stderr.log

+ deepspeed --master_port 52199 --module safe_rlhf.finetune --train_datasets inverse-json::/home/hansirui_1st/jiayi/resist/imdb_data/train/pos/5000/train.json --model_name_or_path /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b --max_length 512 --trust_remote_code True --epochs 1 --per_device_train_batch_size 1 --per_device_eval_batch_size 4 --gradient_accumulation_steps 8 --gradient_checkpointing --learning_rate 1e-5 --lr_warmup_ratio 0 --weight_decay 0.0 --lr_scheduler_type constant --weight_decay 0.0 --seed 42 --output_dir /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-0.5T/tinyllama-0.5T-s3-Q1-5000 --log_type wandb --log_run_name imdb-tinyllama-0.5T-s3-Q1-5000 --log_project Inverse_Alignment_IMDb --zero_stage 3 --offload none --bf16 True --tf32 True --save_16bit
[rank4]:[W529 16:22:59.685402862 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 4]  using GPU 4 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank2]:[W529 16:22:59.723000259 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 2]  using GPU 2 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank1]:[W529 16:22:59.041239071 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 1]  using GPU 1 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank6]:[W529 16:22:59.044379032 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 6]  using GPU 6 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank5]:[W529 16:22:59.128345993 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 5]  using GPU 5 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank3]:[W529 16:23:00.226011042 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 3]  using GPU 3 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank0]:[W529 16:23:00.464226689 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 0]  using GPU 0 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
[rank7]:[W529 16:23:00.968501999 ProcessGroupNCCL.cpp:4561] [PG ID 0 PG GUID 0 Rank 7]  using GPU 7 to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect. Specify device_ids in barrier() to force use of a particular device, or call init_process_group() with a device_id.
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/config.json
Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

Model config LlamaConfig {
  "architectures": [
    "LlamaForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 1,
  "eos_token_id": 2,
  "head_dim": 64,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "initializer_range": 0.02,
  "intermediate_size": 5632,
  "max_position_embeddings": 2048,
  "mlp_bias": false,
  "model_type": "llama",
  "num_attention_heads": 32,
  "num_hidden_layers": 22,
  "num_key_value_heads": 4,
  "pretraining_tp": 1,
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 10000.0,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.52.1",
  "use_cache": true,
  "vocab_size": 32000
}

loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Will use torch_dtype=torch.float32 as defined in model's config object
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Instantiating LlamaForCausalLM model under default dtype torch.float32.
loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

loading weights file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/model.safetensors
Will use torch_dtype=torch.float32 as defined in model's config object
Instantiating LlamaForCausalLM model under default dtype torch.float32.
Detected DeepSpeed ZeRO-3: activating zero.init() for this model
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2
}

All model checkpoint weights were used when initializing LlamaForCausalLM.

All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
All model checkpoint weights were used when initializing LlamaForCausalLM.

All model checkpoint weights were used when initializing LlamaForCausalLM.

All model checkpoint weights were used when initializing LlamaForCausalLM.

All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
All model checkpoint weights were used when initializing LlamaForCausalLM.

All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
All model checkpoint weights were used when initializing LlamaForCausalLM.

All model checkpoint weights were used when initializing LlamaForCausalLM.

All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

loading file tokenizer.model
loading file tokenizer.json
loading file added_tokens.json
loading file special_tokens_map.json
loading file tokenizer_config.json
loading file chat_template.jinja
loading file tokenizer.model
loading file tokenizer.model
loading file tokenizer.json
loading file tokenizer.json
loading file added_tokens.json
loading file added_tokens.json
loading file special_tokens_map.json
loading file tokenizer_config.json
loading file special_tokens_map.json
loading file chat_template.jinja
loading file tokenizer_config.json
loading file chat_template.jinja
loading file tokenizer.model
loading file tokenizer.model
loading file tokenizer.json
loading file tokenizer.json
loading file added_tokens.json
loading file added_tokens.json
loading file special_tokens_map.json
loading file special_tokens_map.json
loading file tokenizer_config.json
loading file tokenizer_config.json
loading file chat_template.jinja
loading file chat_template.jinja
loading file tokenizer.model
loading file tokenizer.json
loading file added_tokens.json
loading file special_tokens_map.json
loading file tokenizer_config.json
loading file chat_template.jinja
loading file tokenizer.model
loading file tokenizer.json
loading file added_tokens.json
loading file special_tokens_map.json
loading file tokenizer_config.json
loading file chat_template.jinja
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
All model checkpoint weights were used when initializing LlamaForCausalLM.

All the weights of LlamaForCausalLM were initialized from the model checkpoint at /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b.
If your task is similar to the task the model of the checkpoint was trained on, you can already use LlamaForCausalLM for predictions without further training.
loading configuration file /aifs4su/hansirui_1st/models/TinyLlama-1.1B-intermediate-step-240k-503b/generation_config.json
Generate config GenerationConfig {
  "bos_token_id": 1,
  "eos_token_id": 2,
  "max_length": 2048,
  "pad_token_id": 0
}

loading file tokenizer.model
loading file tokenizer.json
loading file added_tokens.json
loading file special_tokens_map.json
loading file tokenizer_config.json
loading file chat_template.jinja
You are resizing the embedding layer without providing a `pad_to_multiple_of` parameter. This means that the new embedding dimension will be 32001. This might induce some performance reduction as *Tensor Cores* will not be available. For more details about this, or help on choosing the correct value for resizing, refer to this guide: https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new embeddings will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
The new lm_head weights will be initialized from a multivariate normal distribution that has old embeddings' mean and covariance. As described in this article: https://nlp.stanford.edu/~johnhew/vocab-expansion.html. To disable this, use `mean_resizing=False`
Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...
Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...
Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...
Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...
Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...
Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...

Using /home/hansirui_1st/.cache/torch_extensions/py311_cu124 as PyTorch extensions root...
Detected CUDA files, patching ldflags
Emitting ninja build file /home/hansirui_1st/.cache/torch_extensions/py311_cu124/fused_adam/build.ninja...
/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/torch/utils/cpp_extension.py:2059: UserWarning: TORCH_CUDA_ARCH_LIST is not set, all archs for visible cards are included for compilation. 
If this is not desired, please set os.environ['TORCH_CUDA_ARCH_LIST'].
  warnings.warn(
Building extension module fused_adam...
Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
Loading extension module fused_adam...
Loading extension module fused_adam...
Loading extension module fused_adam...
Loading extension module fused_adam...
Loading extension module fused_adam...
Loading extension module fused_adam...
Loading extension module fused_adam...
Loading extension module fused_adam...
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.
wandb: Currently logged in as: xtom to https://api.wandb.ai. Use `wandb login --relogin` to force relogin
wandb: Tracking run with wandb version 0.19.11
wandb: Run data is saved locally in /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-0.5T/tinyllama-0.5T-s3-Q1-5000/wandb/run-20250529_162336-hgyfe4xf
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run imdb-tinyllama-0.5T-s3-Q1-5000
wandb: ⭐️ View project at https://wandb.ai/xtom/Inverse_Alignment_IMDb
wandb: 🚀 View run at https://wandb.ai/xtom/Inverse_Alignment_IMDb/runs/hgyfe4xf

Training 1/1 epoch:   0%|          | 0/625 [00:00<?, ?it/s]`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`.

