File size: 17,629 Bytes
7547e01 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 |
06/17/2024 19:50:24 - INFO - transformers.models.auto.tokenization_auto - Could not locate the tokenizer configuration file, will try to use the model config instead. 06/17/2024 19:50:25 - INFO - transformers.configuration_utils - loading configuration file config.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/config.json 06/17/2024 19:50:25 - INFO - transformers.configuration_utils - Model config Qwen2Config { "_name_or_path": "Qwen/Qwen2-1.5B-Instruct", "architectures": [ "Qwen2ForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151645, "hidden_act": "silu", "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 8960, "max_position_embeddings": 32768, "max_window_layers": 28, "model_type": "qwen2", "num_attention_heads": 12, "num_hidden_layers": 28, "num_key_value_heads": 2, "rms_norm_eps": 1e-06, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": true, "torch_dtype": "bfloat16", "transformers_version": "4.41.2", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 } 06/17/2024 19:50:30 - INFO - transformers.tokenization_utils_base - loading file vocab.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/vocab.json 06/17/2024 19:50:30 - INFO - transformers.tokenization_utils_base - loading file merges.txt from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/merges.txt 06/17/2024 19:50:30 - INFO - transformers.tokenization_utils_base - loading file tokenizer.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/tokenizer.json 06/17/2024 19:50:30 - INFO - transformers.tokenization_utils_base - loading file added_tokens.json from cache at None 06/17/2024 19:50:30 - INFO - transformers.tokenization_utils_base - loading file special_tokens_map.json from cache at None 06/17/2024 19:50:30 - INFO - transformers.tokenization_utils_base - loading file tokenizer_config.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/tokenizer_config.json 06/17/2024 19:50:31 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained. 06/17/2024 19:50:31 - INFO - llamafactory.data.template - Replace eos token: <|im_end|> 06/17/2024 19:50:31 - WARNING - transformers.tokenization_utils_base - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained. 06/17/2024 19:50:31 - INFO - llamafactory.data.template - Replace eos token: <|im_end|> 06/17/2024 19:50:31 - INFO - llamafactory.data.loader - Loading dataset llamafactory/glaive_toolcall_zh... 06/17/2024 19:50:37 - INFO - llamafactory.data.loader - Loading dataset llamafactory/glaive_toolcall_en... 06/17/2024 19:50:43 - INFO - llamafactory.data.loader - Loading dataset llamafactory/glaive_toolcall_zh... 06/17/2024 19:50:47 - INFO - llamafactory.data.loader - Loading dataset llamafactory/glaive_toolcall_en... 06/17/2024 19:50:52 - INFO - transformers.configuration_utils - loading configuration file config.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/config.json 06/17/2024 19:50:52 - INFO - transformers.configuration_utils - Model config Qwen2Config { "_name_or_path": "Qwen/Qwen2-1.5B-Instruct", "architectures": [ "Qwen2ForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151645, "hidden_act": "silu", "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 8960, "max_position_embeddings": 32768, "max_window_layers": 28, "model_type": "qwen2", "num_attention_heads": 12, "num_hidden_layers": 28, "num_key_value_heads": 2, "rms_norm_eps": 1e-06, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": true, "torch_dtype": "bfloat16", "transformers_version": "4.41.2", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 } 06/17/2024 19:50:52 - INFO - llamafactory.model.model_utils.quantization - Quantizing model to 4 bit. 06/17/2024 19:50:52 - INFO - llamafactory.model.model_utils.quantization - Quantizing model to 4 bit. 06/17/2024 19:50:59 - INFO - transformers.modeling_utils - loading weights file model.safetensors from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/model.safetensors 06/17/2024 19:50:59 - INFO - transformers.modeling_utils - Instantiating Qwen2ForCausalLM model under default dtype torch.float16. 