3.1654 train_time:439826ms step_avg:145.06ms step:3043/3200 train_loss:3.3240 train_time:439968ms step_avg:145.06ms step:3044/3200 train_loss:3.3543 train_time:440110ms step_avg:145.06ms step:3045/3200 train_loss:3.3608 train_time:440252ms step_avg:145.06ms step:3046/3200 train_loss:3.4349 train_time:440395ms step_avg:145.06ms step:3047/3200 train_loss:3.2580 train_time:440544ms step_avg:145.06ms step:3048/3200 train_loss:3.3868 train_time:440689ms step_avg:145.06ms step:3049/3200 train_loss:3.3268 train_time:440831ms step_avg:145.06ms step:3050/3200 train_loss:3.2533 train_time:440973ms step_avg:145.06ms step:3051/3200 train_loss:3.3814 train_time:441115ms step_avg:145.06ms step:3052/3200 train_loss:3.2230 train_time:441258ms step_avg:145.06ms step:3053/3200 train_loss:3.4628 train_time:441401ms step_avg:145.05ms step:3054/3200 train_loss:3.4187 train_time:441547ms step_avg:145.05ms step:3055/3200 train_loss:3.3941 train_time:441692ms step_avg:145.05ms step:3056/3200 train_loss:3.3966 train_time:441835ms step_avg:145.05ms step:3057/3200 train_loss:3.2763 train_time:441979ms step_avg:145.05ms step:3058/3200 train_loss:3.3019 train_time:442121ms step_avg:145.05ms step:3059/3200 train_loss:3.3712 train_time:442263ms step_avg:145.05ms step:3060/3200 train_loss:3.2895 train_time:442407ms step_avg:145.05ms step:3061/3200 train_loss:3.3460 train_time:442551ms step_avg:145.05ms step:3062/3200 train_loss:3.3464 train_time:442695ms step_avg:145.05ms step:3063/3200 train_loss:3.2914 train_time:442838ms step_avg:145.05ms step:3064/3200 train_loss:3.2556 train_time:442981ms step_avg:145.05ms step:3065/3200 train_loss:3.2765 train_time:443123ms step_avg:145.05ms step:3066/3200 train_loss:3.2575 train_time:443267ms step_avg:145.05ms step:3067/3200 train_loss:3.2456 train_time:443410ms step_avg:145.05ms step:3068/3200 train_loss:3.2034 train_time:443554ms step_avg:145.05ms step:3069/3200 train_loss:3.2436 train_time:443697ms step_avg:145.05ms step:3070/3200 train_loss:3.2347 train_time:443839ms step_avg:145.05ms step:3071/3200 train_loss:3.4219 train_time:443984ms step_avg:145.05ms step:3072/3200 train_loss:3.3546 train_time:444128ms step_avg:145.04ms step:3073/3200 train_loss:3.3969 train_time:444271ms step_avg:145.04ms step:3074/3200 train_loss:3.3840 train_time:444414ms step_avg:145.04ms step:3075/3200 train_loss:3.3329 train_time:444559ms step_avg:145.04ms step:3076/3200 train_loss:3.3835 train_time:444702ms step_avg:145.04ms step:3077/3200 train_loss:3.4393 train_time:444845ms step_avg:145.04ms step:3078/3200 train_loss:3.2365 train_time:444991ms step_avg:145.04ms step:3079/3200 train_loss:3.7710 train_time:445134ms step_avg:145.04ms step:3080/3200 train_loss:3.3272 train_time:445277ms step_avg:145.04ms step:3081/3200 train_loss:3.2877 train_time:445422ms step_avg:145.04ms step:3082/3200 train_loss:3.4360 train_time:445566ms step_avg:145.04ms step:3083/3200 train_loss:3.2442 train_time:445711ms step_avg:145.04ms step:3084/3200 train_loss:3.2748 train_time:445854ms step_avg:145.04ms step:3085/3200 train_loss:3.3299 train_time:445998ms step_avg:145.04ms step:3086/3200 train_loss:3.4242 train_time:446143ms step_avg:145.04ms step:3087/3200 train_loss:3.3313 train_time:446287ms step_avg:145.04ms step:3088/3200 train_loss:3.2423 train_time:446430ms step_avg:145.04ms step:3089/3200 train_loss:3.3952 train_time:446573ms step_avg:145.04ms step:3090/3200 train_loss:3.2534 train_time:446716ms step_avg:145.04ms step:3091/3200 train_loss:3.5194 train_time:446859ms step_avg:145.04ms step:3092/3200 train_loss:4.0806 train_time:447003ms step_avg:145.04ms step:3093/3200 train_loss:3.3588 train_time:447148ms step_avg:145.04ms step:3094/3200 train_loss:3.2476 train_time:447292ms step_avg:145.04ms step:3095/3200 train_loss:3.1997 train_time:447435ms step_avg:145.04ms