Training 1/1 epoch (loss 4.4668):   0%|          | 0/625 [00:05<?, ?it/s]
Training 1/1 epoch (loss 4.4668):   0%|          | 1/625 [00:05<1:00:05,  5.78s/it]
Training 1/1 epoch (loss 4.6779):   0%|          | 1/625 [00:07<1:00:05,  5.78s/it]
Training 1/1 epoch (loss 4.6779):   0%|          | 2/625 [00:07<34:37,  3.33s/it]  
Training 1/1 epoch (loss 4.5013):   0%|          | 2/625 [00:07<34:37,  3.33s/it]
Training 1/1 epoch (loss 4.5013):   0%|          | 3/625 [00:07<20:39,  1.99s/it]
Training 1/1 epoch (loss 4.5720):   0%|          | 3/625 [00:08<20:39,  1.99s/it]
Training 1/1 epoch (loss 4.5720):   1%|          | 4/625 [00:08<13:57,  1.35s/it]
Training 1/1 epoch (loss 4.7567):   1%|          | 4/625 [00:08<13:57,  1.35s/it]
Training 1/1 epoch (loss 4.7567):   1%|          | 5/625 [00:08<10:17,  1.00it/s]
Training 1/1 epoch (loss 4.7881):   1%|          | 5/625 [00:08<10:17,  1.00it/s]
Training 1/1 epoch (loss 4.7881):   1%|          | 6/625 [00:08<07:54,  1.31it/s]
Training 1/1 epoch (loss 4.7019):   1%|          | 6/625 [00:09<07:54,  1.31it/s]
Training 1/1 epoch (loss 4.7019):   1%|          | 7/625 [00:09<06:26,  1.60it/s]
Training 1/1 epoch (loss 4.8541):   1%|          | 7/625 [00:09<06:26,  1.60it/s]
Training 1/1 epoch (loss 4.8541):   1%|▏         | 8/625 [00:09<05:51,  1.76it/s]
Training 1/1 epoch (loss 4.4458):   1%|▏         | 8/625 [00:09<05:51,  1.76it/s]
Training 1/1 epoch (loss 4.4458):   1%|▏         | 9/625 [00:09<05:09,  1.99it/s]
Training 1/1 epoch (loss 4.7437):   1%|▏         | 9/625 [00:10<05:09,  1.99it/s]
Training 1/1 epoch (loss 4.7437):   2%|▏         | 10/625 [00:10<04:43,  2.17it/s]
Training 1/1 epoch (loss 4.4261):   2%|▏         | 10/625 [00:10<04:43,  2.17it/s]
Training 1/1 epoch (loss 4.4261):   2%|▏         | 11/625 [00:10<04:17,  2.38it/s]
Training 1/1 epoch (loss 4.5333):   2%|▏         | 11/625 [00:11<04:17,  2.38it/s]
Training 1/1 epoch (loss 4.5333):   2%|▏         | 12/625 [00:11<04:01,  2.54it/s]
Training 1/1 epoch (loss 4.5288):   2%|▏         | 12/625 [00:11<04:01,  2.54it/s]
Training 1/1 epoch (loss 4.5288):   2%|▏         | 13/625 [00:11<03:43,  2.74it/s]
Training 1/1 epoch (loss 4.9279):   2%|▏         | 13/625 [00:11<03:43,  2.74it/s]
Training 1/1 epoch (loss 4.9279):   2%|▏         | 14/625 [00:11<03:37,  2.81it/s]
Training 1/1 epoch (loss 4.4261):   2%|▏         | 14/625 [00:11<03:37,  2.81it/s]
Training 1/1 epoch (loss 4.4261):   2%|▏         | 15/625 [00:11<03:30,  2.89it/s]
Training 1/1 epoch (loss 4.7965):   2%|▏         | 15/625 [00:12<03:30,  2.89it/s]
Training 1/1 epoch (loss 4.7965):   3%|▎         | 16/625 [00:12<03:39,  2.77it/s]
Training 1/1 epoch (loss 4.4697):   3%|▎         | 16/625 [00:12<03:39,  2.77it/s]
Training 1/1 epoch (loss 4.4697):   3%|▎         | 17/625 [00:12<03:36,  2.81it/s]
Training 1/1 epoch (loss 4.3186):   3%|▎         | 17/625 [00:13<03:36,  2.81it/s]
Training 1/1 epoch (loss 4.3186):   3%|▎         | 18/625 [00:13<03:35,  2.81it/s]
Training 1/1 epoch (loss 4.6209):   3%|▎         | 18/625 [00:13<03:35,  2.81it/s]
Training 1/1 epoch (loss 4.6209):   3%|▎         | 19/625 [00:13<03:29,  2.89it/s]
Training 1/1 epoch (loss 4.6410):   3%|▎         | 19/625 [00:13<03:29,  2.89it/s]
Training 1/1 epoch (loss 4.6410):   3%|▎         | 20/625 [00:13<03:22,  2.99it/s]
Training 1/1 epoch (loss 4.3288):   3%|▎         | 20/625 [00:14<03:22,  2.99it/s]
Training 1/1 epoch (loss 4.3288):   3%|▎         | 21/625 [00:14<03:21,  3.00it/s]
Training 1/1 epoch (loss 4.4718):   3%|▎         | 21/625 [00:14<03:21,  3.00it/s]
Training 1/1 epoch (loss 4.4718):   4%|▎         | 22/625 [00:14<03:24,  2.95it/s]
Training 1/1 epoch (loss 4.4318):   4%|▎         | 22/625 [00:14<03:24,  2.95it/s]
Training 1/1 epoch (loss 4.4318):   4%|▎         | 23/625 [00:14<03:24,  2.94it/s]
Training 1/1 epoch (loss 4.2327):   4%|▎         | 23/625 [00:15<03:24,  2.94it/s]
Training 1/1 epoch (loss 4.2327):   4%|▍         | 24/625 [00:15<03:30,  2.85it/s]
Training 1/1 epoch (loss 4.0163):   4%|▍         | 24/625 [00:15<03:30,  2.85it/s]
Training 1/1 epoch (loss 4.0163):   4%|▍         | 25/625 [00:15<03:33,  2.81it/s]
Training 1/1 epoch (loss 4.5317):   4%|▍         | 25/625 [00:15<03:33,  2.81it/s]
Training 1/1 epoch (loss 4.5317):   4%|▍         | 26/625 [00:15<03:26,  2.90it/s]
Training 1/1 epoch (loss 3.7876):   4%|▍         | 26/625 [00:16<03:26,  2.90it/s]
Training 1/1 epoch (loss 3.7876):   4%|▍         | 27/625 [00:16<03:28,  2.87it/s]
Training 1/1 epoch (loss 4.4582):   4%|▍         | 27/625 [00:16<03:28,  2.87it/s]
Training 1/1 epoch (loss 4.4582):   4%|▍         | 28/625 [00:16<03:36,  2.76it/s]
Training 1/1 epoch (loss 4.1466):   4%|▍         | 28/625 [00:16<03:36,  2.76it/s]
Training 1/1 epoch (loss 4.1466):   5%|▍         | 29/625 [00:16<03:28,  2.85it/s]
Training 1/1 epoch (loss 4.3357):   5%|▍         | 29/625 [00:17<03:28,  2.85it/s]
Training 1/1 epoch (loss 4.3357):   5%|▍         | 30/625 [00:17<03:23,  2.92it/s]
Training 1/1 epoch (loss 4.0999):   5%|▍         | 30/625 [00:17<03:23,  2.92it/s]
Training 1/1 epoch (loss 4.0999):   5%|▍         | 31/625 [00:17<03:20,  2.96it/s]
Training 1/1 epoch (loss 4.2150):   5%|▍         | 31/625 [00:17<03:20,  2.96it/s]
Training 1/1 epoch (loss 4.2150):   5%|▌         | 32/625 [00:17<03:30,  2.81it/s]
Training 1/1 epoch (loss 3.9162):   5%|▌         | 32/625 [00:18<03:30,  2.81it/s]
Training 1/1 epoch (loss 3.9162):   5%|▌         | 33/625 [00:18<03:54,  2.52it/s]
Training 1/1 epoch (loss 3.7561):   5%|▌         | 33/625 [00:18<03:54,  2.52it/s]
Training 1/1 epoch (loss 3.7561):   5%|▌         | 34/625 [00:18<03:47,  2.59it/s]
Training 1/1 epoch (loss 3.9071):   5%|▌         | 34/625 [00:19<03:47,  2.59it/s]
Training 1/1 epoch (loss 3.9071):   6%|▌         | 35/625 [00:19<03:36,  2.72it/s]
Training 1/1 epoch (loss 3.7704):   6%|▌         | 35/625 [00:19<03:36,  2.72it/s]
Training 1/1 epoch (loss 3.7704):   6%|▌         | 36/625 [00:19<03:25,  2.86it/s]
Training 1/1 epoch (loss 4.2088):   6%|▌         | 36/625 [00:19<03:25,  2.86it/s]
Training 1/1 epoch (loss 4.2088):   6%|▌         | 37/625 [00:19<03:24,  2.88it/s]
Training 1/1 epoch (loss 3.7491):   6%|▌         | 37/625 [00:20<03:24,  2.88it/s]
Training 1/1 epoch (loss 3.7491):   6%|▌         | 38/625 [00:20<03:15,  3.01it/s]
Training 1/1 epoch (loss 4.1078):   6%|▌         | 38/625 [00:20<03:15,  3.01it/s]
Training 1/1 epoch (loss 4.1078):   6%|▌         | 39/625 [00:20<03:22,  2.89it/s]
Training 1/1 epoch (loss 3.8806):   6%|▌         | 39/625 [00:20<03:22,  2.89it/s]
Training 1/1 epoch (loss 3.8806):   6%|▋         | 40/625 [00:20<03:28,  2.81it/s]
Training 1/1 epoch (loss 3.9669):   6%|▋         | 40/625 [00:21<03:28,  2.81it/s]
Training 1/1 epoch (loss 3.9669):   7%|▋         | 41/625 [00:21<03:26,  2.83it/s]
Training 1/1 epoch (loss 3.9035):   7%|▋         | 41/625 [00:21<03:26,  2.83it/s]
Training 1/1 epoch (loss 3.9035):   7%|▋         | 42/625 [00:21<03:17,  2.96it/s]
Training 1/1 epoch (loss 3.9459):   7%|▋         | 42/625 [00:21<03:17,  2.96it/s]
Training 1/1 epoch (loss 3.9459):   7%|▋         | 43/625 [00:21<03:13,  3.01it/s]
Training 1/1 epoch (loss 3.8424):   7%|▋         | 43/625 [00:22<03:13,  3.01it/s]
Training 1/1 epoch (loss 3.8424):   7%|▋         | 44/625 [00:22<03:14,  2.99it/s]
Training 1/1 epoch (loss 4.0493):   7%|▋         | 44/625 [00:22<03:14,  2.99it/s]
Training 1/1 epoch (loss 4.0493):   7%|▋         | 45/625 [00:22<03:17,  2.94it/s]
Training 1/1 epoch (loss 3.7841):   7%|▋         | 45/625 [00:22<03:17,  2.94it/s]
Training 1/1 epoch (loss 3.7841):   7%|▋         | 46/625 [00:22<03:15,  2.96it/s]
Training 1/1 epoch (loss 3.8318):   7%|▋         | 46/625 [00:23<03:15,  2.96it/s]
Training 1/1 epoch (loss 3.8318):   8%|▊         | 47/625 [00:23<03:17,  2.92it/s]
Training 1/1 epoch (loss 3.7664):   8%|▊         | 47/625 [00:23<03:17,  2.92it/s]
Training 1/1 epoch (loss 3.7664):   8%|▊         | 48/625 [00:23<04:03,  2.37it/s]
Training 1/1 epoch (loss 3.9139):   8%|▊         | 48/625 [00:24<04:03,  2.37it/s]
Training 1/1 epoch (loss 3.9139):   8%|▊         | 49/625 [00:24<04:38,  2.07it/s]
Training 1/1 epoch (loss 4.1815):   8%|▊         | 49/625 [00:24<04:38,  2.07it/s]
Training 1/1 epoch (loss 4.1815):   8%|▊         | 50/625 [00:24<04:07,  2.33it/s]
Training 1/1 epoch (loss 3.8383):   8%|▊         | 50/625 [00:25<04:07,  2.33it/s]
Training 1/1 epoch (loss 3.8383):   8%|▊         | 51/625 [00:25<03:48,  2.51it/s]
Training 1/1 epoch (loss 3.7954):   8%|▊         | 51/625 [00:25<03:48,  2.51it/s]
Training 1/1 epoch (loss 3.7954):   8%|▊         | 52/625 [00:25<03:45,  2.54it/s]
Training 1/1 epoch (loss 3.7260):   8%|▊         | 52/625 [00:25<03:45,  2.54it/s]
Training 1/1 epoch (loss 3.7260):   8%|▊         | 53/625 [00:25<03:38,  2.62it/s]
Training 1/1 epoch (loss 3.6432):   8%|▊         | 53/625 [00:26<03:38,  2.62it/s]
Training 1/1 epoch (loss 3.6432):   9%|▊         | 54/625 [00:26<03:29,  2.72it/s]
Training 1/1 epoch (loss 3.9700):   9%|▊         | 54/625 [00:26<03:29,  2.72it/s]
Training 1/1 epoch (loss 3.9700):   9%|▉         | 55/625 [00:26<03:33,  2.67it/s]
Training 1/1 epoch (loss 3.9824):   9%|▉         | 55/625 [00:26<03:33,  2.67it/s]
Training 1/1 epoch (loss 3.9824):   9%|▉         | 56/625 [00:26<03:34,  2.66it/s]
Training 1/1 epoch (loss 3.5208):   9%|▉         | 56/625 [00:27<03:34,  2.66it/s]
Training 1/1 epoch (loss 3.5208):   9%|▉         | 57/625 [00:27<03:27,  2.74it/s]
Training 1/1 epoch (loss 3.4617):   9%|▉         | 57/625 [00:27<03:27,  2.74it/s]
Training 1/1 epoch (loss 3.4617):   9%|▉         | 58/625 [00:27<03:18,  2.86it/s]
Training 1/1 epoch (loss 3.4368):   9%|▉         | 58/625 [00:27<03:18,  2.86it/s]
Training 1/1 epoch (loss 3.4368):   9%|▉         | 59/625 [00:27<03:17,  2.87it/s]
Training 1/1 epoch (loss 3.3069):   9%|▉         | 59/625 [00:28<03:17,  2.87it/s]
Training 1/1 epoch (loss 3.3069):  10%|▉         | 60/625 [00:28<03:22,  2.79it/s]
Training 1/1 epoch (loss 3.4806):  10%|▉         | 60/625 [00:28<03:22,  2.79it/s]
Training 1/1 epoch (loss 3.4806):  10%|▉         | 61/625 [00:28<03:24,  2.76it/s]
Training 1/1 epoch (loss 3.5428):  10%|▉         | 61/625 [00:28<03:24,  2.76it/s]
Training 1/1 epoch (loss 3.5428):  10%|▉         | 62/625 [00:28<03:26,  2.73it/s]
Training 1/1 epoch (loss 3.4872):  10%|▉         | 62/625 [00:29<03:26,  2.73it/s]
Training 1/1 epoch (loss 3.4872):  10%|█         | 63/625 [00:29<03:22,  2.78it/s]
Training 1/1 epoch (loss 3.4505):  10%|█         | 63/625 [00:29<03:22,  2.78it/s]
Training 1/1 epoch (loss 3.4505):  10%|█         | 64/625 [00:29<03:20,  2.80it/s]
Training 1/1 epoch (loss 3.2411):  10%|█         | 64/625 [00:30<03:20,  2.80it/s]
Training 1/1 epoch (loss 3.2411):  10%|█         | 65/625 [00:30<03:19,  2.81it/s]
Training 1/1 epoch (loss 3.5053):  10%|█         | 65/625 [00:30<03:19,  2.81it/s]
Training 1/1 epoch (loss 3.5053):  11%|█         | 66/625 [00:30<03:24,  2.74it/s]
Training 1/1 epoch (loss 3.4040):  11%|█         | 66/625 [00:30<03:24,  2.74it/s]
Training 1/1 epoch (loss 3.4040):  11%|█         | 67/625 [00:30<03:17,  2.83it/s]
Training 1/1 epoch (loss 3.1889):  11%|█         | 67/625 [00:31<03:17,  2.83it/s]
Training 1/1 epoch (loss 3.1889):  11%|█         | 68/625 [00:31<03:11,  2.91it/s]
Training 1/1 epoch (loss 3.4726):  11%|█         | 68/625 [00:31<03:11,  2.91it/s]
Training 1/1 epoch (loss 3.4726):  11%|█         | 69/625 [00:31<03:10,  2.93it/s]
Training 1/1 epoch (loss 3.2033):  11%|█         | 69/625 [00:31<03:10,  2.93it/s]
Training 1/1 epoch (loss 3.2033):  11%|█         | 70/625 [00:31<03:02,  3.03it/s]
Training 1/1 epoch (loss 3.3759):  11%|█         | 70/625 [00:32<03:02,  3.03it/s]
Training 1/1 epoch (loss 3.3759):  11%|█▏        | 71/625 [00:32<02:58,  3.10it/s]
Training 1/1 epoch (loss 3.2108):  11%|█▏        | 71/625 [00:32<02:58,  3.10it/s]
Training 1/1 epoch (loss 3.2108):  12%|█▏        | 72/625 [00:32<03:06,  2.97it/s]
Training 1/1 epoch (loss 3.2089):  12%|█▏        | 72/625 [00:32<03:06,  2.97it/s]
Training 1/1 epoch (loss 3.2089):  12%|█▏        | 73/625 [00:32<03:07,  2.95it/s]
Training 1/1 epoch (loss 3.5316):  12%|█▏        | 73/625 [00:33<03:07,  2.95it/s]
Training 1/1 epoch (loss 3.5316):  12%|█▏        | 74/625 [00:33<03:10,  2.89it/s]
Training 1/1 epoch (loss 3.0627):  12%|█▏        | 74/625 [00:33<03:10,  2.89it/s]
Training 1/1 epoch (loss 3.0627):  12%|█▏        | 75/625 [00:33<03:04,  2.98it/s]
Training 1/1 epoch (loss 3.1479):  12%|█▏        | 75/625 [00:33<03:04,  2.98it/s]
Training 1/1 epoch (loss 3.1479):  12%|█▏        | 76/625 [00:33<02:59,  3.06it/s]
Training 1/1 epoch (loss 3.2397):  12%|█▏        | 76/625 [00:34<02:59,  3.06it/s]
Training 1/1 epoch (loss 3.2397):  12%|█▏        | 77/625 [00:34<02:57,  3.08it/s]
Training 1/1 epoch (loss 3.2617):  12%|█▏        | 77/625 [00:34<02:57,  3.08it/s]
Training 1/1 epoch (loss 3.2617):  12%|█▏        | 78/625 [00:34<03:16,  2.78it/s]
Training 1/1 epoch (loss 3.4527):  12%|█▏        | 78/625 [00:34<03:16,  2.78it/s]
Training 1/1 epoch (loss 3.4527):  13%|█▎        | 79/625 [00:34<03:12,  2.84it/s]
Training 1/1 epoch (loss 3.2472):  13%|█▎        | 79/625 [00:35<03:12,  2.84it/s]
Training 1/1 epoch (loss 3.2472):  13%|█▎        | 80/625 [00:35<03:08,  2.88it/s]
Training 1/1 epoch (loss 3.2342):  13%|█▎        | 80/625 [00:35<03:08,  2.88it/s]
Training 1/1 epoch (loss 3.2342):  13%|█▎        | 81/625 [00:35<03:10,  2.86it/s]
Training 1/1 epoch (loss 3.2009):  13%|█▎        | 81/625 [00:35<03:10,  2.86it/s]
Training 1/1 epoch (loss 3.2009):  13%|█▎        | 82/625 [00:35<03:06,  2.92it/s]
Training 1/1 epoch (loss 3.1946):  13%|█▎        | 82/625 [00:36<03:06,  2.92it/s]
Training 1/1 epoch (loss 3.1946):  13%|█▎        | 83/625 [00:36<03:01,  2.99it/s]
Training 1/1 epoch (loss 2.9060):  13%|█▎        | 83/625 [00:36<03:01,  2.99it/s]
Training 1/1 epoch (loss 2.9060):  13%|█▎        | 84/625 [00:36<03:05,  2.92it/s]
Training 1/1 epoch (loss 3.2977):  13%|█▎        | 84/625 [00:36<03:05,  2.92it/s]
Training 1/1 epoch (loss 3.2977):  14%|█▎        | 85/625 [00:36<03:02,  2.95it/s]
Training 1/1 epoch (loss 3.1080):  14%|█▎        | 85/625 [00:37<03:02,  2.95it/s]
Training 1/1 epoch (loss 3.1080):  14%|█▍        | 86/625 [00:37<03:00,  2.99it/s]
Training 1/1 epoch (loss 3.2930):  14%|█▍        | 86/625 [00:37<03:00,  2.99it/s]
Training 1/1 epoch (loss 3.2930):  14%|█▍        | 87/625 [00:37<02:58,  3.02it/s]
Training 1/1 epoch (loss 3.3207):  14%|█▍        | 87/625 [00:37<02:58,  3.02it/s]
Training 1/1 epoch (loss 3.3207):  14%|█▍        | 88/625 [00:37<03:06,  2.87it/s]
Training 1/1 epoch (loss 3.3501):  14%|█▍        | 88/625 [00:38<03:06,  2.87it/s]
Training 1/1 epoch (loss 3.3501):  14%|█▍        | 89/625 [00:38<03:07,  2.85it/s]
Training 1/1 epoch (loss 3.3579):  14%|█▍        | 89/625 [00:38<03:07,  2.85it/s]
Training 1/1 epoch (loss 3.3579):  14%|█▍        | 90/625 [00:38<03:08,  2.84it/s]
Training 1/1 epoch (loss 2.9940):  14%|█▍        | 90/625 [00:38<03:08,  2.84it/s]
Training 1/1 epoch (loss 2.9940):  15%|█▍        | 91/625 [00:38<03:18,  2.69it/s]
Training 1/1 epoch (loss 3.1116):  15%|█▍        | 91/625 [00:39<03:18,  2.69it/s]
Training 1/1 epoch (loss 3.1116):  15%|█▍        | 92/625 [00:39<03:09,  2.81it/s]
Training 1/1 epoch (loss 3.1253):  15%|█▍        | 92/625 [00:39<03:09,  2.81it/s]