06/17/2024 19:50:59 - INFO - transformers.generation.configuration_utils - Generate config GenerationConfig { "bos_token_id": 151643, "eos_token_id": 151645 } 06/17/2024 19:51:06 - INFO - transformers.modeling_utils - All model checkpoint weights were used when initializing Qwen2ForCausalLM. 06/17/2024 19:51:06 - INFO - transformers.modeling_utils - All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at Qwen/Qwen2-1.5B-Instruct. If your task is similar to the task the model of the checkpoint was trained on, you can already use Qwen2ForCausalLM for predictions without further training. 06/17/2024 19:51:06 - INFO - transformers.generation.configuration_utils - loading configuration file generation_config.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/generation_config.json 06/17/2024 19:51:06 - INFO - transformers.generation.configuration_utils - Generate config GenerationConfig { "bos_token_id": 151643, "do_sample": true, "eos_token_id": [ 151645, 151643 ], "pad_token_id": 151643, "repetition_penalty": 1.1, "temperature": 0.7, "top_k": 20, "top_p": 0.8 } 06/17/2024 19:51:07 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled. 06/17/2024 19:51:07 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference. 06/17/2024 19:51:07 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32. 06/17/2024 19:51:07 - INFO - llamafactory.model.adapter - Fine-tuning method: LoRA 06/17/2024 19:51:07 - INFO - llamafactory.model.model_utils.misc - Found linear modules: o_proj,v_proj,down_proj,gate_proj,k_proj,up_proj,q_proj 06/17/2024 19:51:07 - INFO - llamafactory.model.model_utils.checkpointing - Gradient checkpointing enabled. 06/17/2024 19:51:07 - INFO - llamafactory.model.model_utils.attention - Using torch SDPA for faster training and inference. 06/17/2024 19:51:07 - INFO - llamafactory.model.adapter - Upcasting trainable params to float32. 06/17/2024 19:51:07 - INFO - llamafactory.model.adapter - Fine-tuning method: LoRA 06/17/2024 19:51:07 - INFO - llamafactory.model.model_utils.misc - Found linear modules: down_proj,k_proj,gate_proj,up_proj,o_proj,v_proj,q_proj 06/17/2024 19:51:07 - INFO - llamafactory.model.loader - trainable params: 9232384 || all params: 1552946688 || trainable%: 0.5945 06/17/2024 19:51:07 - INFO - llamafactory.model.loader - trainable params: 9232384 || all params: 1552946688 || trainable%: 0.5945 06/17/2024 19:51:07 - WARNING - accelerate.utils.other - Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher. 06/17/2024 19:51:07 - INFO - transformers.trainer - Using auto half precision backend 06/17/2024 19:51:07 - INFO - transformers.trainer - ***** Running training ***** 06/17/2024 19:51:07 - INFO - transformers.trainer - Num examples = 2,000 06/17/2024 19:51:07 - INFO - transformers.trainer - Num Epochs = 3 06/17/2024 19:51:07 - INFO - transformers.trainer - Instantaneous batch size per device = 2 06/17/2024 19:51:07 - INFO - transformers.trainer - Total train batch size (w. parallel, distributed & accumulation) = 32 06/17/2024 19:51:07 - INFO - transformers.trainer - Gradient Accumulation steps = 8 06/17/2024 19:51:07 - INFO - transformers.trainer - Total optimization steps = 186 06/17/2024 19:51:07 - INFO - transformers.trainer - Number of trainable parameters = 9,232,384 06/17/2024 19:51:52 - INFO - llamafactory.extras.callbacks - {'loss': 0.8099, 'learning_rate': 4.9911e-05, 'epoch': 0.08, 'throughput': 2566.78} 06/17/2024 19:52:37 - INFO - llamafactory.extras.callbacks - {'loss': 0.9580, 'learning_rate': 4.9644e-05, 'epoch': 0.16, 'throughput': 2542.66} 06/17/2024 19:53:18 - INFO - llamafactory.extras.callbacks - {'loss': 0.7150, 'learning_rate': 4.9202e-05, 'epoch': 0.24, 'throughput': 2505.70} 06/17/2024 19:54:02 - INFO - llamafactory.extras.callbacks - {'loss': 0.7585, 'learning_rate': 4.8587e-05, 'epoch': 0.32, 