| [INFO|2025-02-28 17:03:19] configuration_utils.py:699 >> loading configuration file config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/config.json | |
| [INFO|2025-02-28 17:03:19] configuration_utils.py:771 >> Model config LlamaConfig { | |
| "_name_or_path": "meta-llama/Llama-3.3-70B-Instruct", | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 128000, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 8192, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 28672, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 64, | |
| "num_hidden_layers": 80, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 8.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.49.0", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2050 >> loading file tokenizer.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/tokenizer.json | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2050 >> loading file tokenizer.model from cache at None | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2050 >> loading file added_tokens.json from cache at None | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2050 >> loading file special_tokens_map.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/special_tokens_map.json | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2050 >> loading file tokenizer_config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/tokenizer_config.json | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2050 >> loading file chat_template.jinja from cache at None | |
| [INFO|2025-02-28 17:03:19] tokenization_utils_base.py:2313 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained. | |
| [INFO|2025-02-28 17:03:19] configuration_utils.py:699 >> loading configuration file config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/config.json | |
| [INFO|2025-02-28 17:03:19] configuration_utils.py:771 >> Model config LlamaConfig { | |
| "_name_or_path": "meta-llama/Llama-3.3-70B-Instruct", | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 128000, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 8192, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 28672, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 64, | |
| "num_hidden_layers": 80, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 8.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.49.0", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2050 >> loading file tokenizer.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/tokenizer.json | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2050 >> loading file tokenizer.model from cache at None | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2050 >> loading file added_tokens.json from cache at None | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2050 >> loading file special_tokens_map.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/special_tokens_map.json | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2050 >> loading file tokenizer_config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/tokenizer_config.json | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2050 >> loading file chat_template.jinja from cache at None | |
| [INFO|2025-02-28 17:03:20] tokenization_utils_base.py:2313 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained. | |
| [INFO|2025-02-28 17:03:20] logging.py:157 >> Add <|eot_id|>,<|eom_id|> to stop words. | |
| [INFO|2025-02-28 17:03:20] logging.py:157 >> Loading dataset jgayed/ets10119... | |
| [INFO|2025-02-28 17:03:21] configuration_utils.py:699 >> loading configuration file config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/config.json | |
| [INFO|2025-02-28 17:03:21] configuration_utils.py:771 >> Model config LlamaConfig { | |
| "_name_or_path": "meta-llama/Llama-3.3-70B-Instruct", | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 128000, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 8192, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 28672, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 64, | |
| "num_hidden_layers": 80, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 8.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.49.0", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } | |
| [INFO|2025-02-28 17:03:21] logging.py:157 >> Quantizing model to 4 bit with bitsandbytes. | |
| [INFO|2025-02-28 17:03:21] modeling_utils.py:3982 >> loading weights file model.safetensors from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/model.safetensors.index.json | |
| [INFO|2025-02-28 17:03:21] modeling_utils.py:1633 >> Instantiating LlamaForCausalLM model under default dtype torch.bfloat16. | |
| [INFO|2025-02-28 17:03:21] configuration_utils.py:1140 >> Generate config GenerationConfig { | |
| "bos_token_id": 128000, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ] | |
| } | |
| [INFO|2025-02-28 17:03:45] modeling_utils.py:4970 >> All model checkpoint weights were used when initializing LlamaForCausalLM. | |
| [INFO|2025-02-28 17:03:45] modeling_utils.py:4978 >> All the weights of LlamaForCausalLM were initialized from the model checkpoint at meta-llama/Llama-3.3-70B-Instruct. | |
| 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. | |
| [INFO|2025-02-28 17:03:45] configuration_utils.py:1095 >> loading configuration file generation_config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/generation_config.json | |
| [INFO|2025-02-28 17:03:45] configuration_utils.py:1140 >> Generate config GenerationConfig { | |
| "bos_token_id": 128000, | |
| "do_sample": true, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "temperature": 0.6, | |