Training 1/1 epoch (loss 3.1253):  15%|█▍        | 93/625 [00:39<03:09,  2.81it/s]
Training 1/1 epoch (loss 3.1985):  15%|█▍        | 93/625 [00:40<03:09,  2.81it/s]
Training 1/1 epoch (loss 3.1985):  15%|█▌        | 94/625 [00:40<03:11,  2.77it/s]
Training 1/1 epoch (loss 3.2941):  15%|█▌        | 94/625 [00:40<03:11,  2.77it/s]
Training 1/1 epoch (loss 3.2941):  15%|█▌        | 95/625 [00:40<03:13,  2.74it/s]
Training 1/1 epoch (loss 3.0701):  15%|█▌        | 95/625 [00:40<03:13,  2.74it/s]
Training 1/1 epoch (loss 3.0701):  15%|█▌        | 96/625 [00:40<03:09,  2.80it/s]
Training 1/1 epoch (loss 3.0740):  15%|█▌        | 96/625 [00:41<03:09,  2.80it/s]
Training 1/1 epoch (loss 3.0740):  16%|█▌        | 97/625 [00:41<03:06,  2.83it/s]
Training 1/1 epoch (loss 3.0261):  16%|█▌        | 97/625 [00:41<03:06,  2.83it/s]
Training 1/1 epoch (loss 3.0261):  16%|█▌        | 98/625 [00:41<03:03,  2.87it/s]
Training 1/1 epoch (loss 3.2906):  16%|█▌        | 98/625 [00:41<03:03,  2.87it/s]
Training 1/1 epoch (loss 3.2906):  16%|█▌        | 99/625 [00:41<02:56,  2.98it/s]
Training 1/1 epoch (loss 3.1607):  16%|█▌        | 99/625 [00:42<02:56,  2.98it/s]
Training 1/1 epoch (loss 3.1607):  16%|█▌        | 100/625 [00:42<03:04,  2.84it/s]
Training 1/1 epoch (loss 3.0722):  16%|█▌        | 100/625 [00:42<03:04,  2.84it/s]
Training 1/1 epoch (loss 3.0722):  16%|█▌        | 101/625 [00:42<03:03,  2.85it/s]
Training 1/1 epoch (loss 3.1539):  16%|█▌        | 101/625 [00:42<03:03,  2.85it/s]
Training 1/1 epoch (loss 3.1539):  16%|█▋        | 102/625 [00:42<02:57,  2.95it/s]
Training 1/1 epoch (loss 3.1304):  16%|█▋        | 102/625 [00:43<02:57,  2.95it/s]
Training 1/1 epoch (loss 3.1304):  16%|█▋        | 103/625 [00:43<02:58,  2.92it/s]
Training 1/1 epoch (loss 3.0015):  16%|█▋        | 103/625 [00:43<02:58,  2.92it/s]
Training 1/1 epoch (loss 3.0015):  17%|█▋        | 104/625 [00:43<02:55,  2.96it/s]
Training 1/1 epoch (loss 3.0181):  17%|█▋        | 104/625 [00:43<02:55,  2.96it/s]
Training 1/1 epoch (loss 3.0181):  17%|█▋        | 105/625 [00:43<02:57,  2.94it/s]
Training 1/1 epoch (loss 2.9742):  17%|█▋        | 105/625 [00:44<02:57,  2.94it/s]
Training 1/1 epoch (loss 2.9742):  17%|█▋        | 106/625 [00:44<02:54,  2.98it/s]
Training 1/1 epoch (loss 3.4916):  17%|█▋        | 106/625 [00:44<02:54,  2.98it/s]
Training 1/1 epoch (loss 3.4916):  17%|█▋        | 107/625 [00:44<02:56,  2.93it/s]
Training 1/1 epoch (loss 3.1583):  17%|█▋        | 107/625 [00:44<02:56,  2.93it/s]
Training 1/1 epoch (loss 3.1583):  17%|█▋        | 108/625 [00:44<02:50,  3.03it/s]
Training 1/1 epoch (loss 3.2220):  17%|█▋        | 108/625 [00:45<02:50,  3.03it/s]
Training 1/1 epoch (loss 3.2220):  17%|█▋        | 109/625 [00:45<02:58,  2.89it/s]
Training 1/1 epoch (loss 3.2727):  17%|█▋        | 109/625 [00:45<02:58,  2.89it/s]
Training 1/1 epoch (loss 3.2727):  18%|█▊        | 110/625 [00:45<02:56,  2.92it/s]
Training 1/1 epoch (loss 3.2626):  18%|█▊        | 110/625 [00:45<02:56,  2.92it/s]
Training 1/1 epoch (loss 3.2626):  18%|█▊        | 111/625 [00:45<02:56,  2.92it/s]
Training 1/1 epoch (loss 3.1803):  18%|█▊        | 111/625 [00:46<02:56,  2.92it/s]
Training 1/1 epoch (loss 3.1803):  18%|█▊        | 112/625 [00:46<03:00,  2.84it/s]
Training 1/1 epoch (loss 2.8998):  18%|█▊        | 112/625 [00:46<03:00,  2.84it/s]
Training 1/1 epoch (loss 2.8998):  18%|█▊        | 113/625 [00:46<03:04,  2.78it/s]
Training 1/1 epoch (loss 3.0520):  18%|█▊        | 113/625 [00:46<03:04,  2.78it/s]
Training 1/1 epoch (loss 3.0520):  18%|█▊        | 114/625 [00:46<02:59,  2.84it/s]
Training 1/1 epoch (loss 3.1207):  18%|█▊        | 114/625 [00:47<02:59,  2.84it/s]
Training 1/1 epoch (loss 3.1207):  18%|█▊        | 115/625 [00:47<02:59,  2.83it/s]
Training 1/1 epoch (loss 3.2280):  18%|█▊        | 115/625 [00:47<02:59,  2.83it/s]
Training 1/1 epoch (loss 3.2280):  19%|█▊        | 116/625 [00:47<02:54,  2.91it/s]
Training 1/1 epoch (loss 3.1190):  19%|█▊        | 116/625 [00:47<02:54,  2.91it/s]
Training 1/1 epoch (loss 3.1190):  19%|█▊        | 117/625 [00:47<02:58,  2.85it/s]
Training 1/1 epoch (loss 2.7914):  19%|█▊        | 117/625 [00:48<02:58,  2.85it/s]
Training 1/1 epoch (loss 2.7914):  19%|█▉        | 118/625 [00:48<02:55,  2.89it/s]
Training 1/1 epoch (loss 2.7535):  19%|█▉        | 118/625 [00:48<02:55,  2.89it/s]
Training 1/1 epoch (loss 2.7535):  19%|█▉        | 119/625 [00:48<02:54,  2.91it/s]
Training 1/1 epoch (loss 2.8812):  19%|█▉        | 119/625 [00:49<02:54,  2.91it/s]
Training 1/1 epoch (loss 2.8812):  19%|█▉        | 120/625 [00:49<02:57,  2.84it/s]
Training 1/1 epoch (loss 2.8459):  19%|█▉        | 120/625 [00:49<02:57,  2.84it/s]
Training 1/1 epoch (loss 2.8459):  19%|█▉        | 121/625 [00:49<02:52,  2.93it/s]
Training 1/1 epoch (loss 2.8998):  19%|█▉        | 121/625 [00:49<02:52,  2.93it/s]
Training 1/1 epoch (loss 2.8998):  20%|█▉        | 122/625 [00:49<02:50,  2.95it/s]
Training 1/1 epoch (loss 2.8491):  20%|█▉        | 122/625 [00:50<02:50,  2.95it/s]
Training 1/1 epoch (loss 2.8491):  20%|█▉        | 123/625 [00:50<02:53,  2.89it/s]
Training 1/1 epoch (loss 3.0644):  20%|█▉        | 123/625 [00:50<02:53,  2.89it/s]
Training 1/1 epoch (loss 3.0644):  20%|█▉        | 124/625 [00:50<03:06,  2.68it/s]
Training 1/1 epoch (loss 2.9004):  20%|█▉        | 124/625 [00:50<03:06,  2.68it/s]
Training 1/1 epoch (loss 2.9004):  20%|██        | 125/625 [00:50<03:01,  2.75it/s]
Training 1/1 epoch (loss 3.0749):  20%|██        | 125/625 [00:51<03:01,  2.75it/s]
Training 1/1 epoch (loss 3.0749):  20%|██        | 126/625 [00:51<03:00,  2.76it/s]
Training 1/1 epoch (loss 3.0616):  20%|██        | 126/625 [00:51<03:00,  2.76it/s]
Training 1/1 epoch (loss 3.0616):  20%|██        | 127/625 [00:51<02:52,  2.88it/s]
Training 1/1 epoch (loss 2.9648):  20%|██        | 127/625 [00:51<02:52,  2.88it/s]
Training 1/1 epoch (loss 2.9648):  20%|██        | 128/625 [00:51<02:50,  2.92it/s]
Training 1/1 epoch (loss 2.8082):  20%|██        | 128/625 [00:52<02:50,  2.92it/s]
Training 1/1 epoch (loss 2.8082):  21%|██        | 129/625 [00:52<02:49,  2.92it/s]
Training 1/1 epoch (loss 3.0368):  21%|██        | 129/625 [00:52<02:49,  2.92it/s]
Training 1/1 epoch (loss 3.0368):  21%|██        | 130/625 [00:52<02:46,  2.98it/s]
Training 1/1 epoch (loss 2.8160):  21%|██        | 130/625 [00:52<02:46,  2.98it/s]
Training 1/1 epoch (loss 2.8160):  21%|██        | 131/625 [00:52<02:48,  2.94it/s]
Training 1/1 epoch (loss 3.1015):  21%|██        | 131/625 [00:53<02:48,  2.94it/s]
Training 1/1 epoch (loss 3.1015):  21%|██        | 132/625 [00:53<02:46,  2.96it/s]
Training 1/1 epoch (loss 3.0130):  21%|██        | 132/625 [00:53<02:46,  2.96it/s]
Training 1/1 epoch (loss 3.0130):  21%|██▏       | 133/625 [00:53<02:56,  2.78it/s]
Training 1/1 epoch (loss 2.8835):  21%|██▏       | 133/625 [00:54<02:56,  2.78it/s]
Training 1/1 epoch (loss 2.8835):  21%|██▏       | 134/625 [00:54<03:27,  2.37it/s]
Training 1/1 epoch (loss 3.0746):  21%|██▏       | 134/625 [00:54<03:27,  2.37it/s]
Training 1/1 epoch (loss 3.0746):  22%|██▏       | 135/625 [00:54<03:19,  2.46it/s]
Training 1/1 epoch (loss 3.1257):  22%|██▏       | 135/625 [00:54<03:19,  2.46it/s]
Training 1/1 epoch (loss 3.1257):  22%|██▏       | 136/625 [00:54<03:12,  2.54it/s]
Training 1/1 epoch (loss 2.9486):  22%|██▏       | 136/625 [00:55<03:12,  2.54it/s]
Training 1/1 epoch (loss 2.9486):  22%|██▏       | 137/625 [00:55<02:59,  2.71it/s]
Training 1/1 epoch (loss 2.9305):  22%|██▏       | 137/625 [00:55<02:59,  2.71it/s]
Training 1/1 epoch (loss 2.9305):  22%|██▏       | 138/625 [00:55<03:23,  2.40it/s]
Training 1/1 epoch (loss 3.0390):  22%|██▏       | 138/625 [00:56<03:23,  2.40it/s]
Training 1/1 epoch (loss 3.0390):  22%|██▏       | 139/625 [00:56<03:13,  2.52it/s]
Training 1/1 epoch (loss 2.8485):  22%|██▏       | 139/625 [00:56<03:13,  2.52it/s]
Training 1/1 epoch (loss 2.8485):  22%|██▏       | 140/625 [00:56<03:10,  2.54it/s]
Training 1/1 epoch (loss 2.8152):  22%|██▏       | 140/625 [00:56<03:10,  2.54it/s]
Training 1/1 epoch (loss 2.8152):  23%|██▎       | 141/625 [00:56<03:04,  2.62it/s]
Training 1/1 epoch (loss 2.9032):  23%|██▎       | 141/625 [00:57<03:04,  2.62it/s]
Training 1/1 epoch (loss 2.9032):  23%|██▎       | 142/625 [00:57<02:56,  2.74it/s]
Training 1/1 epoch (loss 2.9801):  23%|██▎       | 142/625 [00:57<02:56,  2.74it/s]
Training 1/1 epoch (loss 2.9801):  23%|██▎       | 143/625 [00:57<02:56,  2.73it/s]
Training 1/1 epoch (loss 3.1263):  23%|██▎       | 143/625 [00:57<02:56,  2.73it/s]
Training 1/1 epoch (loss 3.1263):  23%|██▎       | 144/625 [00:57<02:54,  2.75it/s]
Training 1/1 epoch (loss 3.1262):  23%|██▎       | 144/625 [00:58<02:54,  2.75it/s]
Training 1/1 epoch (loss 3.1262):  23%|██▎       | 145/625 [00:58<02:56,  2.72it/s]
Training 1/1 epoch (loss 3.0276):  23%|██▎       | 145/625 [00:58<02:56,  2.72it/s]
Training 1/1 epoch (loss 3.0276):  23%|██▎       | 146/625 [00:58<02:55,  2.73it/s]
Training 1/1 epoch (loss 2.9039):  23%|██▎       | 146/625 [00:58<02:55,  2.73it/s]
Training 1/1 epoch (loss 2.9039):  24%|██▎       | 147/625 [00:58<02:50,  2.81it/s]
Training 1/1 epoch (loss 3.0482):  24%|██▎       | 147/625 [00:59<02:50,  2.81it/s]
Training 1/1 epoch (loss 3.0482):  24%|██▎       | 148/625 [00:59<02:45,  2.88it/s]
Training 1/1 epoch (loss 3.0285):  24%|██▎       | 148/625 [00:59<02:45,  2.88it/s]
Training 1/1 epoch (loss 3.0285):  24%|██▍       | 149/625 [00:59<02:40,  2.96it/s]
Training 1/1 epoch (loss 2.9273):  24%|██▍       | 149/625 [00:59<02:40,  2.96it/s]
Training 1/1 epoch (loss 2.9273):  24%|██▍       | 150/625 [00:59<02:40,  2.96it/s]
Training 1/1 epoch (loss 2.9248):  24%|██▍       | 150/625 [01:00<02:40,  2.96it/s]
Training 1/1 epoch (loss 2.9248):  24%|██▍       | 151/625 [01:00<02:44,  2.88it/s]
Training 1/1 epoch (loss 2.9653):  24%|██▍       | 151/625 [01:00<02:44,  2.88it/s]
Training 1/1 epoch (loss 2.9653):  24%|██▍       | 152/625 [01:00<02:56,  2.68it/s]
Training 1/1 epoch (loss 2.9838):  24%|██▍       | 152/625 [01:01<02:56,  2.68it/s]
Training 1/1 epoch (loss 2.9838):  24%|██▍       | 153/625 [01:01<02:48,  2.80it/s]
Training 1/1 epoch (loss 2.9592):  24%|██▍       | 153/625 [01:01<02:48,  2.80it/s]
Training 1/1 epoch (loss 2.9592):  25%|██▍       | 154/625 [01:01<02:45,  2.84it/s]
Training 1/1 epoch (loss 3.0122):  25%|██▍       | 154/625 [01:01<02:45,  2.84it/s]
Training 1/1 epoch (loss 3.0122):  25%|██▍       | 155/625 [01:01<02:39,  2.95it/s]
Training 1/1 epoch (loss 2.8762):  25%|██▍       | 155/625 [01:01<02:39,  2.95it/s]
Training 1/1 epoch (loss 2.8762):  25%|██▍       | 156/625 [01:01<02:35,  3.01it/s]
Training 1/1 epoch (loss 3.2393):  25%|██▍       | 156/625 [01:02<02:35,  3.01it/s]
Training 1/1 epoch (loss 3.2393):  25%|██▌       | 157/625 [01:02<02:33,  3.05it/s]
Training 1/1 epoch (loss 2.9452):  25%|██▌       | 157/625 [01:02<02:33,  3.05it/s]
Training 1/1 epoch (loss 2.9452):  25%|██▌       | 158/625 [01:02<02:42,  2.88it/s]
Training 1/1 epoch (loss 3.1170):  25%|██▌       | 158/625 [01:02<02:42,  2.88it/s]
Training 1/1 epoch (loss 3.1170):  25%|██▌       | 159/625 [01:02<02:33,  3.03it/s]
Training 1/1 epoch (loss 2.8022):  25%|██▌       | 159/625 [01:03<02:33,  3.03it/s]
Training 1/1 epoch (loss 2.8022):  26%|██▌       | 160/625 [01:03<02:35,  3.00it/s]
Training 1/1 epoch (loss 3.0377):  26%|██▌       | 160/625 [01:03<02:35,  3.00it/s]
Training 1/1 epoch (loss 3.0377):  26%|██▌       | 161/625 [01:03<02:35,  2.99it/s]
Training 1/1 epoch (loss 3.0006):  26%|██▌       | 161/625 [01:03<02:35,  2.99it/s]
Training 1/1 epoch (loss 3.0006):  26%|██▌       | 162/625 [01:03<02:32,  3.03it/s]
Training 1/1 epoch (loss 3.0049):  26%|██▌       | 162/625 [01:04<02:32,  3.03it/s]
Training 1/1 epoch (loss 3.0049):  26%|██▌       | 163/625 [01:04<02:36,  2.94it/s]
Training 1/1 epoch (loss 2.9746):  26%|██▌       | 163/625 [01:04<02:36,  2.94it/s]
Training 1/1 epoch (loss 2.9746):  26%|██▌       | 164/625 [01:04<02:39,  2.89it/s]
Training 1/1 epoch (loss 2.9568):  26%|██▌       | 164/625 [01:05<02:39,  2.89it/s]
Training 1/1 epoch (loss 2.9568):  26%|██▋       | 165/625 [01:05<02:34,  2.97it/s]
Training 1/1 epoch (loss 2.8485):  26%|██▋       | 165/625 [01:05<02:34,  2.97it/s]
Training 1/1 epoch (loss 2.8485):  27%|██▋       | 166/625 [01:05<02:32,  3.02it/s]
Training 1/1 epoch (loss 2.7434):  27%|██▋       | 166/625 [01:05<02:32,  3.02it/s]
Training 1/1 epoch (loss 2.7434):  27%|██▋       | 167/625 [01:05<02:32,  3.01it/s]
Training 1/1 epoch (loss 2.8790):  27%|██▋       | 167/625 [01:06<02:32,  3.01it/s]
Training 1/1 epoch (loss 2.8790):  27%|██▋       | 168/625 [01:06<02:34,  2.97it/s]
Training 1/1 epoch (loss 2.6690):  27%|██▋       | 168/625 [01:06<02:34,  2.97it/s]
Training 1/1 epoch (loss 2.6690):  27%|██▋       | 169/625 [01:06<02:45,  2.76it/s]
Training 1/1 epoch (loss 2.9139):  27%|██▋       | 169/625 [01:06<02:45,  2.76it/s]
Training 1/1 epoch (loss 2.9139):  27%|██▋       | 170/625 [01:06<02:48,  2.70it/s]
Training 1/1 epoch (loss 2.8911):  27%|██▋       | 170/625 [01:07<02:48,  2.70it/s]
Training 1/1 epoch (loss 2.8911):  27%|██▋       | 171/625 [01:07<02:40,  2.82it/s]
Training 1/1 epoch (loss 3.0765):  27%|██▋       | 171/625 [01:07<02:40,  2.82it/s]
Training 1/1 epoch (loss 3.0765):  28%|██▊       | 172/625 [01:07<02:36,  2.89it/s]
Training 1/1 epoch (loss 2.8429):  28%|██▊       | 172/625 [01:07<02:36,  2.89it/s]
Training 1/1 epoch (loss 2.8429):  28%|██▊       | 173/625 [01:07<02:42,  2.78it/s]
Training 1/1 epoch (loss 2.6896):  28%|██▊       | 173/625 [01:08<02:42,  2.78it/s]
Training 1/1 epoch (loss 2.6896):  28%|██▊       | 174/625 [01:08<02:37,  2.86it/s]
Training 1/1 epoch (loss 2.9691):  28%|██▊       | 174/625 [01:08<02:37,  2.86it/s]
Training 1/1 epoch (loss 2.9691):  28%|██▊       | 175/625 [01:08<02:34,  2.92it/s]
Training 1/1 epoch (loss 2.8055):  28%|██▊       | 175/625 [01:08<02:34,  2.92it/s]
Training 1/1 epoch (loss 2.8055):  28%|██▊       | 176/625 [01:08<02:43,  2.74it/s]
Training 1/1 epoch (loss 2.9631):  28%|██▊       | 176/625 [01:09<02:43,  2.74it/s]
Training 1/1 epoch (loss 2.9631):  28%|██▊       | 177/625 [01:09<02:40,  2.80it/s]
Training 1/1 epoch (loss 2.7783):  28%|██▊       | 177/625 [01:09<02:40,  2.80it/s]
Training 1/1 epoch (loss 2.7783):  28%|██▊       | 178/625 [01:09<02:34,  2.89it/s]
Training 1/1 epoch (loss 2.8081):  28%|██▊       | 178/625 [01:09<02:34,  2.89it/s]
Training 1/1 epoch (loss 2.8081):  29%|██▊       | 179/625 [01:09<02:33,  2.91it/s]
Training 1/1 epoch (loss 2.8410):  29%|██▊       | 179/625 [01:10<02:33,  2.91it/s]
Training 1/1 epoch (loss 2.8410):  29%|██▉       | 180/625 [01:10<02:35,  2.86it/s]
Training 1/1 epoch (loss 2.8966):  29%|██▉       | 180/625 [01:10<02:35,  2.86it/s]
Training 1/1 epoch (loss 2.8966):  29%|██▉       | 181/625 [01:10<02:35,  2.86it/s]
Training 1/1 epoch (loss 2.9512):  29%|██▉       | 181/625 [01:10<02:35,  2.86it/s]
Training 1/1 epoch (loss 2.9512):  29%|██▉       | 182/625 [01:10<02:33,  2.89it/s]
Training 1/1 epoch (loss 3.0908):  29%|██▉       | 182/625 [01:11<02:33,  2.89it/s]
Training 1/1 epoch (loss 3.0908):  29%|██▉       | 183/625 [01:11<02:34,  2.85it/s]
Training 1/1 epoch (loss 2.9080):  29%|██▉       | 183/625 [01:11<02:34,  2.85it/s]