step:3096/3200 train_loss:3.3692 train_time:447578ms step_avg:145.03ms step:3097/3200 train_loss:3.5052 train_time:447721ms step_avg:145.03ms step:3098/3200 train_loss:3.2704 train_time:447865ms step_avg:145.03ms step:3099/3200 train_loss:3.3050 train_time:448010ms step_avg:145.03ms step:3100/3200 train_loss:3.4764 train_time:448154ms step_avg:145.03ms step:3101/3200 train_loss:3.3873 train_time:448297ms step_avg:145.03ms step:3102/3200 train_loss:3.3780 train_time:448442ms step_avg:145.03ms step:3103/3200 train_loss:3.2926 train_time:448586ms step_avg:145.03ms step:3104/3200 train_loss:3.5462 train_time:448730ms step_avg:145.03ms step:3105/3200 train_loss:3.3696 train_time:448873ms step_avg:145.03ms step:3106/3200 train_loss:3.2210 train_time:449015ms step_avg:145.03ms step:3107/3200 train_loss:3.2516 train_time:449159ms step_avg:145.03ms step:3108/3200 train_loss:3.2080 train_time:449303ms step_avg:145.03ms step:3109/3200 train_loss:3.4346 train_time:449449ms step_avg:145.03ms step:3110/3200 train_loss:3.3204 train_time:449592ms step_avg:145.03ms step:3111/3200 train_loss:3.3526 train_time:449734ms step_avg:145.03ms step:3112/3200 train_loss:3.3303 train_time:449878ms step_avg:145.03ms step:3113/3200 train_loss:3.3811 train_time:450021ms step_avg:145.03ms step:3114/3200 train_loss:3.3401 train_time:450164ms step_avg:145.03ms step:3115/3200 train_loss:3.3349 train_time:450310ms step_avg:145.03ms step:3116/3200 train_loss:3.3791 train_time:450454ms step_avg:145.03ms step:3117/3200 train_loss:3.2253 train_time:450597ms step_avg:145.03ms step:3118/3200 train_loss:3.2479 train_time:450742ms step_avg:145.03ms step:3119/3200 train_loss:3.4276 train_time:450882ms step_avg:145.02ms step:3120/3200 train_loss:3.4078 train_time:451026ms step_avg:145.02ms step:3121/3200 train_loss:3.1966 train_time:451170ms step_avg:145.02ms step:3122/3200 train_loss:3.3945 train_time:451314ms step_avg:145.02ms step:3123/3200 train_loss:3.4532 train_time:451457ms step_avg:145.02ms step:3124/3200 train_loss:3.4153 train_time:451601ms step_avg:145.02ms step:3125/3200 train_loss:3.2125 train_time:451744ms step_avg:145.02ms step:3125/3200 val_loss:3.3196 train_time:451799ms step_avg:145.04ms step:3126/3200 train_loss:3.2996 train_time:451895ms step_avg:145.02ms step:3127/3200 train_loss:3.3258 train_time:452043ms step_avg:145.03ms step:3128/3200 train_loss:3.4210 train_time:452187ms step_avg:145.02ms step:3129/3200 train_loss:3.4867 train_time:452329ms step_avg:145.02ms step:3130/3200 train_loss:3.1942 train_time:452471ms step_avg:145.02ms step:3131/3200 train_loss:3.3608 train_time:452613ms step_avg:145.02ms step:3132/3200 train_loss:3.3519 train_time:452755ms step_avg:145.02ms step:3133/3200 train_loss:3.3848 train_time:452901ms step_avg:145.02ms step:3134/3200 train_loss:3.2730 train_time:453047ms step_avg:145.02ms step:3135/3200 train_loss:3.3972 train_time:453189ms step_avg:145.02ms step:3136/3200 train_loss:3.3031 train_time:453332ms step_avg:145.02ms step:3137/3200 train_loss:3.3741 train_time:453474ms step_avg:145.02ms step:3138/3200 train_loss:3.5574 train_time:453616ms step_avg:145.02ms step:3139/3200 train_loss:3.5383 train_time:453759ms step_avg:145.02ms step:3140/3200 train_loss:3.3013 train_time:453904ms step_avg:145.02ms step:3141/3200 train_loss:3.3209 train_time:454049ms step_avg:145.02ms step:3142/3200 train_loss:3.2427 train_time:454192ms step_avg:145.02ms step:3143/3200 train_loss:3.3360 train_time:454335ms step_avg:145.02ms step:3144/3200 train_loss:3.1349 train_time:454477ms step_avg:145.01ms step:3145/3200 train_loss:3.3773 train_time:454620ms step_avg:145.01ms step:3146/3200 train_loss:3.2885 train_time:454764ms step_avg:145.01ms step:3147/3200 train_loss:3.3085 train_time:454909ms step_avg:145.01ms step:3148/3200 train_loss:3.4816 