| "top_p": 0.9 | |
| } | |
| [INFO|2025-02-28 17:03:45] logging.py:157 >> Gradient checkpointing enabled. | |
| [INFO|2025-02-28 17:03:45] logging.py:157 >> Using torch SDPA for faster training and inference. | |
| [INFO|2025-02-28 17:03:45] logging.py:157 >> Upcasting trainable params to float32. | |
| [INFO|2025-02-28 17:03:45] logging.py:157 >> Fine-tuning method: LoRA | |
| [INFO|2025-02-28 17:03:45] logging.py:157 >> Found linear modules: k_proj,up_proj,v_proj,down_proj,q_proj,o_proj,gate_proj | |
| [INFO|2025-02-28 17:03:55] logging.py:157 >> trainable params: 207,093,760 || all params: 70,760,800,256 || trainable%: 0.2927 | |
| [INFO|2025-02-28 17:03:55] trainer.py:746 >> Using auto half precision backend | |
| [WARNING|2025-02-28 17:03:55] trainer.py:781 >> No label_names provided for model class `PeftModelForCausalLM`. Since `PeftModel` hides base models input arguments, if label_names is not given, label_names can't be set automatically within `Trainer`. Note that empty label_names list will be used instead. | |
| [INFO|2025-02-28 17:03:55] trainer.py:2405 >> ***** Running training ***** | |
| [INFO|2025-02-28 17:03:55] trainer.py:2406 >> Num examples = 101 | |
| [INFO|2025-02-28 17:03:55] trainer.py:2407 >> Num Epochs = 3 | |
| [INFO|2025-02-28 17:03:55] trainer.py:2408 >> Instantaneous batch size per device = 1 | |
| [INFO|2025-02-28 17:03:55] trainer.py:2411 >> Total train batch size (w. parallel, distributed & accumulation) = 8 | |
| [INFO|2025-02-28 17:03:55] trainer.py:2412 >> Gradient Accumulation steps = 8 | |
| [INFO|2025-02-28 17:03:55] trainer.py:2413 >> Total optimization steps = 36 | |
| [INFO|2025-02-28 17:03:55] trainer.py:2414 >> Number of trainable parameters = 207,093,760 | |
| [INFO|2025-02-28 17:04:54] logging.py:157 >> {'loss': 10.5953, 'learning_rate': 4.7658e-05, 'epoch': 0.40, 'throughput': 593.47} | |
| [INFO|2025-02-28 17:05:55] logging.py:157 >> {'loss': 2.1097, 'learning_rate': 4.1070e-05, 'epoch': 0.79, 'throughput': 598.49} | |
| [INFO|2025-02-28 17:06:49] logging.py:157 >> {'loss': 0.4752, 'learning_rate': 3.1470e-05, 'epoch': 1.16, 'throughput': 595.11} | |
| [INFO|2025-02-28 17:07:49] logging.py:157 >> {'loss': 0.4311, 'learning_rate': 2.0659e-05, 'epoch': 1.55, 'throughput': 596.05} | |
| [INFO|2025-02-28 17:08:49] logging.py:157 >> {'loss': 0.4116, 'learning_rate': 1.0661e-05, 'epoch': 1.95, 'throughput': 597.38} | |
| [INFO|2025-02-28 17:09:44] logging.py:157 >> {'loss': 0.3330, 'learning_rate': 3.3494e-06, 'epoch': 2.32, 'throughput': 596.59} | |
| [INFO|2025-02-28 17:10:44] logging.py:157 >> {'loss': 0.3408, 'learning_rate': 9.5133e-08, 'epoch': 2.71, 'throughput': 596.82} | |
| [INFO|2025-02-28 17:10:55] trainer.py:3942 >> Saving model checkpoint to saves/Llama-3.3-70B-Instruct/lora/test/checkpoint-36 | |
| [INFO|2025-02-28 17:10:56] configuration_utils.py:699 >> loading configuration file config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/config.json | |
| [INFO|2025-02-28 17:10:56] configuration_utils.py:771 >> Model config LlamaConfig { | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 128000, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 8192, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 28672, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 64, | |
| "num_hidden_layers": 80, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 8.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.49.0", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } | |
| [INFO|2025-02-28 17:10:56] tokenization_utils_base.py:2500 >> tokenizer config file saved in saves/Llama-3.3-70B-Instruct/lora/test/checkpoint-36/tokenizer_config.json | |
| [INFO|2025-02-28 17:10:56] tokenization_utils_base.py:2509 >> Special tokens file saved in saves/Llama-3.3-70B-Instruct/lora/test/checkpoint-36/special_tokens_map.json | |
| [INFO|2025-02-28 17:10:58] trainer.py:2657 >> | |
| Training completed. Do not forget to share your model on huggingface.co/models =) | |
| [INFO|2025-02-28 17:10:58] trainer.py:3942 >> Saving model checkpoint to saves/Llama-3.3-70B-Instruct/lora/test | |
| [INFO|2025-02-28 17:10:58] configuration_utils.py:699 >> loading configuration file config.json from cache at /workspace/hf_home/hub/models--meta-llama--Llama-3.3-70B-Instruct/snapshots/6f6073b423013f6a7d4d9f39144961bfbfbc386b/config.json | |
| [INFO|2025-02-28 17:10:58] configuration_utils.py:771 >> Model config LlamaConfig { | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 128000, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 8192, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 28672, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 64, | |
| "num_hidden_layers": 80, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 8.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.49.0", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } | |
| [INFO|2025-02-28 17:10:59] tokenization_utils_base.py:2500 >> tokenizer config file saved in saves/Llama-3.3-70B-Instruct/lora/test/tokenizer_config.json | |
| [INFO|2025-02-28 17:10:59] tokenization_utils_base.py:2509 >> Special tokens file saved in saves/Llama-3.3-70B-Instruct/lora/test/special_tokens_map.json | |
| [WARNING|2025-02-28 17:10:59] logging.py:162 >> No metric eval_loss to plot. | |
| [WARNING|2025-02-28 17:10:59] logging.py:162 >> No metric eval_accuracy to plot. | |
| [INFO|2025-02-28 17:10:59] modelcard.py:449 >> Dropping the following result as it does not have all the necessary fields: | |
| {'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}} | |