Training 1/1 epoch (loss 2.9080):  29%|██▉       | 184/625 [01:11<02:32,  2.89it/s]
Training 1/1 epoch (loss 2.7342):  29%|██▉       | 184/625 [01:12<02:32,  2.89it/s]
Training 1/1 epoch (loss 2.7342):  30%|██▉       | 185/625 [01:12<02:30,  2.92it/s]
Training 1/1 epoch (loss 2.8068):  30%|██▉       | 185/625 [01:12<02:30,  2.92it/s]
Training 1/1 epoch (loss 2.8068):  30%|██▉       | 186/625 [01:12<02:27,  2.97it/s]
Training 1/1 epoch (loss 3.0225):  30%|██▉       | 186/625 [01:12<02:27,  2.97it/s]
Training 1/1 epoch (loss 3.0225):  30%|██▉       | 187/625 [01:12<02:31,  2.89it/s]
Training 1/1 epoch (loss 2.6802):  30%|██▉       | 187/625 [01:13<02:31,  2.89it/s]
Training 1/1 epoch (loss 2.6802):  30%|███       | 188/625 [01:13<02:30,  2.91it/s]
Training 1/1 epoch (loss 2.9101):  30%|███       | 188/625 [01:13<02:30,  2.91it/s]
Training 1/1 epoch (loss 2.9101):  30%|███       | 189/625 [01:13<02:23,  3.03it/s]
Training 1/1 epoch (loss 3.0885):  30%|███       | 189/625 [01:13<02:23,  3.03it/s]
Training 1/1 epoch (loss 3.0885):  30%|███       | 190/625 [01:13<02:21,  3.07it/s]
Training 1/1 epoch (loss 2.8741):  30%|███       | 190/625 [01:13<02:21,  3.07it/s]
Training 1/1 epoch (loss 2.8741):  31%|███       | 191/625 [01:13<02:18,  3.13it/s]
Training 1/1 epoch (loss 2.8933):  31%|███       | 191/625 [01:14<02:18,  3.13it/s]
Training 1/1 epoch (loss 2.8933):  31%|███       | 192/625 [01:14<02:28,  2.91it/s]
Training 1/1 epoch (loss 2.7871):  31%|███       | 192/625 [01:14<02:28,  2.91it/s]
Training 1/1 epoch (loss 2.7871):  31%|███       | 193/625 [01:14<02:29,  2.88it/s]
Training 1/1 epoch (loss 2.8790):  31%|███       | 193/625 [01:15<02:29,  2.88it/s]
Training 1/1 epoch (loss 2.8790):  31%|███       | 194/625 [01:15<02:34,  2.79it/s]
Training 1/1 epoch (loss 2.8414):  31%|███       | 194/625 [01:15<02:34,  2.79it/s]
Training 1/1 epoch (loss 2.8414):  31%|███       | 195/625 [01:15<02:33,  2.81it/s]
Training 1/1 epoch (loss 3.1511):  31%|███       | 195/625 [01:15<02:33,  2.81it/s]
Training 1/1 epoch (loss 3.1511):  31%|███▏      | 196/625 [01:15<02:27,  2.91it/s]
Training 1/1 epoch (loss 3.0681):  31%|███▏      | 196/625 [01:16<02:27,  2.91it/s]
Training 1/1 epoch (loss 3.0681):  32%|███▏      | 197/625 [01:16<02:24,  2.97it/s]
Training 1/1 epoch (loss 2.8412):  32%|███▏      | 197/625 [01:16<02:24,  2.97it/s]
Training 1/1 epoch (loss 2.8412):  32%|███▏      | 198/625 [01:16<02:37,  2.72it/s]
Training 1/1 epoch (loss 2.8792):  32%|███▏      | 198/625 [01:16<02:37,  2.72it/s]
Training 1/1 epoch (loss 2.8792):  32%|███▏      | 199/625 [01:16<02:34,  2.75it/s]
Training 1/1 epoch (loss 3.0193):  32%|███▏      | 199/625 [01:17<02:34,  2.75it/s]
Training 1/1 epoch (loss 3.0193):  32%|███▏      | 200/625 [01:17<02:27,  2.88it/s]
Training 1/1 epoch (loss 3.0538):  32%|███▏      | 200/625 [01:17<02:27,  2.88it/s]
Training 1/1 epoch (loss 3.0538):  32%|███▏      | 201/625 [01:17<02:33,  2.76it/s]
Training 1/1 epoch (loss 2.7906):  32%|███▏      | 201/625 [01:17<02:33,  2.76it/s]
Training 1/1 epoch (loss 2.7906):  32%|███▏      | 202/625 [01:17<02:35,  2.73it/s]
Training 1/1 epoch (loss 2.8510):  32%|███▏      | 202/625 [01:18<02:35,  2.73it/s]
Training 1/1 epoch (loss 2.8510):  32%|███▏      | 203/625 [01:18<02:31,  2.78it/s]
Training 1/1 epoch (loss 2.9483):  32%|███▏      | 203/625 [01:18<02:31,  2.78it/s]
Training 1/1 epoch (loss 2.9483):  33%|███▎      | 204/625 [01:18<02:37,  2.68it/s]
Training 1/1 epoch (loss 2.9379):  33%|███▎      | 204/625 [01:19<02:37,  2.68it/s]
Training 1/1 epoch (loss 2.9379):  33%|███▎      | 205/625 [01:19<02:34,  2.72it/s]
Training 1/1 epoch (loss 3.0693):  33%|███▎      | 205/625 [01:19<02:34,  2.72it/s]
Training 1/1 epoch (loss 3.0693):  33%|███▎      | 206/625 [01:19<02:29,  2.80it/s]
Training 1/1 epoch (loss 2.6167):  33%|███▎      | 206/625 [01:19<02:29,  2.80it/s]
Training 1/1 epoch (loss 2.6167):  33%|███▎      | 207/625 [01:19<02:27,  2.83it/s]
Training 1/1 epoch (loss 2.8742):  33%|███▎      | 207/625 [01:20<02:27,  2.83it/s]
Training 1/1 epoch (loss 2.8742):  33%|███▎      | 208/625 [01:20<02:27,  2.83it/s]
Training 1/1 epoch (loss 2.7096):  33%|███▎      | 208/625 [01:20<02:27,  2.83it/s]
Training 1/1 epoch (loss 2.7096):  33%|███▎      | 209/625 [01:20<02:32,  2.73it/s]
Training 1/1 epoch (loss 2.9113):  33%|███▎      | 209/625 [01:20<02:32,  2.73it/s]
Training 1/1 epoch (loss 2.9113):  34%|███▎      | 210/625 [01:20<02:33,  2.70it/s]
Training 1/1 epoch (loss 3.0747):  34%|███▎      | 210/625 [01:21<02:33,  2.70it/s]
Training 1/1 epoch (loss 3.0747):  34%|███▍      | 211/625 [01:21<02:24,  2.86it/s]
Training 1/1 epoch (loss 3.0178):  34%|███▍      | 211/625 [01:21<02:24,  2.86it/s]
Training 1/1 epoch (loss 3.0178):  34%|███▍      | 212/625 [01:21<02:22,  2.90it/s]
Training 1/1 epoch (loss 3.1520):  34%|███▍      | 212/625 [01:21<02:22,  2.90it/s]
Training 1/1 epoch (loss 3.1520):  34%|███▍      | 213/625 [01:21<02:25,  2.82it/s]
Training 1/1 epoch (loss 3.0570):  34%|███▍      | 213/625 [01:22<02:25,  2.82it/s]
Training 1/1 epoch (loss 3.0570):  34%|███▍      | 214/625 [01:22<02:19,  2.94it/s]
Training 1/1 epoch (loss 2.6713):  34%|███▍      | 214/625 [01:22<02:19,  2.94it/s]
Training 1/1 epoch (loss 2.6713):  34%|███▍      | 215/625 [01:22<02:14,  3.05it/s]
Training 1/1 epoch (loss 2.6194):  34%|███▍      | 215/625 [01:22<02:14,  3.05it/s]
Training 1/1 epoch (loss 2.6194):  35%|███▍      | 216/625 [01:22<02:22,  2.88it/s]
Training 1/1 epoch (loss 2.8831):  35%|███▍      | 216/625 [01:23<02:22,  2.88it/s]
Training 1/1 epoch (loss 2.8831):  35%|███▍      | 217/625 [01:23<02:18,  2.94it/s]
Training 1/1 epoch (loss 2.7221):  35%|███▍      | 217/625 [01:23<02:18,  2.94it/s]
Training 1/1 epoch (loss 2.7221):  35%|███▍      | 218/625 [01:23<02:32,  2.67it/s]
Training 1/1 epoch (loss 3.0408):  35%|███▍      | 218/625 [01:24<02:32,  2.67it/s]
Training 1/1 epoch (loss 3.0408):  35%|███▌      | 219/625 [01:24<02:54,  2.33it/s]
Training 1/1 epoch (loss 3.0810):  35%|███▌      | 219/625 [01:24<02:54,  2.33it/s]
Training 1/1 epoch (loss 3.0810):  35%|███▌      | 220/625 [01:24<02:42,  2.50it/s]
Training 1/1 epoch (loss 2.9967):  35%|███▌      | 220/625 [01:24<02:42,  2.50it/s]
Training 1/1 epoch (loss 2.9967):  35%|███▌      | 221/625 [01:24<02:36,  2.58it/s]
Training 1/1 epoch (loss 2.8240):  35%|███▌      | 221/625 [01:25<02:36,  2.58it/s]
Training 1/1 epoch (loss 2.8240):  36%|███▌      | 222/625 [01:25<02:25,  2.76it/s]
Training 1/1 epoch (loss 3.2021):  36%|███▌      | 222/625 [01:25<02:25,  2.76it/s]
Training 1/1 epoch (loss 3.2021):  36%|███▌      | 223/625 [01:25<02:40,  2.50it/s]
Training 1/1 epoch (loss 2.9030):  36%|███▌      | 223/625 [01:26<02:40,  2.50it/s]
Training 1/1 epoch (loss 2.9030):  36%|███▌      | 224/625 [01:26<02:32,  2.63it/s]
Training 1/1 epoch (loss 2.9290):  36%|███▌      | 224/625 [01:26<02:32,  2.63it/s]
Training 1/1 epoch (loss 2.9290):  36%|███▌      | 225/625 [01:26<02:33,  2.61it/s]
Training 1/1 epoch (loss 3.1035):  36%|███▌      | 225/625 [01:26<02:33,  2.61it/s]
Training 1/1 epoch (loss 3.1035):  36%|███▌      | 226/625 [01:26<02:29,  2.67it/s]
Training 1/1 epoch (loss 2.7940):  36%|███▌      | 226/625 [01:27<02:29,  2.67it/s]
Training 1/1 epoch (loss 2.7940):  36%|███▋      | 227/625 [01:27<02:31,  2.63it/s]
Training 1/1 epoch (loss 2.7052):  36%|███▋      | 227/625 [01:27<02:31,  2.63it/s]
Training 1/1 epoch (loss 2.7052):  36%|███▋      | 228/625 [01:27<02:20,  2.82it/s]
Training 1/1 epoch (loss 3.1114):  36%|███▋      | 228/625 [01:27<02:20,  2.82it/s]
Training 1/1 epoch (loss 3.1114):  37%|███▋      | 229/625 [01:27<02:16,  2.90it/s]
Training 1/1 epoch (loss 2.9670):  37%|███▋      | 229/625 [01:28<02:16,  2.90it/s]
Training 1/1 epoch (loss 2.9670):  37%|███▋      | 230/625 [01:28<02:22,  2.77it/s]
Training 1/1 epoch (loss 2.9156):  37%|███▋      | 230/625 [01:28<02:22,  2.77it/s]
Training 1/1 epoch (loss 2.9156):  37%|███▋      | 231/625 [01:28<02:20,  2.79it/s]
Training 1/1 epoch (loss 3.0619):  37%|███▋      | 231/625 [01:28<02:20,  2.79it/s]
Training 1/1 epoch (loss 3.0619):  37%|███▋      | 232/625 [01:28<02:24,  2.72it/s]
Training 1/1 epoch (loss 2.7630):  37%|███▋      | 232/625 [01:29<02:24,  2.72it/s]
Training 1/1 epoch (loss 2.7630):  37%|███▋      | 233/625 [01:29<02:20,  2.80it/s]
Training 1/1 epoch (loss 2.9133):  37%|███▋      | 233/625 [01:29<02:20,  2.80it/s]
Training 1/1 epoch (loss 2.9133):  37%|███▋      | 234/625 [01:29<02:14,  2.91it/s]
Training 1/1 epoch (loss 2.8945):  37%|███▋      | 234/625 [01:29<02:14,  2.91it/s]
Training 1/1 epoch (loss 2.8945):  38%|███▊      | 235/625 [01:29<02:11,  2.97it/s]
Training 1/1 epoch (loss 2.9304):  38%|███▊      | 235/625 [01:30<02:11,  2.97it/s]
Training 1/1 epoch (loss 2.9304):  38%|███▊      | 236/625 [01:30<02:12,  2.94it/s]
Training 1/1 epoch (loss 2.9389):  38%|███▊      | 236/625 [01:30<02:12,  2.94it/s]
Training 1/1 epoch (loss 2.9389):  38%|███▊      | 237/625 [01:30<02:15,  2.87it/s]
Training 1/1 epoch (loss 2.7989):  38%|███▊      | 237/625 [01:30<02:15,  2.87it/s]
Training 1/1 epoch (loss 2.7989):  38%|███▊      | 238/625 [01:30<02:11,  2.93it/s]
Training 1/1 epoch (loss 2.8456):  38%|███▊      | 238/625 [01:31<02:11,  2.93it/s]
Training 1/1 epoch (loss 2.8456):  38%|███▊      | 239/625 [01:31<02:07,  3.02it/s]
Training 1/1 epoch (loss 2.7048):  38%|███▊      | 239/625 [01:31<02:07,  3.02it/s]
Training 1/1 epoch (loss 2.7048):  38%|███▊      | 240/625 [01:31<02:09,  2.96it/s]
Training 1/1 epoch (loss 2.7900):  38%|███▊      | 240/625 [01:31<02:09,  2.96it/s]
Training 1/1 epoch (loss 2.7900):  39%|███▊      | 241/625 [01:31<02:08,  2.98it/s]
Training 1/1 epoch (loss 3.1135):  39%|███▊      | 241/625 [01:32<02:08,  2.98it/s]
Training 1/1 epoch (loss 3.1135):  39%|███▊      | 242/625 [01:32<02:12,  2.89it/s]
Training 1/1 epoch (loss 2.8850):  39%|███▊      | 242/625 [01:32<02:12,  2.89it/s]
Training 1/1 epoch (loss 2.8850):  39%|███▉      | 243/625 [01:32<02:12,  2.87it/s]
Training 1/1 epoch (loss 2.8714):  39%|███▉      | 243/625 [01:32<02:12,  2.87it/s]
Training 1/1 epoch (loss 2.8714):  39%|███▉      | 244/625 [01:32<02:09,  2.94it/s]
Training 1/1 epoch (loss 2.9641):  39%|███▉      | 244/625 [01:33<02:09,  2.94it/s]
Training 1/1 epoch (loss 2.9641):  39%|███▉      | 245/625 [01:33<02:07,  2.99it/s]
Training 1/1 epoch (loss 2.8210):  39%|███▉      | 245/625 [01:33<02:07,  2.99it/s]
Training 1/1 epoch (loss 2.8210):  39%|███▉      | 246/625 [01:33<02:05,  3.01it/s]
Training 1/1 epoch (loss 2.8480):  39%|███▉      | 246/625 [01:33<02:05,  3.01it/s]
Training 1/1 epoch (loss 2.8480):  40%|███▉      | 247/625 [01:33<02:03,  3.05it/s]
Training 1/1 epoch (loss 2.7082):  40%|███▉      | 247/625 [01:34<02:03,  3.05it/s]
Training 1/1 epoch (loss 2.7082):  40%|███▉      | 248/625 [01:34<02:07,  2.97it/s]
Training 1/1 epoch (loss 2.8173):  40%|███▉      | 248/625 [01:34<02:07,  2.97it/s]
Training 1/1 epoch (loss 2.8173):  40%|███▉      | 249/625 [01:34<02:10,  2.88it/s]
Training 1/1 epoch (loss 3.1584):  40%|███▉      | 249/625 [01:35<02:10,  2.88it/s]
Training 1/1 epoch (loss 3.1584):  40%|████      | 250/625 [01:35<02:13,  2.82it/s]
Training 1/1 epoch (loss 2.8013):  40%|████      | 250/625 [01:35<02:13,  2.82it/s]
Training 1/1 epoch (loss 2.8013):  40%|████      | 251/625 [01:35<02:05,  2.98it/s]
Training 1/1 epoch (loss 3.0233):  40%|████      | 251/625 [01:35<02:05,  2.98it/s]
Training 1/1 epoch (loss 3.0233):  40%|████      | 252/625 [01:35<02:04,  2.99it/s]
Training 1/1 epoch (loss 2.9415):  40%|████      | 252/625 [01:35<02:04,  2.99it/s]
Training 1/1 epoch (loss 2.9415):  40%|████      | 253/625 [01:35<02:00,  3.08it/s]
Training 1/1 epoch (loss 2.7488):  40%|████      | 253/625 [01:36<02:00,  3.08it/s]
Training 1/1 epoch (loss 2.7488):  41%|████      | 254/625 [01:36<02:03,  3.01it/s]
Training 1/1 epoch (loss 2.9373):  41%|████      | 254/625 [01:36<02:03,  3.01it/s]
Training 1/1 epoch (loss 2.9373):  41%|████      | 255/625 [01:36<02:05,  2.95it/s]
Training 1/1 epoch (loss 2.8492):  41%|████      | 255/625 [01:37<02:05,  2.95it/s]
Training 1/1 epoch (loss 2.8492):  41%|████      | 256/625 [01:37<02:04,  2.96it/s]
Training 1/1 epoch (loss 2.7330):  41%|████      | 256/625 [01:37<02:04,  2.96it/s]
Training 1/1 epoch (loss 2.7330):  41%|████      | 257/625 [01:37<02:07,  2.88it/s]
Training 1/1 epoch (loss 2.7910):  41%|████      | 257/625 [01:37<02:07,  2.88it/s]
Training 1/1 epoch (loss 2.7910):  41%|████▏     | 258/625 [01:37<02:07,  2.88it/s]
Training 1/1 epoch (loss 3.0964):  41%|████▏     | 258/625 [01:38<02:07,  2.88it/s]
Training 1/1 epoch (loss 3.0964):  41%|████▏     | 259/625 [01:38<02:06,  2.90it/s]
Training 1/1 epoch (loss 2.9284):  41%|████▏     | 259/625 [01:38<02:06,  2.90it/s]
Training 1/1 epoch (loss 2.9284):  42%|████▏     | 260/625 [01:38<02:11,  2.78it/s]
Training 1/1 epoch (loss 2.8488):  42%|████▏     | 260/625 [01:38<02:11,  2.78it/s]
Training 1/1 epoch (loss 2.8488):  42%|████▏     | 261/625 [01:38<02:19,  2.62it/s]
Training 1/1 epoch (loss 2.9464):  42%|████▏     | 261/625 [01:39<02:19,  2.62it/s]
Training 1/1 epoch (loss 2.9464):  42%|████▏     | 262/625 [01:39<02:11,  2.76it/s]
Training 1/1 epoch (loss 2.8296):  42%|████▏     | 262/625 [01:39<02:11,  2.76it/s]
Training 1/1 epoch (loss 2.8296):  42%|████▏     | 263/625 [01:39<02:03,  2.93it/s]
Training 1/1 epoch (loss 2.5757):  42%|████▏     | 263/625 [01:39<02:03,  2.93it/s]
Training 1/1 epoch (loss 2.5757):  42%|████▏     | 264/625 [01:39<02:02,  2.96it/s]
Training 1/1 epoch (loss 2.7696):  42%|████▏     | 264/625 [01:40<02:02,  2.96it/s]
Training 1/1 epoch (loss 2.7696):  42%|████▏     | 265/625 [01:40<02:02,  2.94it/s]
Training 1/1 epoch (loss 2.6879):  42%|████▏     | 265/625 [01:40<02:02,  2.94it/s]
Training 1/1 epoch (loss 2.6879):  43%|████▎     | 266/625 [01:40<02:05,  2.85it/s]
Training 1/1 epoch (loss 2.9676):  43%|████▎     | 266/625 [01:40<02:05,  2.85it/s]
Training 1/1 epoch (loss 2.9676):  43%|████▎     | 267/625 [01:40<02:07,  2.81it/s]
Training 1/1 epoch (loss 2.8130):  43%|████▎     | 267/625 [01:41<02:07,  2.81it/s]
Training 1/1 epoch (loss 2.8130):  43%|████▎     | 268/625 [01:41<02:03,  2.90it/s]
Training 1/1 epoch (loss 2.9101):  43%|████▎     | 268/625 [01:41<02:03,  2.90it/s]
Training 1/1 epoch (loss 2.9101):  43%|████▎     | 269/625 [01:41<02:02,  2.90it/s]
Training 1/1 epoch (loss 2.7579):  43%|████▎     | 269/625 [01:41<02:02,  2.90it/s]
Training 1/1 epoch (loss 2.7579):  43%|████▎     | 270/625 [01:41<01:54,  3.09it/s]
Training 1/1 epoch (loss 2.9286):  43%|████▎     | 270/625 [01:42<01:54,  3.09it/s]
Training 1/1 epoch (loss 2.9286):  43%|████▎     | 271/625 [01:42<01:56,  3.04it/s]
Training 1/1 epoch (loss 2.8555):  43%|████▎     | 271/625 [01:42<01:56,  3.04it/s]
Training 1/1 epoch (loss 2.8555):  44%|████▎     | 272/625 [01:42<01:57,  3.01it/s]
Training 1/1 epoch (loss 2.8910):  44%|████▎     | 272/625 [01:43<01:57,  3.01it/s]
Training 1/1 epoch (loss 2.8910):  44%|████▎     | 273/625 [01:43<02:09,  2.72it/s]
Training 1/1 epoch (loss 2.7419):  44%|████▎     | 273/625 [01:43<02:09,  2.72it/s]
Training 1/1 epoch (loss 2.7419):  44%|████▍     | 274/625 [01:43<02:03,  2.85it/s]
Training 1/1 epoch (loss 2.6593):  44%|████▍     | 274/625 [01:43<02:03,  2.85it/s]
Training 1/1 epoch (loss 2.6593):  44%|████▍     | 275/625 [01:43<01:57,  2.98it/s]
