05/16/2024 23:41:32 - INFO - transformers.tokenization_utils_base - loading file tokenizer.model from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/tokenizer.model 05/16/2024 23:41:32 - INFO - transformers.tokenization_utils_base - loading file tokenizer.json from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/tokenizer.json 05/16/2024 23:41:32 - INFO - transformers.tokenization_utils_base - loading file added_tokens.json from cache at None 05/16/2024 23:41:32 - INFO - transformers.tokenization_utils_base - loading file special_tokens_map.json from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/special_tokens_map.json 05/16/2024 23:41:32 - INFO - transformers.tokenization_utils_base - loading file tokenizer_config.json from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/tokenizer_config.json 05/16/2024 23:41:32 - INFO - llmtuner.data.template - Add pad token: 05/16/2024 23:41:32 - INFO - llmtuner.data.loader - Loading dataset MediaMeter/Chapterization... 05/16/2024 23:41:38 - INFO - transformers.configuration_utils - loading configuration file config.json from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/config.json 05/16/2024 23:41:38 - INFO - transformers.configuration_utils - Model config MistralConfig { "_name_or_path": "mistralai/Mistral-7B-Instruct-v0.2", "architectures": [ "MistralForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 1, "eos_token_id": 2, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 14336, "max_position_embeddings": 32768, "model_type": "mistral", "num_attention_heads": 32, "num_hidden_layers": 32, "num_key_value_heads": 8, "rms_norm_eps": 1e-05, "rope_theta": 1000000.0, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "bfloat16", "transformers_version": "4.40.2", "use_cache": true, "vocab_size": 32000 } 05/16/2024 23:41:38 - WARNING - llmtuner.model.utils.attention - FlashAttention-2 is not installed. 05/16/2024 23:41:38 - INFO - llmtuner.model.utils.quantization - Quantizing model to 4 bit. 05/16/2024 23:41:38 - INFO - transformers.modeling_utils - loading weights file model.safetensors from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/model.safetensors.index.json 05/16/2024 23:41:38 - INFO - transformers.modeling_utils - Instantiating MistralForCausalLM model under default dtype torch.bfloat16. 05/16/2024 23:41:38 - INFO - transformers.generation.configuration_utils - Generate config GenerationConfig { "bos_token_id": 1, "eos_token_id": 2 } 05/16/2024 23:41:41 - INFO - transformers.modeling_utils - All model checkpoint weights were used when initializing MistralForCausalLM. 05/16/2024 23:41:41 - INFO - transformers.modeling_utils - All the weights of MistralForCausalLM were initialized from the model checkpoint at mistralai/Mistral-7B-Instruct-v0.2. If your task is similar to the task the model of the checkpoint was trained on, you can already use MistralForCausalLM for predictions without further training. 05/16/2024 23:41:41 - INFO - transformers.generation.configuration_utils - loading configuration file generation_config.json from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/generation_config.json 05/16/2024 23:41:41 - INFO - transformers.generation.configuration_utils - Generate config GenerationConfig { "bos_token_id": 1, "eos_token_id": 2 } 05/16/2024 23:41:41 - INFO - llmtuner.model.utils.checkpointing - Gradient checkpointing enabled. 05/16/2024 23:41:41 - INFO - llmtuner.model.utils.attention - Using torch SDPA for faster training and inference. 05/16/2024 23:41:41 - INFO - llmtuner.model.adapter - Fine-tuning method: LoRA 05/16/2024 23:41:41 - INFO - llmtuner.model.loader - trainable params: 3407872 || all params: 7245139968 || trainable%: 0.0470 05/16/2024 23:41:41 - INFO - transformers.trainer - Using auto half precision backend 05/16/2024 23:41:41 - INFO - transformers.trainer - ***** Running training ***** 05/16/2024 23:41:41 - INFO - transformers.trainer - Num examples = 743 05/16/2024 23:41:41 - INFO - transformers.trainer - Num Epochs = 1 05/16/2024 23:41:41 - INFO - transformers.trainer - Instantaneous batch size per device = 16 05/16/2024 23:41:41 - INFO - transformers.trainer - Total train batch size (w. parallel, distributed & accumulation) = 64 05/16/2024 23:41:41 - INFO - transformers.trainer - Gradient Accumulation steps = 4 05/16/2024 23:41:41 - INFO - transformers.trainer - Total optimization steps = 11 05/16/2024 23:41:41 - INFO - transformers.trainer - Number of trainable parameters = 3,407,872 05/16/2024 23:43:41 - INFO - llmtuner.extras.callbacks - {'loss': 1.9309, 'learning_rate': 1.1423e-04, 'epoch': 0.43} 05/16/2024 23:45:40 - INFO - llmtuner.extras.callbacks - {'loss': 1.7404, 'learning_rate': 4.0507e-06, 'epoch': 0.85} 05/16/2024 23:46:04 - INFO - transformers.trainer - Training completed. Do not forget to share your model on huggingface.co/models =) 05/16/2024 23:46:04 - INFO - transformers.trainer - Saving model checkpoint to saves/Mistral-7B-v0.2-Chat/lora/Chapterization_Mistral-7B-v0.2-Chat_0.1.0 05/16/2024 23:46:05 - INFO - transformers.configuration_utils - loading configuration file config.json from cache at /home/marx/.cache/huggingface/hub/models--mistralai--Mistral-7B-Instruct-v0.2/snapshots/41b61a33a2483885c981aa79e0df6b32407ed873/config.json 05/16/2024 23:46:05 - INFO - transformers.configuration_utils - Model config MistralConfig { "architectures": [ "MistralForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 1, "eos_token_id": 2, "hidden_act": "silu", "hidden_size": 4096, "initializer_range": 0.02, "intermediate_size": 14336, "max_position_embeddings": 32768, "model_type": "mistral", "num_attention_heads": 32, "num_hidden_layers": 32, "num_key_value_heads": 8, "rms_norm_eps": 1e-05, "rope_theta": 1000000.0, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "bfloat16", "transformers_version": "4.40.2", "use_cache": true, "vocab_size": 32000 } 05/16/2024 23:46:05 - INFO - transformers.tokenization_utils_base - tokenizer config file saved in saves/Mistral-7B-v0.2-Chat/lora/Chapterization_Mistral-7B-v0.2-Chat_0.1.0/tokenizer_config.json 05/16/2024 23:46:05 - INFO - transformers.tokenization_utils_base - Special tokens file saved in saves/Mistral-7B-v0.2-Chat/lora/Chapterization_Mistral-7B-v0.2-Chat_0.1.0/special_tokens_map.json 05/16/2024 23:46:05 - INFO - transformers.modelcard - Dropping the following result as it does not have all the necessary fields: {'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}}