'throughput': 2511.00} 06/17/2024 19:54:37 - INFO - llamafactory.extras.callbacks - {'loss': 0.7342, 'learning_rate': 4.7804e-05, 'epoch': 0.40, 'throughput': 2533.17} 06/17/2024 19:55:16 - INFO - llamafactory.extras.callbacks - {'loss': 0.6904, 'learning_rate': 4.6859e-05, 'epoch': 0.48, 'throughput': 2557.91} 06/17/2024 19:55:54 - INFO - llamafactory.extras.callbacks - {'loss': 0.8254, 'learning_rate': 4.5757e-05, 'epoch': 0.56, 'throughput': 2587.35} 06/17/2024 19:56:34 - INFO - llamafactory.extras.callbacks - {'loss': 0.7551, 'learning_rate': 4.4508e-05, 'epoch': 0.64, 'throughput': 2578.43} 06/17/2024 19:57:15 - INFO - llamafactory.extras.callbacks - {'loss': 0.7747, 'learning_rate': 4.3120e-05, 'epoch': 0.72, 'throughput': 2580.98} 06/17/2024 19:57:54 - INFO - llamafactory.extras.callbacks - {'loss': 0.7027, 'learning_rate': 4.1602e-05, 'epoch': 0.80, 'throughput': 2578.88} 06/17/2024 19:58:32 - INFO - llamafactory.extras.callbacks - {'loss': 0.7581, 'learning_rate': 3.9967e-05, 'epoch': 0.88, 'throughput': 2580.17} 06/17/2024 19:59:13 - INFO - llamafactory.extras.callbacks - {'loss': 0.7221, 'learning_rate': 3.8224e-05, 'epoch': 0.96, 'throughput': 2583.14} 06/17/2024 19:59:55 - INFO - llamafactory.extras.callbacks - {'loss': 0.8214, 'learning_rate': 3.6387e-05, 'epoch': 1.04, 'throughput': 2587.56} 06/17/2024 20:00:36 - INFO - llamafactory.extras.callbacks - {'loss': 0.6304, 'learning_rate': 3.4469e-05, 'epoch': 1.12, 'throughput': 2580.91} 06/17/2024 20:01:12 - INFO - llamafactory.extras.callbacks - {'loss': 0.6434, 'learning_rate': 3.2484e-05, 'epoch': 1.20, 'throughput': 2577.50} 06/17/2024 20:01:55 - INFO - llamafactory.extras.callbacks - {'loss': 0.6796, 'learning_rate': 3.0445e-05, 'epoch': 1.28, 'throughput': 2575.34} 06/17/2024 20:02:35 - INFO - llamafactory.extras.callbacks - {'loss': 0.6651, 'learning_rate': 2.8368e-05, 'epoch': 1.36, 'throughput': 2580.63} 06/17/2024 20:03:17 - INFO - llamafactory.extras.callbacks - {'loss': 0.7844, 'learning_rate': 2.6266e-05, 'epoch': 1.44, 'throughput': 2586.23} 06/17/2024 20:04:01 - INFO - llamafactory.extras.callbacks - {'loss': 0.8139, 'learning_rate': 2.4156e-05, 'epoch': 1.52, 'throughput': 2581.43} 06/17/2024 20:04:43 - INFO - llamafactory.extras.callbacks - {'loss': 0.6717, 'learning_rate': 2.2051e-05, 'epoch': 1.60, 'throughput': 2578.63} 06/17/2024 20:04:43 - INFO - transformers.trainer - Saving model checkpoint to saves/Qwen2-1.5B-Chat/lora/train_2024-06-17-19-49-05/checkpoint-100 06/17/2024 20:04:44 - INFO - transformers.configuration_utils - loading configuration file config.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/config.json 06/17/2024 20:04:44 - INFO - transformers.configuration_utils - Model config Qwen2Config { "architectures": [ "Qwen2ForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151645, "hidden_act": "silu", "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 8960, "max_position_embeddings": 32768, "max_window_layers": 28, "model_type": "qwen2", "num_attention_heads": 12, "num_hidden_layers": 28, "num_key_value_heads": 2, "rms_norm_eps": 1e-06, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": true, "torch_dtype": "bfloat16", "transformers_version": "4.41.2", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 } 06/17/2024 20:04:44 - INFO - transformers.tokenization_utils_base - tokenizer config file saved in saves/Qwen2-1.5B-Chat/lora/train_2024-06-17-19-49-05/checkpoint-100/tokenizer_config.json 06/17/2024 20:04:44 - INFO - transformers.tokenization_utils_base - Special tokens file saved in saves/Qwen2-1.5B-Chat/lora/train_2024-06-17-19-49-05/checkpoint-100/special_tokens_map.json 06/17/2024 20:05:27 - INFO - llamafactory.extras.callbacks - {'loss': 0.7524, 'learning_rate': 1.9968e-05, 'epoch': 1.68, 'throughput': 2577.77} 06/17/2024 20:06:06 - INFO - llamafactory.extras.callbacks - {'loss': 0.6310, 'learning_rate': 1.7920e-05, 'epoch': 1.76, 'throughput': 2578.49} 06/17/2024 20:06:48 - INFO - llamafactory.extras.callbacks - {'loss': 0.7462, 'learning_rate': 1.5923e-05, 'epoch': 1.84, 'throughput': 2578.25} 06/17/2024 20:07:32 - INFO - llamafactory.extras.callbacks - {'loss': 0.6148, 'learning_rate': 1.3990e-05, 'epoch': 1.92, 'throughput': 2578.95} 06/17/2024 20:08:10 - INFO - llamafactory.extras.callbacks - {'loss': 0.7145, 'learning_rate': 1.2136e-05, 'epoch': 2.00, 'throughput': 2582.83} 06/17/2024 20:08:52 - INFO - llamafactory.extras.callbacks - {'loss': 0.6798, 'learning_rate': 1.0374e-05, 'epoch': 2.08, 'throughput': 2583.35} 06/17/2024 20:09:31 - INFO - llamafactory.extras.callbacks - {'loss': 0.6754, 'learning_rate': 8.7157e-06, 'epoch': 2.16, 'throughput': 2584.30} 06/17/2024 20:10:10 - INFO - llamafactory.extras.callbacks - {'loss': 0.6708, 'learning_rate': 7.1737e-06, 'epoch': 2.24, 'throughput': 2584.52} 06/17/2024 20:10:46 - INFO - llamafactory.extras.callbacks - {'loss': 0.6386, 'learning_rate': 5.7587e-06, 'epoch': 2.32, 'throughput': 2587.95} 06/17/2024 20:11:24 - INFO - llamafactory.extras.callbacks - {'loss': 0.6995, 'learning_rate': 4.4809e-06, 'epoch': 2.40, 'throughput': 2590.06} 06/17/2024 20:12:06 - INFO - llamafactory.extras.callbacks - {'loss': 0.6691, 'learning_rate': 3.3494e-06, 'epoch': 2.48, 'throughput': 2593.15} 06/17/2024 20:12:50 - INFO - llamafactory.extras.callbacks - {'loss': 0.6024, 'learning_rate': 2.3721e-06, 'epoch': 2.56, 'throughput': 2588.49} 06/17/2024 20:13:28 - INFO - llamafactory.extras.callbacks - {'loss': 0.6484, 'learning_rate': 1.5562e-06, 'epoch': 2.64, 'throughput': 2591.26} 06/17/2024 20:14:10 - INFO - llamafactory.extras.callbacks - {'loss': 0.7137, 'learning_rate': 9.0736e-07, 'epoch': 2.72, 'throughput': 2585.69} 06/17/2024 20:14:56 - INFO - llamafactory.extras.callbacks - {'loss': 0.7770, 'learning_rate': 4.3025e-07, 'epoch': 2.80, 'throughput': 2589.29} 06/17/2024 20:15:36 - INFO - llamafactory.extras.callbacks - {'loss': 0.6529, 'learning_rate': 1.2827e-07, 'epoch': 2.88, 'throughput': 2590.50} 06/17/2024 20:16:16 - INFO - llamafactory.extras.callbacks - {'loss': 0.6996, 'learning_rate': 3.5659e-09, 'epoch': 2.96, 'throughput': 2590.41} 06/17/2024 20:16:26 - INFO - transformers.trainer - Training completed. Do not forget to share your model on huggingface.co/models =) 06/17/2024 20:16:26 - INFO - transformers.trainer - Saving model checkpoint to saves/Qwen2-1.5B-Chat/lora/train_2024-06-17-19-49-05 06/17/2024 20:16:27 - INFO - transformers.configuration_utils - loading configuration file config.json from cache at /home/featurize/.cache/huggingface/hub/models--Qwen--Qwen2-1.5B-Instruct/snapshots/ba1cf1846d7df0a0591d6c00649f57e798519da8/config.json 06/17/2024 20:16:27 - INFO - transformers.configuration_utils - Model config Qwen2Config { "architectures": [ "Qwen2ForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 151643, "eos_token_id": 151645, "hidden_act": "silu", "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 8960, "max_position_embeddings": 32768, "max_window_layers": 28, "model_type": "qwen2", "num_attention_heads": 12, "num_hidden_layers": 28, "num_key_value_heads": 2, "rms_norm_eps": 1e-06, "rope_theta": 1000000.0, "sliding_window": 32768, "tie_word_embeddings": true, "torch_dtype": "bfloat16", "transformers_version": "4.41.2", "use_cache": true, "use_sliding_window": false, "vocab_size": 151936 } 06/17/2024 20:16:27 - INFO - transformers.tokenization_utils_base - tokenizer config file saved in saves/Qwen2-1.5B-Chat/lora/train_2024-06-17-19-49-05/tokenizer_config.json 06/17/2024 20:16:27 - INFO - transformers.tokenization_utils_base - Special tokens file saved in saves/Qwen2-1.5B-Chat/lora/train_2024-06-17-19-49-05/special_tokens_map.json 06/17/2024 20:16:27 - WARNING - llamafactory.extras.ploting - No metric eval_loss to plot. 06/17/2024 20:16:27 - INFO - transformers.modelcard - Dropping the following result as it does not have all the necessary fields: {'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}} |