Training 1/1 epoch (loss 2.7345):  44%|████▍     | 275/625 [01:43<01:57,  2.98it/s]
Training 1/1 epoch (loss 2.7345):  44%|████▍     | 276/625 [01:43<01:54,  3.05it/s]
Training 1/1 epoch (loss 2.8554):  44%|████▍     | 276/625 [01:44<01:54,  3.05it/s]
Training 1/1 epoch (loss 2.8554):  44%|████▍     | 277/625 [01:44<01:55,  3.02it/s]
Training 1/1 epoch (loss 2.8825):  44%|████▍     | 277/625 [01:44<01:55,  3.02it/s]
Training 1/1 epoch (loss 2.8825):  44%|████▍     | 278/625 [01:44<01:54,  3.02it/s]
Training 1/1 epoch (loss 2.8773):  44%|████▍     | 278/625 [01:44<01:54,  3.02it/s]
Training 1/1 epoch (loss 2.8773):  45%|████▍     | 279/625 [01:44<01:54,  3.03it/s]
Training 1/1 epoch (loss 2.8234):  45%|████▍     | 279/625 [01:45<01:54,  3.03it/s]
Training 1/1 epoch (loss 2.8234):  45%|████▍     | 280/625 [01:45<02:02,  2.83it/s]
Training 1/1 epoch (loss 3.0202):  45%|████▍     | 280/625 [01:45<02:02,  2.83it/s]
Training 1/1 epoch (loss 3.0202):  45%|████▍     | 281/625 [01:45<02:00,  2.85it/s]
Training 1/1 epoch (loss 2.9086):  45%|████▍     | 281/625 [01:45<02:00,  2.85it/s]
Training 1/1 epoch (loss 2.9086):  45%|████▌     | 282/625 [01:45<01:57,  2.92it/s]
Training 1/1 epoch (loss 2.8511):  45%|████▌     | 282/625 [01:46<01:57,  2.92it/s]
Training 1/1 epoch (loss 2.8511):  45%|████▌     | 283/625 [01:46<01:58,  2.88it/s]
Training 1/1 epoch (loss 3.0290):  45%|████▌     | 283/625 [01:46<01:58,  2.88it/s]
Training 1/1 epoch (loss 3.0290):  45%|████▌     | 284/625 [01:46<01:58,  2.88it/s]
Training 1/1 epoch (loss 2.7807):  45%|████▌     | 284/625 [01:47<01:58,  2.88it/s]
Training 1/1 epoch (loss 2.7807):  46%|████▌     | 285/625 [01:47<02:02,  2.77it/s]
Training 1/1 epoch (loss 2.8170):  46%|████▌     | 285/625 [01:47<02:02,  2.77it/s]
Training 1/1 epoch (loss 2.8170):  46%|████▌     | 286/625 [01:47<01:58,  2.86it/s]
Training 1/1 epoch (loss 2.8256):  46%|████▌     | 286/625 [01:47<01:58,  2.86it/s]
Training 1/1 epoch (loss 2.8256):  46%|████▌     | 287/625 [01:47<01:55,  2.91it/s]
Training 1/1 epoch (loss 2.6855):  46%|████▌     | 287/625 [01:48<01:55,  2.91it/s]
Training 1/1 epoch (loss 2.6855):  46%|████▌     | 288/625 [01:48<02:07,  2.65it/s]
Training 1/1 epoch (loss 2.8864):  46%|████▌     | 288/625 [01:48<02:07,  2.65it/s]
Training 1/1 epoch (loss 2.8864):  46%|████▌     | 289/625 [01:48<02:03,  2.71it/s]
Training 1/1 epoch (loss 3.0540):  46%|████▌     | 289/625 [01:48<02:03,  2.71it/s]
Training 1/1 epoch (loss 3.0540):  46%|████▋     | 290/625 [01:48<02:06,  2.65it/s]
Training 1/1 epoch (loss 2.8701):  46%|████▋     | 290/625 [01:49<02:06,  2.65it/s]
Training 1/1 epoch (loss 2.8701):  47%|████▋     | 291/625 [01:49<02:00,  2.77it/s]
Training 1/1 epoch (loss 3.2075):  47%|████▋     | 291/625 [01:49<02:00,  2.77it/s]
Training 1/1 epoch (loss 3.2075):  47%|████▋     | 292/625 [01:49<01:57,  2.82it/s]
Training 1/1 epoch (loss 2.9390):  47%|████▋     | 292/625 [01:49<01:57,  2.82it/s]
Training 1/1 epoch (loss 2.9390):  47%|████▋     | 293/625 [01:49<01:56,  2.86it/s]
Training 1/1 epoch (loss 2.6657):  47%|████▋     | 293/625 [01:50<01:56,  2.86it/s]
Training 1/1 epoch (loss 2.6657):  47%|████▋     | 294/625 [01:50<01:53,  2.91it/s]
Training 1/1 epoch (loss 2.8654):  47%|████▋     | 294/625 [01:50<01:53,  2.91it/s]
Training 1/1 epoch (loss 2.8654):  47%|████▋     | 295/625 [01:50<01:52,  2.94it/s]
Training 1/1 epoch (loss 2.7386):  47%|████▋     | 295/625 [01:51<01:52,  2.94it/s]
Training 1/1 epoch (loss 2.7386):  47%|████▋     | 296/625 [01:51<02:00,  2.73it/s]
Training 1/1 epoch (loss 2.9759):  47%|████▋     | 296/625 [01:51<02:00,  2.73it/s]
Training 1/1 epoch (loss 2.9759):  48%|████▊     | 297/625 [01:51<01:58,  2.78it/s]
Training 1/1 epoch (loss 2.6290):  48%|████▊     | 297/625 [01:51<01:58,  2.78it/s]
Training 1/1 epoch (loss 2.6290):  48%|████▊     | 298/625 [01:51<01:54,  2.85it/s]
Training 1/1 epoch (loss 2.4788):  48%|████▊     | 298/625 [01:52<01:54,  2.85it/s]
Training 1/1 epoch (loss 2.4788):  48%|████▊     | 299/625 [01:52<01:52,  2.90it/s]
Training 1/1 epoch (loss 2.7123):  48%|████▊     | 299/625 [01:52<01:52,  2.90it/s]
Training 1/1 epoch (loss 2.7123):  48%|████▊     | 300/625 [01:52<01:48,  2.98it/s]
Training 1/1 epoch (loss 2.7751):  48%|████▊     | 300/625 [01:52<01:48,  2.98it/s]
Training 1/1 epoch (loss 2.7751):  48%|████▊     | 301/625 [01:52<01:51,  2.90it/s]
Training 1/1 epoch (loss 2.8971):  48%|████▊     | 301/625 [01:53<01:51,  2.90it/s]
Training 1/1 epoch (loss 2.8971):  48%|████▊     | 302/625 [01:53<01:55,  2.80it/s]
Training 1/1 epoch (loss 2.7301):  48%|████▊     | 302/625 [01:53<01:55,  2.80it/s]
Training 1/1 epoch (loss 2.7301):  48%|████▊     | 303/625 [01:53<02:07,  2.53it/s]
Training 1/1 epoch (loss 2.9733):  48%|████▊     | 303/625 [01:54<02:07,  2.53it/s]
Training 1/1 epoch (loss 2.9733):  49%|████▊     | 304/625 [01:54<02:22,  2.26it/s]
Training 1/1 epoch (loss 2.7999):  49%|████▊     | 304/625 [01:54<02:22,  2.26it/s]
Training 1/1 epoch (loss 2.7999):  49%|████▉     | 305/625 [01:54<02:16,  2.35it/s]
Training 1/1 epoch (loss 2.6789):  49%|████▉     | 305/625 [01:54<02:16,  2.35it/s]
Training 1/1 epoch (loss 2.6789):  49%|████▉     | 306/625 [01:54<02:08,  2.49it/s]
Training 1/1 epoch (loss 3.0435):  49%|████▉     | 306/625 [01:55<02:08,  2.49it/s]
Training 1/1 epoch (loss 3.0435):  49%|████▉     | 307/625 [01:55<02:04,  2.56it/s]
Training 1/1 epoch (loss 2.8550):  49%|████▉     | 307/625 [01:55<02:04,  2.56it/s]
Training 1/1 epoch (loss 2.8550):  49%|████▉     | 308/625 [01:55<02:03,  2.56it/s]
Training 1/1 epoch (loss 2.8375):  49%|████▉     | 308/625 [01:56<02:03,  2.56it/s]
Training 1/1 epoch (loss 2.8375):  49%|████▉     | 309/625 [01:56<02:12,  2.38it/s]
Training 1/1 epoch (loss 2.7100):  49%|████▉     | 309/625 [01:56<02:12,  2.38it/s]
Training 1/1 epoch (loss 2.7100):  50%|████▉     | 310/625 [01:56<02:18,  2.28it/s]
Training 1/1 epoch (loss 2.6405):  50%|████▉     | 310/625 [01:56<02:18,  2.28it/s]
Training 1/1 epoch (loss 2.6405):  50%|████▉     | 311/625 [01:56<02:07,  2.47it/s]
Training 1/1 epoch (loss 2.9408):  50%|████▉     | 311/625 [01:57<02:07,  2.47it/s]
Training 1/1 epoch (loss 2.9408):  50%|████▉     | 312/625 [01:57<02:01,  2.57it/s]
Training 1/1 epoch (loss 2.7525):  50%|████▉     | 312/625 [01:57<02:01,  2.57it/s]
Training 1/1 epoch (loss 2.7525):  50%|█████     | 313/625 [01:57<01:58,  2.64it/s]
Training 1/1 epoch (loss 2.7530):  50%|█████     | 313/625 [01:58<01:58,  2.64it/s]
Training 1/1 epoch (loss 2.7530):  50%|█████     | 314/625 [01:58<01:56,  2.67it/s]
Training 1/1 epoch (loss 3.0355):  50%|█████     | 314/625 [01:58<01:56,  2.67it/s]
Training 1/1 epoch (loss 3.0355):  50%|█████     | 315/625 [01:58<01:59,  2.59it/s]
Training 1/1 epoch (loss 2.9396):  50%|█████     | 315/625 [01:58<01:59,  2.59it/s]
Training 1/1 epoch (loss 2.9396):  51%|█████     | 316/625 [01:58<01:53,  2.72it/s]
Training 1/1 epoch (loss 2.8896):  51%|█████     | 316/625 [01:59<01:53,  2.72it/s]
Training 1/1 epoch (loss 2.8896):  51%|█████     | 317/625 [01:59<02:02,  2.52it/s]
Training 1/1 epoch (loss 2.7484):  51%|█████     | 317/625 [01:59<02:02,  2.52it/s]
Training 1/1 epoch (loss 2.7484):  51%|█████     | 318/625 [01:59<01:53,  2.71it/s]
Training 1/1 epoch (loss 2.9326):  51%|█████     | 318/625 [01:59<01:53,  2.71it/s]
Training 1/1 epoch (loss 2.9326):  51%|█████     | 319/625 [01:59<01:47,  2.85it/s]
Training 1/1 epoch (loss 2.7682):  51%|█████     | 319/625 [02:00<01:47,  2.85it/s]
Training 1/1 epoch (loss 2.7682):  51%|█████     | 320/625 [02:00<01:53,  2.68it/s]
Training 1/1 epoch (loss 2.8991):  51%|█████     | 320/625 [02:00<01:53,  2.68it/s]
Training 1/1 epoch (loss 2.8991):  51%|█████▏    | 321/625 [02:00<01:47,  2.83it/s]
Training 1/1 epoch (loss 2.8205):  51%|█████▏    | 321/625 [02:00<01:47,  2.83it/s]
Training 1/1 epoch (loss 2.8205):  52%|█████▏    | 322/625 [02:00<01:47,  2.81it/s]
Training 1/1 epoch (loss 2.8870):  52%|█████▏    | 322/625 [02:01<01:47,  2.81it/s]
Training 1/1 epoch (loss 2.8870):  52%|█████▏    | 323/625 [02:01<01:44,  2.90it/s]
Training 1/1 epoch (loss 2.8745):  52%|█████▏    | 323/625 [02:01<01:44,  2.90it/s]
Training 1/1 epoch (loss 2.8745):  52%|█████▏    | 324/625 [02:01<01:43,  2.90it/s]
Training 1/1 epoch (loss 3.0285):  52%|█████▏    | 324/625 [02:01<01:43,  2.90it/s]
Training 1/1 epoch (loss 3.0285):  52%|█████▏    | 325/625 [02:01<01:41,  2.95it/s]
Training 1/1 epoch (loss 2.7211):  52%|█████▏    | 325/625 [02:02<01:41,  2.95it/s]
Training 1/1 epoch (loss 2.7211):  52%|█████▏    | 326/625 [02:02<01:39,  3.00it/s]
Training 1/1 epoch (loss 2.8275):  52%|█████▏    | 326/625 [02:02<01:39,  3.00it/s]
Training 1/1 epoch (loss 2.8275):  52%|█████▏    | 327/625 [02:02<01:43,  2.88it/s]
Training 1/1 epoch (loss 2.7601):  52%|█████▏    | 327/625 [02:02<01:43,  2.88it/s]
Training 1/1 epoch (loss 2.7601):  52%|█████▏    | 328/625 [02:02<01:40,  2.94it/s]
Training 1/1 epoch (loss 2.7470):  52%|█████▏    | 328/625 [02:03<01:40,  2.94it/s]
Training 1/1 epoch (loss 2.7470):  53%|█████▎    | 329/625 [02:03<01:44,  2.84it/s]
Training 1/1 epoch (loss 2.7938):  53%|█████▎    | 329/625 [02:03<01:44,  2.84it/s]
Training 1/1 epoch (loss 2.7938):  53%|█████▎    | 330/625 [02:03<01:42,  2.87it/s]
Training 1/1 epoch (loss 2.8559):  53%|█████▎    | 330/625 [02:03<01:42,  2.87it/s]
Training 1/1 epoch (loss 2.8559):  53%|█████▎    | 331/625 [02:03<01:37,  3.01it/s]
Training 1/1 epoch (loss 3.0565):  53%|█████▎    | 331/625 [02:04<01:37,  3.01it/s]
Training 1/1 epoch (loss 3.0565):  53%|█████▎    | 332/625 [02:04<01:35,  3.05it/s]
Training 1/1 epoch (loss 2.6913):  53%|█████▎    | 332/625 [02:04<01:35,  3.05it/s]
Training 1/1 epoch (loss 2.6913):  53%|█████▎    | 333/625 [02:04<01:39,  2.93it/s]
Training 1/1 epoch (loss 2.8509):  53%|█████▎    | 333/625 [02:04<01:39,  2.93it/s]
Training 1/1 epoch (loss 2.8509):  53%|█████▎    | 334/625 [02:04<01:40,  2.89it/s]
Training 1/1 epoch (loss 2.7224):  53%|█████▎    | 334/625 [02:05<01:40,  2.89it/s]
Training 1/1 epoch (loss 2.7224):  54%|█████▎    | 335/625 [02:05<01:39,  2.91it/s]
Training 1/1 epoch (loss 2.9649):  54%|█████▎    | 335/625 [02:05<01:39,  2.91it/s]
Training 1/1 epoch (loss 2.9649):  54%|█████▍    | 336/625 [02:05<01:38,  2.94it/s]
Training 1/1 epoch (loss 2.9490):  54%|█████▍    | 336/625 [02:05<01:38,  2.94it/s]
Training 1/1 epoch (loss 2.9490):  54%|█████▍    | 337/625 [02:05<01:36,  2.99it/s]
Training 1/1 epoch (loss 2.9024):  54%|█████▍    | 337/625 [02:06<01:36,  2.99it/s]
Training 1/1 epoch (loss 2.9024):  54%|█████▍    | 338/625 [02:06<01:34,  3.04it/s]
Training 1/1 epoch (loss 2.7625):  54%|█████▍    | 338/625 [02:06<01:34,  3.04it/s]
Training 1/1 epoch (loss 2.7625):  54%|█████▍    | 339/625 [02:06<01:37,  2.94it/s]
Training 1/1 epoch (loss 2.9434):  54%|█████▍    | 339/625 [02:07<01:37,  2.94it/s]
Training 1/1 epoch (loss 2.9434):  54%|█████▍    | 340/625 [02:07<01:39,  2.88it/s]
Training 1/1 epoch (loss 2.6812):  54%|█████▍    | 340/625 [02:07<01:39,  2.88it/s]
Training 1/1 epoch (loss 2.6812):  55%|█████▍    | 341/625 [02:07<01:35,  2.97it/s]
Training 1/1 epoch (loss 2.7866):  55%|█████▍    | 341/625 [02:07<01:35,  2.97it/s]
Training 1/1 epoch (loss 2.7866):  55%|█████▍    | 342/625 [02:07<01:35,  2.95it/s]
Training 1/1 epoch (loss 2.7744):  55%|█████▍    | 342/625 [02:08<01:35,  2.95it/s]
Training 1/1 epoch (loss 2.7744):  55%|█████▍    | 343/625 [02:08<01:34,  2.98it/s]
Training 1/1 epoch (loss 2.6939):  55%|█████▍    | 343/625 [02:08<01:34,  2.98it/s]
Training 1/1 epoch (loss 2.6939):  55%|█████▌    | 344/625 [02:08<01:38,  2.85it/s]
Training 1/1 epoch (loss 2.7828):  55%|█████▌    | 344/625 [02:08<01:38,  2.85it/s]
Training 1/1 epoch (loss 2.7828):  55%|█████▌    | 345/625 [02:08<01:42,  2.73it/s]
Training 1/1 epoch (loss 2.6326):  55%|█████▌    | 345/625 [02:09<01:42,  2.73it/s]
Training 1/1 epoch (loss 2.6326):  55%|█████▌    | 346/625 [02:09<01:46,  2.63it/s]
Training 1/1 epoch (loss 2.8010):  55%|█████▌    | 346/625 [02:09<01:46,  2.63it/s]
Training 1/1 epoch (loss 2.8010):  56%|█████▌    | 347/625 [02:09<01:41,  2.73it/s]
Training 1/1 epoch (loss 3.0749):  56%|█████▌    | 347/625 [02:09<01:41,  2.73it/s]
Training 1/1 epoch (loss 3.0749):  56%|█████▌    | 348/625 [02:09<01:38,  2.81it/s]
Training 1/1 epoch (loss 2.7890):  56%|█████▌    | 348/625 [02:10<01:38,  2.81it/s]
Training 1/1 epoch (loss 2.7890):  56%|█████▌    | 349/625 [02:10<01:35,  2.89it/s]
Training 1/1 epoch (loss 2.8778):  56%|█████▌    | 349/625 [02:10<01:35,  2.89it/s]
Training 1/1 epoch (loss 2.8778):  56%|█████▌    | 350/625 [02:10<01:34,  2.90it/s]
Training 1/1 epoch (loss 2.6726):  56%|█████▌    | 350/625 [02:10<01:34,  2.90it/s]
Training 1/1 epoch (loss 2.6726):  56%|█████▌    | 351/625 [02:10<01:36,  2.83it/s]
Training 1/1 epoch (loss 3.0540):  56%|█████▌    | 351/625 [02:11<01:36,  2.83it/s]
Training 1/1 epoch (loss 3.0540):  56%|█████▋    | 352/625 [02:11<01:36,  2.84it/s]
Training 1/1 epoch (loss 3.2049):  56%|█████▋    | 352/625 [02:11<01:36,  2.84it/s]
Training 1/1 epoch (loss 3.2049):  56%|█████▋    | 353/625 [02:11<01:35,  2.86it/s]
Training 1/1 epoch (loss 2.8681):  56%|█████▋    | 353/625 [02:11<01:35,  2.86it/s]
Training 1/1 epoch (loss 2.8681):  57%|█████▋    | 354/625 [02:11<01:29,  3.01it/s]
Training 1/1 epoch (loss 2.8644):  57%|█████▋    | 354/625 [02:12<01:29,  3.01it/s]
Training 1/1 epoch (loss 2.8644):  57%|█████▋    | 355/625 [02:12<01:31,  2.95it/s]
Training 1/1 epoch (loss 2.7387):  57%|█████▋    | 355/625 [02:12<01:31,  2.95it/s]
Training 1/1 epoch (loss 2.7387):  57%|█████▋    | 356/625 [02:12<01:31,  2.95it/s]
Training 1/1 epoch (loss 2.9670):  57%|█████▋    | 356/625 [02:12<01:31,  2.95it/s]
Training 1/1 epoch (loss 2.9670):  57%|█████▋    | 357/625 [02:12<01:32,  2.90it/s]
Training 1/1 epoch (loss 2.7463):  57%|█████▋    | 357/625 [02:13<01:32,  2.90it/s]
Training 1/1 epoch (loss 2.7463):  57%|█████▋    | 358/625 [02:13<01:28,  3.01it/s]
Training 1/1 epoch (loss 3.0121):  57%|█████▋    | 358/625 [02:13<01:28,  3.01it/s]
Training 1/1 epoch (loss 3.0121):  57%|█████▋    | 359/625 [02:13<01:27,  3.05it/s]
Training 1/1 epoch (loss 2.6433):  57%|█████▋    | 359/625 [02:13<01:27,  3.05it/s]
Training 1/1 epoch (loss 2.6433):  58%|█████▊    | 360/625 [02:13<01:26,  3.06it/s]
Training 1/1 epoch (loss 2.8228):  58%|█████▊    | 360/625 [02:14<01:26,  3.06it/s]
Training 1/1 epoch (loss 2.8228):  58%|█████▊    | 361/625 [02:14<01:33,  2.82it/s]
Training 1/1 epoch (loss 2.7380):  58%|█████▊    | 361/625 [02:14<01:33,  2.82it/s]
Training 1/1 epoch (loss 2.7380):  58%|█████▊    | 362/625 [02:14<01:31,  2.86it/s]
Training 1/1 epoch (loss 2.8398):  58%|█████▊    | 362/625 [02:15<01:31,  2.86it/s]
Training 1/1 epoch (loss 2.8398):  58%|█████▊    | 363/625 [02:15<01:35,  2.76it/s]
Training 1/1 epoch (loss 2.8425):  58%|█████▊    | 363/625 [02:15<01:35,  2.76it/s]
Training 1/1 epoch (loss 2.8425):  58%|█████▊    | 364/625 [02:15<01:34,  2.77it/s]
Training 1/1 epoch (loss 2.6096):  58%|█████▊    | 364/625 [02:15<01:34,  2.77it/s]
Training 1/1 epoch (loss 2.6096):  58%|█████▊    | 365/625 [02:15<01:33,  2.79it/s]
Training 1/1 epoch (loss 2.7548):  58%|█████▊    | 365/625 [02:16<01:33,  2.79it/s]
Training 1/1 epoch (loss 2.7548):  59%|█████▊    | 366/625 [02:16<01:28,  2.92it/s]