train_time:455053ms step_avg:145.01ms step:3149/3200 train_loss:3.5602 train_time:455196ms step_avg:145.01ms step:3150/3200 train_loss:3.4379 train_time:455341ms step_avg:145.01ms step:3151/3200 train_loss:3.2472 train_time:455486ms step_avg:145.01ms step:3152/3200 train_loss:3.2982 train_time:455628ms step_avg:145.01ms step:3153/3200 train_loss:3.2701 train_time:455771ms step_avg:145.01ms step:3154/3200 train_loss:3.3936 train_time:455914ms step_avg:145.01ms step:3155/3200 train_loss:3.2060 train_time:456058ms step_avg:145.01ms step:3156/3200 train_loss:3.3402 train_time:456201ms step_avg:145.01ms step:3157/3200 train_loss:3.2858 train_time:456344ms step_avg:145.01ms step:3158/3200 train_loss:3.4147 train_time:456488ms step_avg:145.01ms step:3159/3200 train_loss:3.4694 train_time:456631ms step_avg:145.01ms step:3160/3200 train_loss:3.3178 train_time:456773ms step_avg:145.01ms step:3161/3200 train_loss:3.3779 train_time:456916ms step_avg:145.01ms step:3162/3200 train_loss:3.4649 train_time:457059ms step_avg:145.01ms step:3163/3200 train_loss:3.3616 train_time:457204ms step_avg:145.01ms step:3164/3200 train_loss:3.4145 train_time:457349ms step_avg:145.01ms step:3165/3200 train_loss:3.2426 train_time:457492ms step_avg:145.01ms step:3166/3200 train_loss:3.2207 train_time:457636ms step_avg:145.00ms step:3167/3200 train_loss:3.2642 train_time:457778ms step_avg:145.00ms step:3168/3200 train_loss:3.0829 train_time:457920ms step_avg:145.00ms step:3169/3200 train_loss:3.2596 train_time:458067ms step_avg:145.00ms step:3170/3200 train_loss:3.3940 train_time:458210ms step_avg:145.00ms step:3171/3200 train_loss:3.4096 train_time:458353ms step_avg:145.00ms step:3172/3200 train_loss:3.3842 train_time:458496ms step_avg:145.00ms step:3173/3200 train_loss:3.3623 train_time:458640ms step_avg:145.00ms step:3174/3200 train_loss:3.3312 train_time:458784ms step_avg:145.00ms step:3175/3200 train_loss:3.3291 train_time:458927ms step_avg:145.00ms step:3176/3200 train_loss:3.3143 train_time:459071ms step_avg:145.00ms step:3177/3200 train_loss:3.2573 train_time:459214ms step_avg:145.00ms step:3178/3200 train_loss:3.3802 train_time:459358ms step_avg:145.00ms step:3179/3200 train_loss:3.4739 train_time:459503ms step_avg:145.00ms step:3180/3200 train_loss:3.3007 train_time:459646ms step_avg:145.00ms step:3181/3200 train_loss:3.2935 train_time:459790ms step_avg:145.00ms step:3182/3200 train_loss:3.3380 train_time:459933ms step_avg:145.00ms step:3183/3200 train_loss:3.4358 train_time:460075ms step_avg:145.00ms step:3184/3200 train_loss:3.4529 train_time:460219ms step_avg:145.00ms step:3185/3200 train_loss:3.3497 train_time:460363ms step_avg:145.00ms step:3186/3200 train_loss:3.4251 train_time:460508ms step_avg:145.00ms step:3187/3200 train_loss:3.4073 train_time:460651ms step_avg:145.00ms step:3188/3200 train_loss:3.2040 train_time:460794ms step_avg:144.99ms step:3189/3200 train_loss:3.2952 train_time:460937ms step_avg:144.99ms step:3190/3200 train_loss:3.3177 train_time:461081ms step_avg:144.99ms step:3191/3200 train_loss:3.3406 train_time:461224ms step_avg:144.99ms step:3192/3200 train_loss:3.3048 train_time:461368ms step_avg:144.99ms step:3193/3200 train_loss:3.2335 train_time:461510ms step_avg:144.99ms step:3194/3200 train_loss:4.2403 train_time:461654ms step_avg:144.99ms step:3195/3200 train_loss:3.3444 train_time:461796ms step_avg:144.99ms step:3196/3200 train_loss:3.1480 train_time:461941ms step_avg:144.99ms step:3197/3200 train_loss:3.2850 train_time:462086ms step_avg:144.99ms step:3198/3200 train_loss:3.1608 train_time:462230ms step_avg:144.99ms step:3199/3200 train_loss:3.2863 train_time:462374ms step_avg:144.99ms step:3200/3200 train_loss:3.2243 train_time:462516ms step_avg:144.99ms step:3200/3200 val_loss:3.3151 train_time:462573ms step_avg:145.01ms