Training 1/1 epoch (loss 2.7297):  59%|█████▊    | 366/625 [02:16<01:28,  2.92it/s]
Training 1/1 epoch (loss 2.7297):  59%|█████▊    | 367/625 [02:16<01:28,  2.92it/s]
Training 1/1 epoch (loss 2.7244):  59%|█████▊    | 367/625 [02:16<01:28,  2.92it/s]
Training 1/1 epoch (loss 2.7244):  59%|█████▉    | 368/625 [02:16<01:29,  2.87it/s]
Training 1/1 epoch (loss 2.6339):  59%|█████▉    | 368/625 [02:17<01:29,  2.87it/s]
Training 1/1 epoch (loss 2.6339):  59%|█████▉    | 369/625 [02:17<01:30,  2.82it/s]
Training 1/1 epoch (loss 2.8481):  59%|█████▉    | 369/625 [02:17<01:30,  2.82it/s]
Training 1/1 epoch (loss 2.8481):  59%|█████▉    | 370/625 [02:17<01:29,  2.84it/s]
Training 1/1 epoch (loss 2.8717):  59%|█████▉    | 370/625 [02:17<01:29,  2.84it/s]
Training 1/1 epoch (loss 2.8717):  59%|█████▉    | 371/625 [02:17<01:29,  2.84it/s]
Training 1/1 epoch (loss 2.9800):  59%|█████▉    | 371/625 [02:18<01:29,  2.84it/s]
Training 1/1 epoch (loss 2.9800):  60%|█████▉    | 372/625 [02:18<01:27,  2.89it/s]
Training 1/1 epoch (loss 2.7415):  60%|█████▉    | 372/625 [02:18<01:27,  2.89it/s]
Training 1/1 epoch (loss 2.7415):  60%|█████▉    | 373/625 [02:18<01:27,  2.87it/s]
Training 1/1 epoch (loss 2.9108):  60%|█████▉    | 373/625 [02:18<01:27,  2.87it/s]
Training 1/1 epoch (loss 2.9108):  60%|█████▉    | 374/625 [02:18<01:28,  2.84it/s]
Training 1/1 epoch (loss 2.8502):  60%|█████▉    | 374/625 [02:19<01:28,  2.84it/s]
Training 1/1 epoch (loss 2.8502):  60%|██████    | 375/625 [02:19<01:29,  2.79it/s]
Training 1/1 epoch (loss 2.7281):  60%|██████    | 375/625 [02:19<01:29,  2.79it/s]
Training 1/1 epoch (loss 2.7281):  60%|██████    | 376/625 [02:19<01:40,  2.47it/s]
Training 1/1 epoch (loss 2.6492):  60%|██████    | 376/625 [02:20<01:40,  2.47it/s]
Training 1/1 epoch (loss 2.6492):  60%|██████    | 377/625 [02:20<01:34,  2.62it/s]
Training 1/1 epoch (loss 2.6770):  60%|██████    | 377/625 [02:20<01:34,  2.62it/s]
Training 1/1 epoch (loss 2.6770):  60%|██████    | 378/625 [02:20<01:32,  2.67it/s]
Training 1/1 epoch (loss 2.9395):  60%|██████    | 378/625 [02:20<01:32,  2.67it/s]
Training 1/1 epoch (loss 2.9395):  61%|██████    | 379/625 [02:20<01:28,  2.79it/s]
Training 1/1 epoch (loss 2.6852):  61%|██████    | 379/625 [02:21<01:28,  2.79it/s]
Training 1/1 epoch (loss 2.6852):  61%|██████    | 380/625 [02:21<01:26,  2.85it/s]
Training 1/1 epoch (loss 2.9518):  61%|██████    | 380/625 [02:21<01:26,  2.85it/s]
Training 1/1 epoch (loss 2.9518):  61%|██████    | 381/625 [02:21<01:23,  2.92it/s]
Training 1/1 epoch (loss 2.7560):  61%|██████    | 381/625 [02:21<01:23,  2.92it/s]
Training 1/1 epoch (loss 2.7560):  61%|██████    | 382/625 [02:21<01:23,  2.90it/s]
Training 1/1 epoch (loss 2.9357):  61%|██████    | 382/625 [02:22<01:23,  2.90it/s]
Training 1/1 epoch (loss 2.9357):  61%|██████▏   | 383/625 [02:22<01:20,  3.01it/s]
Training 1/1 epoch (loss 2.7042):  61%|██████▏   | 383/625 [02:22<01:20,  3.01it/s]
Training 1/1 epoch (loss 2.7042):  61%|██████▏   | 384/625 [02:22<01:21,  2.97it/s]
Training 1/1 epoch (loss 2.7122):  61%|██████▏   | 384/625 [02:22<01:21,  2.97it/s]
Training 1/1 epoch (loss 2.7122):  62%|██████▏   | 385/625 [02:22<01:22,  2.92it/s]
Training 1/1 epoch (loss 2.8466):  62%|██████▏   | 385/625 [02:23<01:22,  2.92it/s]
Training 1/1 epoch (loss 2.8466):  62%|██████▏   | 386/625 [02:23<01:23,  2.86it/s]
Training 1/1 epoch (loss 2.8726):  62%|██████▏   | 386/625 [02:23<01:23,  2.86it/s]
Training 1/1 epoch (loss 2.8726):  62%|██████▏   | 387/625 [02:23<01:45,  2.25it/s]
Training 1/1 epoch (loss 2.8319):  62%|██████▏   | 387/625 [02:24<01:45,  2.25it/s]
Training 1/1 epoch (loss 2.8319):  62%|██████▏   | 388/625 [02:24<01:43,  2.30it/s]
Training 1/1 epoch (loss 2.6759):  62%|██████▏   | 388/625 [02:24<01:43,  2.30it/s]
Training 1/1 epoch (loss 2.6759):  62%|██████▏   | 389/625 [02:24<01:39,  2.38it/s]
Training 1/1 epoch (loss 2.8952):  62%|██████▏   | 389/625 [02:24<01:39,  2.38it/s]
Training 1/1 epoch (loss 2.8952):  62%|██████▏   | 390/625 [02:24<01:34,  2.48it/s]
Training 1/1 epoch (loss 2.8186):  62%|██████▏   | 390/625 [02:25<01:34,  2.48it/s]
Training 1/1 epoch (loss 2.8186):  63%|██████▎   | 391/625 [02:25<01:29,  2.62it/s]
Training 1/1 epoch (loss 2.9721):  63%|██████▎   | 391/625 [02:25<01:29,  2.62it/s]
Training 1/1 epoch (loss 2.9721):  63%|██████▎   | 392/625 [02:25<01:32,  2.52it/s]
Training 1/1 epoch (loss 2.7893):  63%|██████▎   | 392/625 [02:26<01:32,  2.52it/s]
Training 1/1 epoch (loss 2.7893):  63%|██████▎   | 393/625 [02:26<01:28,  2.61it/s]
Training 1/1 epoch (loss 2.7919):  63%|██████▎   | 393/625 [02:26<01:28,  2.61it/s]
Training 1/1 epoch (loss 2.7919):  63%|██████▎   | 394/625 [02:26<01:26,  2.68it/s]
Training 1/1 epoch (loss 2.8539):  63%|██████▎   | 394/625 [02:26<01:26,  2.68it/s]
Training 1/1 epoch (loss 2.8539):  63%|██████▎   | 395/625 [02:26<01:23,  2.75it/s]
Training 1/1 epoch (loss 2.6532):  63%|██████▎   | 395/625 [02:27<01:23,  2.75it/s]
Training 1/1 epoch (loss 2.6532):  63%|██████▎   | 396/625 [02:27<01:23,  2.75it/s]
Training 1/1 epoch (loss 2.8154):  63%|██████▎   | 396/625 [02:27<01:23,  2.75it/s]
Training 1/1 epoch (loss 2.8154):  64%|██████▎   | 397/625 [02:27<01:19,  2.86it/s]
Training 1/1 epoch (loss 2.7874):  64%|██████▎   | 397/625 [02:27<01:19,  2.86it/s]
Training 1/1 epoch (loss 2.7874):  64%|██████▎   | 398/625 [02:27<01:18,  2.89it/s]
Training 1/1 epoch (loss 2.7940):  64%|██████▎   | 398/625 [02:28<01:18,  2.89it/s]
Training 1/1 epoch (loss 2.7940):  64%|██████▍   | 399/625 [02:28<01:14,  3.03it/s]
Training 1/1 epoch (loss 2.6816):  64%|██████▍   | 399/625 [02:28<01:14,  3.03it/s]
Training 1/1 epoch (loss 2.6816):  64%|██████▍   | 400/625 [02:28<01:17,  2.89it/s]
Training 1/1 epoch (loss 2.7575):  64%|██████▍   | 400/625 [02:28<01:17,  2.89it/s]
Training 1/1 epoch (loss 2.7575):  64%|██████▍   | 401/625 [02:28<01:17,  2.87it/s]
Training 1/1 epoch (loss 2.8556):  64%|██████▍   | 401/625 [02:29<01:17,  2.87it/s]
Training 1/1 epoch (loss 2.8556):  64%|██████▍   | 402/625 [02:29<01:19,  2.80it/s]
Training 1/1 epoch (loss 3.0696):  64%|██████▍   | 402/625 [02:29<01:19,  2.80it/s]
Training 1/1 epoch (loss 3.0696):  64%|██████▍   | 403/625 [02:29<01:16,  2.90it/s]
Training 1/1 epoch (loss 2.8554):  64%|██████▍   | 403/625 [02:29<01:16,  2.90it/s]
Training 1/1 epoch (loss 2.8554):  65%|██████▍   | 404/625 [02:29<01:19,  2.77it/s]
Training 1/1 epoch (loss 2.7960):  65%|██████▍   | 404/625 [02:30<01:19,  2.77it/s]
Training 1/1 epoch (loss 2.7960):  65%|██████▍   | 405/625 [02:30<01:21,  2.71it/s]
Training 1/1 epoch (loss 2.7661):  65%|██████▍   | 405/625 [02:30<01:21,  2.71it/s]
Training 1/1 epoch (loss 2.7661):  65%|██████▍   | 406/625 [02:30<01:21,  2.67it/s]
Training 1/1 epoch (loss 2.8643):  65%|██████▍   | 406/625 [02:31<01:21,  2.67it/s]
Training 1/1 epoch (loss 2.8643):  65%|██████▌   | 407/625 [02:31<01:17,  2.80it/s]
Training 1/1 epoch (loss 2.8489):  65%|██████▌   | 407/625 [02:31<01:17,  2.80it/s]
Training 1/1 epoch (loss 2.8489):  65%|██████▌   | 408/625 [02:31<01:16,  2.83it/s]
Training 1/1 epoch (loss 2.9108):  65%|██████▌   | 408/625 [02:31<01:16,  2.83it/s]
Training 1/1 epoch (loss 2.9108):  65%|██████▌   | 409/625 [02:31<01:14,  2.92it/s]
Training 1/1 epoch (loss 3.0654):  65%|██████▌   | 409/625 [02:31<01:14,  2.92it/s]
Training 1/1 epoch (loss 3.0654):  66%|██████▌   | 410/625 [02:31<01:10,  3.04it/s]
Training 1/1 epoch (loss 2.4783):  66%|██████▌   | 410/625 [02:32<01:10,  3.04it/s]
Training 1/1 epoch (loss 2.4783):  66%|██████▌   | 411/625 [02:32<01:11,  3.01it/s]
Training 1/1 epoch (loss 2.8513):  66%|██████▌   | 411/625 [02:32<01:11,  3.01it/s]
Training 1/1 epoch (loss 2.8513):  66%|██████▌   | 412/625 [02:32<01:12,  2.92it/s]
Training 1/1 epoch (loss 2.8908):  66%|██████▌   | 412/625 [02:33<01:12,  2.92it/s]
Training 1/1 epoch (loss 2.8908):  66%|██████▌   | 413/625 [02:33<01:11,  2.97it/s]
Training 1/1 epoch (loss 2.9881):  66%|██████▌   | 413/625 [02:33<01:11,  2.97it/s]
Training 1/1 epoch (loss 2.9881):  66%|██████▌   | 414/625 [02:33<01:09,  3.04it/s]
Training 1/1 epoch (loss 2.7992):  66%|██████▌   | 414/625 [02:33<01:09,  3.04it/s]
Training 1/1 epoch (loss 2.7992):  66%|██████▋   | 415/625 [02:33<01:10,  2.99it/s]
Training 1/1 epoch (loss 2.5900):  66%|██████▋   | 415/625 [02:33<01:10,  2.99it/s]
Training 1/1 epoch (loss 2.5900):  67%|██████▋   | 416/625 [02:33<01:09,  3.02it/s]
Training 1/1 epoch (loss 3.0126):  67%|██████▋   | 416/625 [02:34<01:09,  3.02it/s]
Training 1/1 epoch (loss 3.0126):  67%|██████▋   | 417/625 [02:34<01:11,  2.92it/s]
Training 1/1 epoch (loss 2.8174):  67%|██████▋   | 417/625 [02:34<01:11,  2.92it/s]
Training 1/1 epoch (loss 2.8174):  67%|██████▋   | 418/625 [02:34<01:11,  2.90it/s]
Training 1/1 epoch (loss 2.9031):  67%|██████▋   | 418/625 [02:35<01:11,  2.90it/s]
Training 1/1 epoch (loss 2.9031):  67%|██████▋   | 419/625 [02:35<01:11,  2.88it/s]
Training 1/1 epoch (loss 2.8303):  67%|██████▋   | 419/625 [02:35<01:11,  2.88it/s]
Training 1/1 epoch (loss 2.8303):  67%|██████▋   | 420/625 [02:35<01:15,  2.73it/s]
Training 1/1 epoch (loss 2.6609):  67%|██████▋   | 420/625 [02:35<01:15,  2.73it/s]
Training 1/1 epoch (loss 2.6609):  67%|██████▋   | 421/625 [02:35<01:12,  2.83it/s]
Training 1/1 epoch (loss 2.7924):  67%|██████▋   | 421/625 [02:36<01:12,  2.83it/s]
Training 1/1 epoch (loss 2.7924):  68%|██████▊   | 422/625 [02:36<01:09,  2.93it/s]
Training 1/1 epoch (loss 2.8266):  68%|██████▊   | 422/625 [02:36<01:09,  2.93it/s]
Training 1/1 epoch (loss 2.8266):  68%|██████▊   | 423/625 [02:36<01:09,  2.90it/s]
Training 1/1 epoch (loss 3.0523):  68%|██████▊   | 423/625 [02:36<01:09,  2.90it/s]
Training 1/1 epoch (loss 3.0523):  68%|██████▊   | 424/625 [02:36<01:07,  2.97it/s]
Training 1/1 epoch (loss 2.8294):  68%|██████▊   | 424/625 [02:37<01:07,  2.97it/s]
Training 1/1 epoch (loss 2.8294):  68%|██████▊   | 425/625 [02:37<01:10,  2.85it/s]
Training 1/1 epoch (loss 2.8115):  68%|██████▊   | 425/625 [02:37<01:10,  2.85it/s]
Training 1/1 epoch (loss 2.8115):  68%|██████▊   | 426/625 [02:37<01:07,  2.95it/s]
Training 1/1 epoch (loss 2.8019):  68%|██████▊   | 426/625 [02:37<01:07,  2.95it/s]
Training 1/1 epoch (loss 2.8019):  68%|██████▊   | 427/625 [02:37<01:07,  2.94it/s]
Training 1/1 epoch (loss 2.7163):  68%|██████▊   | 427/625 [02:38<01:07,  2.94it/s]
Training 1/1 epoch (loss 2.7163):  68%|██████▊   | 428/625 [02:38<01:04,  3.05it/s]
Training 1/1 epoch (loss 2.8808):  68%|██████▊   | 428/625 [02:38<01:04,  3.05it/s]
Training 1/1 epoch (loss 2.8808):  69%|██████▊   | 429/625 [02:38<01:05,  3.01it/s]
Training 1/1 epoch (loss 3.1122):  69%|██████▊   | 429/625 [02:38<01:05,  3.01it/s]
Training 1/1 epoch (loss 3.1122):  69%|██████▉   | 430/625 [02:38<01:06,  2.95it/s]
Training 1/1 epoch (loss 2.7987):  69%|██████▉   | 430/625 [02:39<01:06,  2.95it/s]
Training 1/1 epoch (loss 2.7987):  69%|██████▉   | 431/625 [02:39<01:08,  2.85it/s]
Training 1/1 epoch (loss 2.4683):  69%|██████▉   | 431/625 [02:39<01:08,  2.85it/s]
Training 1/1 epoch (loss 2.4683):  69%|██████▉   | 432/625 [02:39<01:07,  2.87it/s]
Training 1/1 epoch (loss 2.4847):  69%|██████▉   | 432/625 [02:39<01:07,  2.87it/s]
Training 1/1 epoch (loss 2.4847):  69%|██████▉   | 433/625 [02:39<01:05,  2.92it/s]
Training 1/1 epoch (loss 2.7610):  69%|██████▉   | 433/625 [02:40<01:05,  2.92it/s]
Training 1/1 epoch (loss 2.7610):  69%|██████▉   | 434/625 [02:40<01:07,  2.82it/s]
Training 1/1 epoch (loss 2.9591):  69%|██████▉   | 434/625 [02:40<01:07,  2.82it/s]
Training 1/1 epoch (loss 2.9591):  70%|██████▉   | 435/625 [02:40<01:11,  2.67it/s]
Training 1/1 epoch (loss 2.8945):  70%|██████▉   | 435/625 [02:41<01:11,  2.67it/s]
Training 1/1 epoch (loss 2.8945):  70%|██████▉   | 436/625 [02:41<01:09,  2.71it/s]
Training 1/1 epoch (loss 2.7559):  70%|██████▉   | 436/625 [02:41<01:09,  2.71it/s]
Training 1/1 epoch (loss 2.7559):  70%|██████▉   | 437/625 [02:41<01:07,  2.77it/s]
Training 1/1 epoch (loss 3.1034):  70%|██████▉   | 437/625 [02:41<01:07,  2.77it/s]
Training 1/1 epoch (loss 3.1034):  70%|███████   | 438/625 [02:41<01:06,  2.82it/s]
Training 1/1 epoch (loss 2.8478):  70%|███████   | 438/625 [02:42<01:06,  2.82it/s]
Training 1/1 epoch (loss 2.8478):  70%|███████   | 439/625 [02:42<01:03,  2.94it/s]
Training 1/1 epoch (loss 2.9272):  70%|███████   | 439/625 [02:42<01:03,  2.94it/s]
Training 1/1 epoch (loss 2.9272):  70%|███████   | 440/625 [02:42<01:07,  2.74it/s]
Training 1/1 epoch (loss 2.7873):  70%|███████   | 440/625 [02:42<01:07,  2.74it/s]
Training 1/1 epoch (loss 2.7873):  71%|███████   | 441/625 [02:42<01:10,  2.61it/s]
Training 1/1 epoch (loss 2.9090):  71%|███████   | 441/625 [02:43<01:10,  2.61it/s]
Training 1/1 epoch (loss 2.9090):  71%|███████   | 442/625 [02:43<01:08,  2.66it/s]
Training 1/1 epoch (loss 2.7945):  71%|███████   | 442/625 [02:43<01:08,  2.66it/s]
Training 1/1 epoch (loss 2.7945):  71%|███████   | 443/625 [02:43<01:05,  2.76it/s]
Training 1/1 epoch (loss 2.7390):  71%|███████   | 443/625 [02:43<01:05,  2.76it/s]
Training 1/1 epoch (loss 2.7390):  71%|███████   | 444/625 [02:43<01:06,  2.72it/s]
Training 1/1 epoch (loss 2.7445):  71%|███████   | 444/625 [02:44<01:06,  2.72it/s]
Training 1/1 epoch (loss 2.7445):  71%|███████   | 445/625 [02:44<01:06,  2.72it/s]
Training 1/1 epoch (loss 2.9483):  71%|███████   | 445/625 [02:44<01:06,  2.72it/s]
Training 1/1 epoch (loss 2.9483):  71%|███████▏  | 446/625 [02:44<01:07,  2.66it/s]
Training 1/1 epoch (loss 2.9519):  71%|███████▏  | 446/625 [02:45<01:07,  2.66it/s]
Training 1/1 epoch (loss 2.9519):  72%|███████▏  | 447/625 [02:45<01:10,  2.53it/s]
Training 1/1 epoch (loss 2.7765):  72%|███████▏  | 447/625 [02:45<01:10,  2.53it/s]
Training 1/1 epoch (loss 2.7765):  72%|███████▏  | 448/625 [02:45<01:10,  2.50it/s]
Training 1/1 epoch (loss 2.9640):  72%|███████▏  | 448/625 [02:45<01:10,  2.50it/s]
Training 1/1 epoch (loss 2.9640):  72%|███████▏  | 449/625 [02:45<01:10,  2.49it/s]
Training 1/1 epoch (loss 2.7502):  72%|███████▏  | 449/625 [02:46<01:10,  2.49it/s]
Training 1/1 epoch (loss 2.7502):  72%|███████▏  | 450/625 [02:46<01:07,  2.60it/s]
Training 1/1 epoch (loss 2.6374):  72%|███████▏  | 450/625 [02:46<01:07,  2.60it/s]
Training 1/1 epoch (loss 2.6374):  72%|███████▏  | 451/625 [02:46<01:06,  2.62it/s]
Training 1/1 epoch (loss 2.7334):  72%|███████▏  | 451/625 [02:46<01:06,  2.62it/s]
Training 1/1 epoch (loss 2.7334):  72%|███████▏  | 452/625 [02:46<01:02,  2.77it/s]
Training 1/1 epoch (loss 2.6804):  72%|███████▏  | 452/625 [02:47<01:02,  2.77it/s]
Training 1/1 epoch (loss 2.6804):  72%|███████▏  | 453/625 [02:47<01:00,  2.84it/s]
Training 1/1 epoch (loss 2.6186):  72%|███████▏  | 453/625 [02:47<01:00,  2.84it/s]
Training 1/1 epoch (loss 2.6186):  73%|███████▎  | 454/625 [02:47<01:00,  2.83it/s]
Training 1/1 epoch (loss 2.9206):  73%|███████▎  | 454/625 [02:48<01:00,  2.83it/s]
Training 1/1 epoch (loss 2.9206):  73%|███████▎  | 455/625 [02:48<01:00,  2.83it/s]
Training 1/1 epoch (loss 2.8221):  73%|███████▎  | 455/625 [02:48<01:00,  2.83it/s]
Training 1/1 epoch (loss 2.8221):  73%|███████▎  | 456/625 [02:48<00:59,  2.86it/s]
Training 1/1 epoch (loss 2.9662):  73%|███████▎  | 456/625 [02:48<00:59,  2.86it/s]
Training 1/1 epoch (loss 2.9662):  73%|███████▎  | 457/625 [02:48<01:03,  2.64it/s]
Training 1/1 epoch (loss 2.8639):  73%|███████▎  | 457/625 [02:49<01:03,  2.64it/s]
Training 1/1 epoch (loss 2.8639):  73%|███████▎  | 458/625 [02:49<01:03,  2.62it/s]
Training 1/1 epoch (loss 2.7692):  73%|███████▎  | 458/625 [02:49<01:03,  2.62it/s]
Training 1/1 epoch (loss 2.7692):  73%|███████▎  | 459/625 [02:49<01:01,  2.70it/s]
Training 1/1 epoch (loss 2.5410):  73%|███████▎  | 459/625 [02:49<01:01,  2.70it/s]
Training 1/1 epoch (loss 2.5410):  74%|███████▎  | 460/625 [02:49<00:58,  2.80it/s]
Training 1/1 epoch (loss 3.0195):  74%|███████▎  | 460/625 [02:50<00:58,  2.80it/s]
Training 1/1 epoch (loss 3.0195):  74%|███████▍  | 461/625 [02:50<00:58,  2.81it/s]
Training 1/1 epoch (loss 2.9179):  74%|███████▍  | 461/625 [02:50<00:58,  2.81it/s]
Training 1/1 epoch (loss 2.9179):  74%|███████▍  | 462/625 [02:50<00:58,  2.81it/s]
Training 1/1 epoch (loss 2.7612):  74%|███████▍  | 462/625 [02:50<00:58,  2.81it/s]
Training 1/1 epoch (loss 2.7612):  74%|███████▍  | 463/625 [02:50<00:58,  2.76it/s]
Training 1/1 epoch (loss 2.9953):  74%|███████▍  | 463/625 [02:51<00:58,  2.76it/s]
Training 1/1 epoch (loss 2.9953):  74%|███████▍  | 464/625 [02:51<00:58,  2.73it/s]
Training 1/1 epoch (loss 2.8104):  74%|███████▍  | 464/625 [02:51<00:58,  2.73it/s]
Training 1/1 epoch (loss 2.8104):  74%|███████▍  | 465/625 [02:51<00:55,  2.87it/s]
Training 1/1 epoch (loss 2.9536):  74%|███████▍  | 465/625 [02:51<00:55,  2.87it/s]
Training 1/1 epoch (loss 2.9536):  75%|███████▍  | 466/625 [02:51<00:54,  2.91it/s]
Training 1/1 epoch (loss 2.7333):  75%|███████▍  | 466/625 [02:52<00:54,  2.91it/s]
Training 1/1 epoch (loss 2.7333):  75%|███████▍  | 467/625 [02:52<00:53,  2.96it/s]
Training 1/1 epoch (loss 2.7844):  75%|███████▍  | 467/625 [02:52<00:53,  2.96it/s]
Training 1/1 epoch (loss 2.7844):  75%|███████▍  | 468/625 [02:52<00:53,  2.94it/s]
Training 1/1 epoch (loss 3.0574):  75%|███████▍  | 468/625 [02:53<00:53,  2.94it/s]
Training 1/1 epoch (loss 3.0574):  75%|███████▌  | 469/625 [02:53<00:55,  2.82it/s]
Training 1/1 epoch (loss 2.7936):  75%|███████▌  | 469/625 [02:53<00:55,  2.82it/s]
Training 1/1 epoch (loss 2.7936):  75%|███████▌  | 470/625 [02:53<00:55,  2.77it/s]
Training 1/1 epoch (loss 2.7412):  75%|███████▌  | 470/625 [02:53<00:55,  2.77it/s]
Training 1/1 epoch (loss 2.7412):  75%|███████▌  | 471/625 [02:53<01:00,  2.53it/s]
Training 1/1 epoch (loss 3.0290):  75%|███████▌  | 471/625 [02:54<01:00,  2.53it/s]
Training 1/1 epoch (loss 3.0290):  76%|███████▌  | 472/625 [02:54<01:05,  2.35it/s]
Training 1/1 epoch (loss 2.5504):  76%|███████▌  | 472/625 [02:54<01:05,  2.35it/s]
Training 1/1 epoch (loss 2.5504):  76%|███████▌  | 473/625 [02:54<01:02,  2.45it/s]
Training 1/1 epoch (loss 2.9463):  76%|███████▌  | 473/625 [02:55<01:02,  2.45it/s]
Training 1/1 epoch (loss 2.9463):  76%|███████▌  | 474/625 [02:55<00:59,  2.56it/s]
Training 1/1 epoch (loss 3.0401):  76%|███████▌  | 474/625 [02:55<00:59,  2.56it/s]
Training 1/1 epoch (loss 3.0401):  76%|███████▌  | 475/625 [02:55<00:59,  2.51it/s]
Training 1/1 epoch (loss 3.0732):  76%|███████▌  | 475/625 [02:55<00:59,  2.51it/s]
Training 1/1 epoch (loss 3.0732):  76%|███████▌  | 476/625 [02:55<00:56,  2.65it/s]
Training 1/1 epoch (loss 2.8242):  76%|███████▌  | 476/625 [02:56<00:56,  2.65it/s]
Training 1/1 epoch (loss 2.8242):  76%|███████▋  | 477/625 [02:56<00:53,  2.75it/s]
Training 1/1 epoch (loss 2.7439):  76%|███████▋  | 477/625 [02:56<00:53,  2.75it/s]
Training 1/1 epoch (loss 2.7439):  76%|███████▋  | 478/625 [02:56<00:55,  2.66it/s]
Training 1/1 epoch (loss 2.7988):  76%|███████▋  | 478/625 [02:56<00:55,  2.66it/s]
Training 1/1 epoch (loss 2.7988):  77%|███████▋  | 479/625 [02:56<00:52,  2.78it/s]
Training 1/1 epoch (loss 2.8290):  77%|███████▋  | 479/625 [02:57<00:52,  2.78it/s]
Training 1/1 epoch (loss 2.8290):  77%|███████▋  | 480/625 [02:57<00:51,  2.80it/s]
Training 1/1 epoch (loss 2.8109):  77%|███████▋  | 480/625 [02:57<00:51,  2.80it/s]
Training 1/1 epoch (loss 2.8109):  77%|███████▋  | 481/625 [02:57<00:51,  2.78it/s]
Training 1/1 epoch (loss 2.5858):  77%|███████▋  | 481/625 [02:57<00:51,  2.78it/s]
Training 1/1 epoch (loss 2.5858):  77%|███████▋  | 482/625 [02:57<00:50,  2.85it/s]
Training 1/1 epoch (loss 2.7331):  77%|███████▋  | 482/625 [02:58<00:50,  2.85it/s]
Training 1/1 epoch (loss 2.7331):  77%|███████▋  | 483/625 [02:58<00:48,  2.94it/s]
Training 1/1 epoch (loss 2.7650):  77%|███████▋  | 483/625 [02:58<00:48,  2.94it/s]
Training 1/1 epoch (loss 2.7650):  77%|███████▋  | 484/625 [02:58<00:48,  2.90it/s]
Training 1/1 epoch (loss 2.7569):  77%|███████▋  | 484/625 [02:58<00:48,  2.90it/s]
Training 1/1 epoch (loss 2.7569):  78%|███████▊  | 485/625 [02:58<00:48,  2.88it/s]
Training 1/1 epoch (loss 2.7927):  78%|███████▊  | 485/625 [02:59<00:48,  2.88it/s]
Training 1/1 epoch (loss 2.7927):  78%|███████▊  | 486/625 [02:59<00:49,  2.78it/s]
Training 1/1 epoch (loss 2.8837):  78%|███████▊  | 486/625 [02:59<00:49,  2.78it/s]
Training 1/1 epoch (loss 2.8837):  78%|███████▊  | 487/625 [02:59<00:47,  2.89it/s]
Training 1/1 epoch (loss 2.7278):  78%|███████▊  | 487/625 [03:00<00:47,  2.89it/s]
Training 1/1 epoch (loss 2.7278):  78%|███████▊  | 488/625 [03:00<00:46,  2.92it/s]
Training 1/1 epoch (loss 2.7559):  78%|███████▊  | 488/625 [03:00<00:46,  2.92it/s]
Training 1/1 epoch (loss 2.7559):  78%|███████▊  | 489/625 [03:00<00:46,  2.95it/s]
Training 1/1 epoch (loss 2.8310):  78%|███████▊  | 489/625 [03:00<00:46,  2.95it/s]
Training 1/1 epoch (loss 2.8310):  78%|███████▊  | 490/625 [03:00<00:47,  2.86it/s]
Training 1/1 epoch (loss 2.8122):  78%|███████▊  | 490/625 [03:01<00:47,  2.86it/s]
Training 1/1 epoch (loss 2.8122):  79%|███████▊  | 491/625 [03:01<00:47,  2.85it/s]
Training 1/1 epoch (loss 2.6936):  79%|███████▊  | 491/625 [03:01<00:47,  2.85it/s]
Training 1/1 epoch (loss 2.6936):  79%|███████▊  | 492/625 [03:01<00:46,  2.84it/s]
Training 1/1 epoch (loss 2.7491):  79%|███████▊  | 492/625 [03:01<00:46,  2.84it/s]
Training 1/1 epoch (loss 2.7491):  79%|███████▉  | 493/625 [03:01<00:46,  2.84it/s]
Training 1/1 epoch (loss 2.5865):  79%|███████▉  | 493/625 [03:02<00:46,  2.84it/s]
Training 1/1 epoch (loss 2.5865):  79%|███████▉  | 494/625 [03:02<00:44,  2.95it/s]
Training 1/1 epoch (loss 2.5514):  79%|███████▉  | 494/625 [03:02<00:44,  2.95it/s]
Training 1/1 epoch (loss 2.5514):  79%|███████▉  | 495/625 [03:02<00:43,  3.01it/s]
Training 1/1 epoch (loss 2.7625):  79%|███████▉  | 495/625 [03:02<00:43,  3.01it/s]
Training 1/1 epoch (loss 2.7625):  79%|███████▉  | 496/625 [03:02<00:44,  2.91it/s]
Training 1/1 epoch (loss 2.8739):  79%|███████▉  | 496/625 [03:03<00:44,  2.91it/s]
Training 1/1 epoch (loss 2.8739):  80%|███████▉  | 497/625 [03:03<00:46,  2.78it/s]
Training 1/1 epoch (loss 2.7449):  80%|███████▉  | 497/625 [03:03<00:46,  2.78it/s]
Training 1/1 epoch (loss 2.7449):  80%|███████▉  | 498/625 [03:03<00:44,  2.83it/s]
Training 1/1 epoch (loss 2.9907):  80%|███████▉  | 498/625 [03:03<00:44,  2.83it/s]
Training 1/1 epoch (loss 2.9907):  80%|███████▉  | 499/625 [03:03<00:42,  2.95it/s]
Training 1/1 epoch (loss 2.7084):  80%|███████▉  | 499/625 [03:04<00:42,  2.95it/s]
Training 1/1 epoch (loss 2.7084):  80%|████████  | 500/625 [03:04<00:41,  3.04it/s]
Training 1/1 epoch (loss 2.7993):  80%|████████  | 500/625 [03:04<00:41,  3.04it/s]
Training 1/1 epoch (loss 2.7993):  80%|████████  | 501/625 [03:04<00:43,  2.88it/s]
Training 1/1 epoch (loss 2.7875):  80%|████████  | 501/625 [03:04<00:43,  2.88it/s]
Training 1/1 epoch (loss 2.7875):  80%|████████  | 502/625 [03:04<00:42,  2.89it/s]
Training 1/1 epoch (loss 2.7844):  80%|████████  | 502/625 [03:05<00:42,  2.89it/s]
Training 1/1 epoch (loss 2.7844):  80%|████████  | 503/625 [03:05<00:45,  2.67it/s]
Training 1/1 epoch (loss 2.9316):  80%|████████  | 503/625 [03:05<00:45,  2.67it/s]
Training 1/1 epoch (loss 2.9316):  81%|████████  | 504/625 [03:05<00:46,  2.61it/s]
Training 1/1 epoch (loss 2.8487):  81%|████████  | 504/625 [03:06<00:46,  2.61it/s]
Training 1/1 epoch (loss 2.8487):  81%|████████  | 505/625 [03:06<00:45,  2.66it/s]
Training 1/1 epoch (loss 2.9062):  81%|████████  | 505/625 [03:06<00:45,  2.66it/s]
Training 1/1 epoch (loss 2.9062):  81%|████████  | 506/625 [03:06<00:43,  2.77it/s]
Training 1/1 epoch (loss 2.6058):  81%|████████  | 506/625 [03:06<00:43,  2.77it/s]
Training 1/1 epoch (loss 2.6058):  81%|████████  | 507/625 [03:06<00:40,  2.90it/s]
Training 1/1 epoch (loss 2.6438):  81%|████████  | 507/625 [03:07<00:40,  2.90it/s]
Training 1/1 epoch (loss 2.6438):  81%|████████▏ | 508/625 [03:07<00:41,  2.85it/s]
Training 1/1 epoch (loss 2.7247):  81%|████████▏ | 508/625 [03:07<00:41,  2.85it/s]
Training 1/1 epoch (loss 2.7247):  81%|████████▏ | 509/625 [03:07<00:40,  2.88it/s]
Training 1/1 epoch (loss 2.7595):  81%|████████▏ | 509/625 [03:07<00:40,  2.88it/s]
Training 1/1 epoch (loss 2.7595):  82%|████████▏ | 510/625 [03:07<00:39,  2.93it/s]
Training 1/1 epoch (loss 2.9722):  82%|████████▏ | 510/625 [03:08<00:39,  2.93it/s]
Training 1/1 epoch (loss 2.9722):  82%|████████▏ | 511/625 [03:08<00:37,  3.01it/s]
Training 1/1 epoch (loss 2.5620):  82%|████████▏ | 511/625 [03:08<00:37,  3.01it/s]
Training 1/1 epoch (loss 2.5620):  82%|████████▏ | 512/625 [03:08<00:37,  3.02it/s]
Training 1/1 epoch (loss 2.5723):  82%|████████▏ | 512/625 [03:08<00:37,  3.02it/s]
Training 1/1 epoch (loss 2.5723):  82%|████████▏ | 513/625 [03:08<00:39,  2.86it/s]
Training 1/1 epoch (loss 2.7609):  82%|████████▏ | 513/625 [03:09<00:39,  2.86it/s]
Training 1/1 epoch (loss 2.7609):  82%|████████▏ | 514/625 [03:09<00:38,  2.88it/s]
Training 1/1 epoch (loss 2.7482):  82%|████████▏ | 514/625 [03:09<00:38,  2.88it/s]
Training 1/1 epoch (loss 2.7482):  82%|████████▏ | 515/625 [03:09<00:39,  2.82it/s]
Training 1/1 epoch (loss 2.9090):  82%|████████▏ | 515/625 [03:09<00:39,  2.82it/s]
Training 1/1 epoch (loss 2.9090):  83%|████████▎ | 516/625 [03:09<00:37,  2.91it/s]
Training 1/1 epoch (loss 2.7319):  83%|████████▎ | 516/625 [03:10<00:37,  2.91it/s]
Training 1/1 epoch (loss 2.7319):  83%|████████▎ | 517/625 [03:10<00:35,  3.06it/s]
Training 1/1 epoch (loss 2.9256):  83%|████████▎ | 517/625 [03:10<00:35,  3.06it/s]
Training 1/1 epoch (loss 2.9256):  83%|████████▎ | 518/625 [03:10<00:36,  2.96it/s]
Training 1/1 epoch (loss 2.9337):  83%|████████▎ | 518/625 [03:10<00:36,  2.96it/s]
Training 1/1 epoch (loss 2.9337):  83%|████████▎ | 519/625 [03:10<00:36,  2.91it/s]
Training 1/1 epoch (loss 2.9572):  83%|████████▎ | 519/625 [03:11<00:36,  2.91it/s]
Training 1/1 epoch (loss 2.9572):  83%|████████▎ | 520/625 [03:11<00:39,  2.66it/s]
Training 1/1 epoch (loss 2.9138):  83%|████████▎ | 520/625 [03:11<00:39,  2.66it/s]
Training 1/1 epoch (loss 2.9138):  83%|████████▎ | 521/625 [03:11<00:38,  2.67it/s]
Training 1/1 epoch (loss 2.7466):  83%|████████▎ | 521/625 [03:11<00:38,  2.67it/s]
Training 1/1 epoch (loss 2.7466):  84%|████████▎ | 522/625 [03:11<00:36,  2.79it/s]
Training 1/1 epoch (loss 2.8385):  84%|████████▎ | 522/625 [03:12<00:36,  2.79it/s]
Training 1/1 epoch (loss 2.8385):  84%|████████▎ | 523/625 [03:12<00:35,  2.86it/s]
Training 1/1 epoch (loss 2.7083):  84%|████████▎ | 523/625 [03:12<00:35,  2.86it/s]
Training 1/1 epoch (loss 2.7083):  84%|████████▍ | 524/625 [03:12<00:34,  2.94it/s]
Training 1/1 epoch (loss 2.8809):  84%|████████▍ | 524/625 [03:12<00:34,  2.94it/s]
Training 1/1 epoch (loss 2.8809):  84%|████████▍ | 525/625 [03:12<00:32,  3.07it/s]
Training 1/1 epoch (loss 3.0197):  84%|████████▍ | 525/625 [03:13<00:32,  3.07it/s]
Training 1/1 epoch (loss 3.0197):  84%|████████▍ | 526/625 [03:13<00:33,  2.95it/s]
Training 1/1 epoch (loss 2.8566):  84%|████████▍ | 526/625 [03:13<00:33,  2.95it/s]
Training 1/1 epoch (loss 2.8566):  84%|████████▍ | 527/625 [03:13<00:33,  2.91it/s]
Training 1/1 epoch (loss 2.7612):  84%|████████▍ | 527/625 [03:13<00:33,  2.91it/s]
Training 1/1 epoch (loss 2.7612):  84%|████████▍ | 528/625 [03:13<00:32,  2.96it/s]
Training 1/1 epoch (loss 2.7683):  84%|████████▍ | 528/625 [03:14<00:32,  2.96it/s]
Training 1/1 epoch (loss 2.7683):  85%|████████▍ | 529/625 [03:14<00:32,  2.92it/s]
Training 1/1 epoch (loss 2.7322):  85%|████████▍ | 529/625 [03:14<00:32,  2.92it/s]
Training 1/1 epoch (loss 2.7322):  85%|████████▍ | 530/625 [03:14<00:33,  2.86it/s]
Training 1/1 epoch (loss 2.9428):  85%|████████▍ | 530/625 [03:14<00:33,  2.86it/s]
Training 1/1 epoch (loss 2.9428):  85%|████████▍ | 531/625 [03:14<00:31,  2.95it/s]
Training 1/1 epoch (loss 2.7373):  85%|████████▍ | 531/625 [03:15<00:31,  2.95it/s]
Training 1/1 epoch (loss 2.7373):  85%|████████▌ | 532/625 [03:15<00:32,  2.82it/s]
Training 1/1 epoch (loss 3.0109):  85%|████████▌ | 532/625 [03:15<00:32,  2.82it/s]
Training 1/1 epoch (loss 3.0109):  85%|████████▌ | 533/625 [03:15<00:31,  2.89it/s]
Training 1/1 epoch (loss 2.9986):  85%|████████▌ | 533/625 [03:15<00:31,  2.89it/s]
Training 1/1 epoch (loss 2.9986):  85%|████████▌ | 534/625 [03:15<00:31,  2.93it/s]
Training 1/1 epoch (loss 2.6860):  85%|████████▌ | 534/625 [03:16<00:31,  2.93it/s]
Training 1/1 epoch (loss 2.6860):  86%|████████▌ | 535/625 [03:16<00:30,  2.98it/s]
Training 1/1 epoch (loss 2.8362):  86%|████████▌ | 535/625 [03:16<00:30,  2.98it/s]
Training 1/1 epoch (loss 2.8362):  86%|████████▌ | 536/625 [03:16<00:30,  2.90it/s]
Training 1/1 epoch (loss 2.9195):  86%|████████▌ | 536/625 [03:17<00:30,  2.90it/s]
Training 1/1 epoch (loss 2.9195):  86%|████████▌ | 537/625 [03:17<00:31,  2.81it/s]
Training 1/1 epoch (loss 2.7081):  86%|████████▌ | 537/625 [03:17<00:31,  2.81it/s]
Training 1/1 epoch (loss 2.7081):  86%|████████▌ | 538/625 [03:17<00:32,  2.66it/s]
Training 1/1 epoch (loss 2.7439):  86%|████████▌ | 538/625 [03:17<00:32,  2.66it/s]
Training 1/1 epoch (loss 2.7439):  86%|████████▌ | 539/625 [03:17<00:31,  2.77it/s]
Training 1/1 epoch (loss 2.6518):  86%|████████▌ | 539/625 [03:18<00:31,  2.77it/s]
Training 1/1 epoch (loss 2.6518):  86%|████████▋ | 540/625 [03:18<00:29,  2.88it/s]
Training 1/1 epoch (loss 2.6319):  86%|████████▋ | 540/625 [03:18<00:29,  2.88it/s]
Training 1/1 epoch (loss 2.6319):  87%|████████▋ | 541/625 [03:18<00:29,  2.86it/s]
Training 1/1 epoch (loss 2.7472):  87%|████████▋ | 541/625 [03:18<00:29,  2.86it/s]
Training 1/1 epoch (loss 2.7472):  87%|████████▋ | 542/625 [03:18<00:28,  2.88it/s]
Training 1/1 epoch (loss 2.6845):  87%|████████▋ | 542/625 [03:19<00:28,  2.88it/s]
Training 1/1 epoch (loss 2.6845):  87%|████████▋ | 543/625 [03:19<00:28,  2.85it/s]
Training 1/1 epoch (loss 2.8964):  87%|████████▋ | 543/625 [03:19<00:28,  2.85it/s]
Training 1/1 epoch (loss 2.8964):  87%|████████▋ | 544/625 [03:19<00:29,  2.77it/s]
Training 1/1 epoch (loss 3.0570):  87%|████████▋ | 544/625 [03:19<00:29,  2.77it/s]
Training 1/1 epoch (loss 3.0570):  87%|████████▋ | 545/625 [03:19<00:27,  2.86it/s]
Training 1/1 epoch (loss 2.8007):  87%|████████▋ | 545/625 [03:20<00:27,  2.86it/s]
Training 1/1 epoch (loss 2.8007):  87%|████████▋ | 546/625 [03:20<00:26,  2.94it/s]
Training 1/1 epoch (loss 2.6792):  87%|████████▋ | 546/625 [03:20<00:26,  2.94it/s]
Training 1/1 epoch (loss 2.6792):  88%|████████▊ | 547/625 [03:20<00:26,  2.93it/s]
Training 1/1 epoch (loss 2.5320):  88%|████████▊ | 547/625 [03:20<00:26,  2.93it/s]
Training 1/1 epoch (loss 2.5320):  88%|████████▊ | 548/625 [03:20<00:25,  2.98it/s]
Training 1/1 epoch (loss 2.8052):  88%|████████▊ | 548/625 [03:21<00:25,  2.98it/s]
Training 1/1 epoch (loss 2.8052):  88%|████████▊ | 549/625 [03:21<00:26,  2.86it/s]
Training 1/1 epoch (loss 2.5935):  88%|████████▊ | 549/625 [03:21<00:26,  2.86it/s]
Training 1/1 epoch (loss 2.5935):  88%|████████▊ | 550/625 [03:21<00:25,  2.98it/s]
Training 1/1 epoch (loss 2.9607):  88%|████████▊ | 550/625 [03:21<00:25,  2.98it/s]
Training 1/1 epoch (loss 2.9607):  88%|████████▊ | 551/625 [03:21<00:24,  3.03it/s]
Training 1/1 epoch (loss 2.7292):  88%|████████▊ | 551/625 [03:22<00:24,  3.03it/s]
Training 1/1 epoch (loss 2.7292):  88%|████████▊ | 552/625 [03:22<00:24,  3.04it/s]
Training 1/1 epoch (loss 2.7166):  88%|████████▊ | 552/625 [03:22<00:24,  3.04it/s]
Training 1/1 epoch (loss 2.7166):  88%|████████▊ | 553/625 [03:22<00:25,  2.87it/s]
Training 1/1 epoch (loss 2.7814):  88%|████████▊ | 553/625 [03:22<00:25,  2.87it/s]
Training 1/1 epoch (loss 2.7814):  89%|████████▊ | 554/625 [03:22<00:24,  2.90it/s]
Training 1/1 epoch (loss 2.8241):  89%|████████▊ | 554/625 [03:23<00:24,  2.90it/s]
Training 1/1 epoch (loss 2.8241):  89%|████████▉ | 555/625 [03:23<00:24,  2.89it/s]
Training 1/1 epoch (loss 2.6266):  89%|████████▉ | 555/625 [03:23<00:24,  2.89it/s]
Training 1/1 epoch (loss 2.6266):  89%|████████▉ | 556/625 [03:23<00:28,  2.44it/s]
Training 1/1 epoch (loss 2.7476):  89%|████████▉ | 556/625 [03:24<00:28,  2.44it/s]
Training 1/1 epoch (loss 2.7476):  89%|████████▉ | 557/625 [03:24<00:28,  2.38it/s]
Training 1/1 epoch (loss 2.8370):  89%|████████▉ | 557/625 [03:24<00:28,  2.38it/s]
Training 1/1 epoch (loss 2.8370):  89%|████████▉ | 558/625 [03:24<00:26,  2.52it/s]
Training 1/1 epoch (loss 2.6017):  89%|████████▉ | 558/625 [03:24<00:26,  2.52it/s]
Training 1/1 epoch (loss 2.6017):  89%|████████▉ | 559/625 [03:24<00:25,  2.60it/s]
Training 1/1 epoch (loss 2.7902):  89%|████████▉ | 559/625 [03:25<00:25,  2.60it/s]
Training 1/1 epoch (loss 2.7902):  90%|████████▉ | 560/625 [03:25<00:25,  2.50it/s]
Training 1/1 epoch (loss 2.8408):  90%|████████▉ | 560/625 [03:25<00:25,  2.50it/s]
Training 1/1 epoch (loss 2.8408):  90%|████████▉ | 561/625 [03:25<00:27,  2.31it/s]
Training 1/1 epoch (loss 2.9371):  90%|████████▉ | 561/625 [03:26<00:27,  2.31it/s]
Training 1/1 epoch (loss 2.9371):  90%|████████▉ | 562/625 [03:26<00:24,  2.53it/s]
Training 1/1 epoch (loss 2.8057):  90%|████████▉ | 562/625 [03:26<00:24,  2.53it/s]
Training 1/1 epoch (loss 2.8057):  90%|█████████ | 563/625 [03:26<00:23,  2.67it/s]
Training 1/1 epoch (loss 2.5171):  90%|█████████ | 563/625 [03:26<00:23,  2.67it/s]
Training 1/1 epoch (loss 2.5171):  90%|█████████ | 564/625 [03:26<00:22,  2.72it/s]
Training 1/1 epoch (loss 2.7437):  90%|█████████ | 564/625 [03:27<00:22,  2.72it/s]
Training 1/1 epoch (loss 2.7437):  90%|█████████ | 565/625 [03:27<00:21,  2.75it/s]
Training 1/1 epoch (loss 2.8151):  90%|█████████ | 565/625 [03:27<00:21,  2.75it/s]
Training 1/1 epoch (loss 2.8151):  91%|█████████ | 566/625 [03:27<00:20,  2.82it/s]
Training 1/1 epoch (loss 2.6687):  91%|█████████ | 566/625 [03:28<00:20,  2.82it/s]
Training 1/1 epoch (loss 2.6687):  91%|█████████ | 567/625 [03:28<00:21,  2.70it/s]
Training 1/1 epoch (loss 2.7038):  91%|█████████ | 567/625 [03:28<00:21,  2.70it/s]
Training 1/1 epoch (loss 2.7038):  91%|█████████ | 568/625 [03:28<00:20,  2.73it/s]
Training 1/1 epoch (loss 2.9593):  91%|█████████ | 568/625 [03:28<00:20,  2.73it/s]
Training 1/1 epoch (loss 2.9593):  91%|█████████ | 569/625 [03:28<00:20,  2.76it/s]
Training 1/1 epoch (loss 2.9866):  91%|█████████ | 569/625 [03:29<00:20,  2.76it/s]
Training 1/1 epoch (loss 2.9866):  91%|█████████ | 570/625 [03:29<00:19,  2.80it/s]
Training 1/1 epoch (loss 2.9397):  91%|█████████ | 570/625 [03:29<00:19,  2.80it/s]
Training 1/1 epoch (loss 2.9397):  91%|█████████▏| 571/625 [03:29<00:19,  2.73it/s]
Training 1/1 epoch (loss 2.9702):  91%|█████████▏| 571/625 [03:29<00:19,  2.73it/s]
Training 1/1 epoch (loss 2.9702):  92%|█████████▏| 572/625 [03:29<00:19,  2.76it/s]
Training 1/1 epoch (loss 2.7411):  92%|█████████▏| 572/625 [03:30<00:19,  2.76it/s]
Training 1/1 epoch (loss 2.7411):  92%|█████████▏| 573/625 [03:30<00:18,  2.84it/s]
Training 1/1 epoch (loss 2.5706):  92%|█████████▏| 573/625 [03:30<00:18,  2.84it/s]
Training 1/1 epoch (loss 2.5706):  92%|█████████▏| 574/625 [03:30<00:18,  2.79it/s]
Training 1/1 epoch (loss 2.6168):  92%|█████████▏| 574/625 [03:30<00:18,  2.79it/s]
Training 1/1 epoch (loss 2.6168):  92%|█████████▏| 575/625 [03:30<00:17,  2.87it/s]
Training 1/1 epoch (loss 2.6007):  92%|█████████▏| 575/625 [03:31<00:17,  2.87it/s]
Training 1/1 epoch (loss 2.6007):  92%|█████████▏| 576/625 [03:31<00:17,  2.74it/s]
Training 1/1 epoch (loss 2.7978):  92%|█████████▏| 576/625 [03:31<00:17,  2.74it/s]
Training 1/1 epoch (loss 2.7978):  92%|█████████▏| 577/625 [03:31<00:18,  2.55it/s]
Training 1/1 epoch (loss 2.7620):  92%|█████████▏| 577/625 [03:32<00:18,  2.55it/s]
Training 1/1 epoch (loss 2.7620):  92%|█████████▏| 578/625 [03:32<00:17,  2.66it/s]
Training 1/1 epoch (loss 2.9856):  92%|█████████▏| 578/625 [03:32<00:17,  2.66it/s]
Training 1/1 epoch (loss 2.9856):  93%|█████████▎| 579/625 [03:32<00:16,  2.80it/s]
Training 1/1 epoch (loss 2.7237):  93%|█████████▎| 579/625 [03:32<00:16,  2.80it/s]
Training 1/1 epoch (loss 2.7237):  93%|█████████▎| 580/625 [03:32<00:15,  2.83it/s]
Training 1/1 epoch (loss 2.6852):  93%|█████████▎| 580/625 [03:33<00:15,  2.83it/s]
Training 1/1 epoch (loss 2.6852):  93%|█████████▎| 581/625 [03:33<00:15,  2.85it/s]
Training 1/1 epoch (loss 2.8118):  93%|█████████▎| 581/625 [03:33<00:15,  2.85it/s]
Training 1/1 epoch (loss 2.8118):  93%|█████████▎| 582/625 [03:33<00:15,  2.79it/s]
Training 1/1 epoch (loss 2.9108):  93%|█████████▎| 582/625 [03:33<00:15,  2.79it/s]
Training 1/1 epoch (loss 2.9108):  93%|█████████▎| 583/625 [03:33<00:14,  2.85it/s]
Training 1/1 epoch (loss 2.5352):  93%|█████████▎| 583/625 [03:34<00:14,  2.85it/s]
Training 1/1 epoch (loss 2.5352):  93%|█████████▎| 584/625 [03:34<00:14,  2.87it/s]
Training 1/1 epoch (loss 2.7558):  93%|█████████▎| 584/625 [03:34<00:14,  2.87it/s]
Training 1/1 epoch (loss 2.7558):  94%|█████████▎| 585/625 [03:34<00:13,  2.88it/s]
Training 1/1 epoch (loss 2.6458):  94%|█████████▎| 585/625 [03:34<00:13,  2.88it/s]
Training 1/1 epoch (loss 2.6458):  94%|█████████▍| 586/625 [03:34<00:13,  2.89it/s]
Training 1/1 epoch (loss 2.9458):  94%|█████████▍| 586/625 [03:35<00:13,  2.89it/s]
Training 1/1 epoch (loss 2.9458):  94%|█████████▍| 587/625 [03:35<00:12,  2.97it/s]
Training 1/1 epoch (loss 2.8020):  94%|█████████▍| 587/625 [03:35<00:12,  2.97it/s]
Training 1/1 epoch (loss 2.8020):  94%|█████████▍| 588/625 [03:35<00:12,  2.97it/s]
Training 1/1 epoch (loss 2.7818):  94%|█████████▍| 588/625 [03:35<00:12,  2.97it/s]
Training 1/1 epoch (loss 2.7818):  94%|█████████▍| 589/625 [03:35<00:12,  2.84it/s]
Training 1/1 epoch (loss 2.6753):  94%|█████████▍| 589/625 [03:36<00:12,  2.84it/s]
Training 1/1 epoch (loss 2.6753):  94%|█████████▍| 590/625 [03:36<00:12,  2.87it/s]
Training 1/1 epoch (loss 3.0763):  94%|█████████▍| 590/625 [03:36<00:12,  2.87it/s]
Training 1/1 epoch (loss 3.0763):  95%|█████████▍| 591/625 [03:36<00:11,  2.91it/s]
Training 1/1 epoch (loss 2.6754):  95%|█████████▍| 591/625 [03:36<00:11,  2.91it/s]
Training 1/1 epoch (loss 2.6754):  95%|█████████▍| 592/625 [03:36<00:11,  2.91it/s]
Training 1/1 epoch (loss 2.7792):  95%|█████████▍| 592/625 [03:37<00:11,  2.91it/s]
Training 1/1 epoch (loss 2.7792):  95%|█████████▍| 593/625 [03:37<00:11,  2.74it/s]
Training 1/1 epoch (loss 2.7998):  95%|█████████▍| 593/625 [03:37<00:11,  2.74it/s]
Training 1/1 epoch (loss 2.7998):  95%|█████████▌| 594/625 [03:37<00:10,  2.83it/s]
Training 1/1 epoch (loss 2.6130):  95%|█████████▌| 594/625 [03:37<00:10,  2.83it/s]
Training 1/1 epoch (loss 2.6130):  95%|█████████▌| 595/625 [03:37<00:10,  2.84it/s]
Training 1/1 epoch (loss 2.7066):  95%|█████████▌| 595/625 [03:38<00:10,  2.84it/s]
Training 1/1 epoch (loss 2.7066):  95%|█████████▌| 596/625 [03:38<00:10,  2.84it/s]
Training 1/1 epoch (loss 2.6338):  95%|█████████▌| 596/625 [03:38<00:10,  2.84it/s]
Training 1/1 epoch (loss 2.6338):  96%|█████████▌| 597/625 [03:38<00:09,  2.85it/s]
Training 1/1 epoch (loss 2.8189):  96%|█████████▌| 597/625 [03:38<00:09,  2.85it/s]
Training 1/1 epoch (loss 2.8189):  96%|█████████▌| 598/625 [03:38<00:09,  2.85it/s]
Training 1/1 epoch (loss 2.7980):  96%|█████████▌| 598/625 [03:39<00:09,  2.85it/s]
Training 1/1 epoch (loss 2.7980):  96%|█████████▌| 599/625 [03:39<00:08,  2.90it/s]
Training 1/1 epoch (loss 2.8033):  96%|█████████▌| 599/625 [03:39<00:08,  2.90it/s]
Training 1/1 epoch (loss 2.8033):  96%|█████████▌| 600/625 [03:39<00:08,  2.89it/s]
Training 1/1 epoch (loss 2.6637):  96%|█████████▌| 600/625 [03:40<00:08,  2.89it/s]
Training 1/1 epoch (loss 2.6637):  96%|█████████▌| 601/625 [03:40<00:08,  2.85it/s]
Training 1/1 epoch (loss 2.7906):  96%|█████████▌| 601/625 [03:40<00:08,  2.85it/s]
Training 1/1 epoch (loss 2.7906):  96%|█████████▋| 602/625 [03:40<00:07,  2.92it/s]
Training 1/1 epoch (loss 2.6611):  96%|█████████▋| 602/625 [03:40<00:07,  2.92it/s]
Training 1/1 epoch (loss 2.6611):  96%|█████████▋| 603/625 [03:40<00:07,  2.94it/s]
Training 1/1 epoch (loss 2.7570):  96%|█████████▋| 603/625 [03:40<00:07,  2.94it/s]
Training 1/1 epoch (loss 2.7570):  97%|█████████▋| 604/625 [03:40<00:06,  3.02it/s]
Training 1/1 epoch (loss 2.8464):  97%|█████████▋| 604/625 [03:41<00:06,  3.02it/s]
Training 1/1 epoch (loss 2.8464):  97%|█████████▋| 605/625 [03:41<00:06,  2.99it/s]
Training 1/1 epoch (loss 2.7590):  97%|█████████▋| 605/625 [03:41<00:06,  2.99it/s]
Training 1/1 epoch (loss 2.7590):  97%|█████████▋| 606/625 [03:41<00:06,  2.91it/s]
Training 1/1 epoch (loss 2.7083):  97%|█████████▋| 606/625 [03:41<00:06,  2.91it/s]
Training 1/1 epoch (loss 2.7083):  97%|█████████▋| 607/625 [03:41<00:05,  3.03it/s]
Training 1/1 epoch (loss 2.8852):  97%|█████████▋| 607/625 [03:42<00:05,  3.03it/s]
Training 1/1 epoch (loss 2.8852):  97%|█████████▋| 608/625 [03:42<00:05,  2.95it/s]
Training 1/1 epoch (loss 2.8266):  97%|█████████▋| 608/625 [03:42<00:05,  2.95it/s]
Training 1/1 epoch (loss 2.8266):  97%|█████████▋| 609/625 [03:42<00:05,  2.96it/s]
Training 1/1 epoch (loss 2.8352):  97%|█████████▋| 609/625 [03:43<00:05,  2.96it/s]
Training 1/1 epoch (loss 2.8352):  98%|█████████▊| 610/625 [03:43<00:05,  2.99it/s]
Training 1/1 epoch (loss 2.7206):  98%|█████████▊| 610/625 [03:43<00:05,  2.99it/s]
Training 1/1 epoch (loss 2.7206):  98%|█████████▊| 611/625 [03:43<00:04,  3.03it/s]
Training 1/1 epoch (loss 3.0205):  98%|█████████▊| 611/625 [03:43<00:04,  3.03it/s]
Training 1/1 epoch (loss 3.0205):  98%|█████████▊| 612/625 [03:43<00:04,  2.88it/s]
Training 1/1 epoch (loss 2.6336):  98%|█████████▊| 612/625 [03:44<00:04,  2.88it/s]
Training 1/1 epoch (loss 2.6336):  98%|█████████▊| 613/625 [03:44<00:04,  2.91it/s]
Training 1/1 epoch (loss 2.8791):  98%|█████████▊| 613/625 [03:44<00:04,  2.91it/s]
Training 1/1 epoch (loss 2.8791):  98%|█████████▊| 614/625 [03:44<00:03,  2.88it/s]
Training 1/1 epoch (loss 2.6587):  98%|█████████▊| 614/625 [03:44<00:03,  2.88it/s]
Training 1/1 epoch (loss 2.6587):  98%|█████████▊| 615/625 [03:44<00:03,  2.92it/s]
Training 1/1 epoch (loss 2.7368):  98%|█████████▊| 615/625 [03:45<00:03,  2.92it/s]
Training 1/1 epoch (loss 2.7368):  99%|█████████▊| 616/625 [03:45<00:03,  2.84it/s]
Training 1/1 epoch (loss 2.8773):  99%|█████████▊| 616/625 [03:45<00:03,  2.84it/s]
Training 1/1 epoch (loss 2.8773):  99%|█████████▊| 617/625 [03:45<00:02,  2.77it/s]
Training 1/1 epoch (loss 2.9389):  99%|█████████▊| 617/625 [03:45<00:02,  2.77it/s]
Training 1/1 epoch (loss 2.9389):  99%|█████████▉| 618/625 [03:45<00:02,  2.82it/s]
Training 1/1 epoch (loss 2.7756):  99%|█████████▉| 618/625 [03:46<00:02,  2.82it/s]
Training 1/1 epoch (loss 2.7756):  99%|█████████▉| 619/625 [03:46<00:02,  2.87it/s]
Training 1/1 epoch (loss 2.6288):  99%|█████████▉| 619/625 [03:46<00:02,  2.87it/s]
Training 1/1 epoch (loss 2.6288):  99%|█████████▉| 620/625 [03:46<00:01,  2.77it/s]
Training 1/1 epoch (loss 2.8780):  99%|█████████▉| 620/625 [03:46<00:01,  2.77it/s]
Training 1/1 epoch (loss 2.8780):  99%|█████████▉| 621/625 [03:46<00:01,  2.86it/s]
Training 1/1 epoch (loss 2.7361):  99%|█████████▉| 621/625 [03:47<00:01,  2.86it/s]
Training 1/1 epoch (loss 2.7361): 100%|█████████▉| 622/625 [03:47<00:01,  2.93it/s]
Training 1/1 epoch (loss 2.8506): 100%|█████████▉| 622/625 [03:47<00:01,  2.93it/s]
Training 1/1 epoch (loss 2.8506): 100%|█████████▉| 623/625 [03:47<00:00,  2.75it/s]
Training 1/1 epoch (loss 2.5905): 100%|█████████▉| 623/625 [03:48<00:00,  2.75it/s]
Training 1/1 epoch (loss 2.5905): 100%|█████████▉| 624/625 [03:48<00:00,  2.71it/s]
Training 1/1 epoch (loss 2.7995): 100%|█████████▉| 624/625 [03:48<00:00,  2.71it/s]
Training 1/1 epoch (loss 2.7995): 100%|██████████| 625/625 [03:48<00:00,  2.78it/s]
Training 1/1 epoch (loss 2.7995): 100%|██████████| 625/625 [03:48<00:00,  2.74it/s]
tokenizer config file saved in /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-0.5T/tinyllama-0.5T-s3-Q1-5000/tokenizer_config.json
Special tokens file saved in /aifs4su/hansirui_1st/jiayi/setting3-imdb/tinyllama-0.5T/tinyllama-0.5T-s3-Q1-5000/special_tokens_map.json
wandb: ERROR Problem finishing run
Exception ignored in atexit callback: <bound method rank_zero_only.<locals>.wrapper of <safe_rlhf.logger.Logger object at 0x15512c3bd750>>
Traceback (most recent call last):
  File "/home/hansirui_1st/jiayi/resist/setting3/safe_rlhf/utils.py", line 212, in wrapper
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File "/home/hansirui_1st/jiayi/resist/setting3/safe_rlhf/logger.py", line 183, in close
    self.wandb.finish()
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 406, in wrapper
    return func(self, *args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 503, in wrapper
    return func(self, *args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 451, in wrapper
    return func(self, *args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2309, in finish
    return self._finish(exit_code)
           ^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 406, in wrapper
    return func(self, *args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2337, in _finish
    self._atexit_cleanup(exit_code=exit_code)
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2550, in _atexit_cleanup
    self._on_finish()
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/wandb_run.py", line 2806, in _on_finish
    wait_with_progress(
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/mailbox/wait_with_progress.py", line 24, in wait_with_progress
    return wait_all_with_progress(
           ^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/mailbox/wait_with_progress.py", line 87, in wait_all_with_progress
    return asyncio_compat.run(progress_loop_with_timeout)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/site-packages/wandb/sdk/lib/asyncio_compat.py", line 27, in run
    future = executor.submit(runner.run, fn)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/aifs4su/hansirui_1st/miniconda3/envs/jy-resist/lib/python3.11/concurrent/futures/thread.py", line 169, in submit
    raise RuntimeError('cannot schedule new futures after '
RuntimeError: cannot schedule new futures after interpreter shutdown