Upload running_log.txt with huggingface_hub

running_log.txt

@@ -0,0 +1,618 @@
+[INFO|2025-02-17 12:12:13] configuration_utils.py:696 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/config.json
+
+[INFO|2025-02-17 12:12:13] configuration_utils.py:768 >> Model config Qwen2Config {
+  "_name_or_path": "Qwen/Qwen2.5-1.5B",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 131072,
+  "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_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file vocab.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/vocab.json
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file merges.txt from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/merges.txt
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file tokenizer.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/tokenizer.json
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file added_tokens.json from cache at None
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file special_tokens_map.json from cache at None
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file tokenizer_config.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/tokenizer_config.json
+
+[INFO|2025-02-17 12:12:17] tokenization_utils_base.py:2034 >> loading file chat_template.jinja from cache at None
+
+[INFO|2025-02-17 12:12:18] tokenization_utils_base.py:2304 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+
+[INFO|2025-02-17 12:12:19] configuration_utils.py:696 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/config.json
+
+[INFO|2025-02-17 12:12:19] configuration_utils.py:768 >> Model config Qwen2Config {
+  "_name_or_path": "Qwen/Qwen2.5-1.5B",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 131072,
+  "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_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file vocab.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/vocab.json
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file merges.txt from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/merges.txt
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file tokenizer.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/tokenizer.json
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file added_tokens.json from cache at None
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file special_tokens_map.json from cache at None
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file tokenizer_config.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/tokenizer_config.json
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2034 >> loading file chat_template.jinja from cache at None
+
+[INFO|2025-02-17 12:12:19] tokenization_utils_base.py:2304 >> Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
+
+[INFO|2025-02-17 12:12:19] logging.py:157 >> Loading dataset limo.json...
+
+[INFO|2025-02-17 12:12:27] configuration_utils.py:696 >> loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/config.json
+
+[INFO|2025-02-17 12:12:27] configuration_utils.py:768 >> Model config Qwen2Config {
+  "_name_or_path": "Qwen/Qwen2.5-1.5B",
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "hidden_act": "silu",
+  "hidden_size": 1536,
+  "initializer_range": 0.02,
+  "intermediate_size": 8960,
+  "max_position_embeddings": 131072,
+  "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_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": null,
+  "tie_word_embeddings": true,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.48.2",
+  "use_cache": true,
+  "use_mrope": false,
+  "use_sliding_window": false,
+  "vocab_size": 151936
+}
+
+
+[INFO|2025-02-17 12:13:41] modeling_utils.py:3904 >> loading weights file model.safetensors from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/model.safetensors
+
+[INFO|2025-02-17 12:13:41] modeling_utils.py:1582 >> Instantiating Qwen2ForCausalLM model under default dtype torch.bfloat16.
+
+[INFO|2025-02-17 12:13:41] configuration_utils.py:1140 >> Generate config GenerationConfig {
+  "bos_token_id": 151643,
+  "eos_token_id": 151643
+}
+
+
+[INFO|2025-02-17 12:13:43] modeling_utils.py:4888 >> All model checkpoint weights were used when initializing Qwen2ForCausalLM.
+
+
+[INFO|2025-02-17 12:13:43] modeling_utils.py:4896 >> All the weights of Qwen2ForCausalLM were initialized from the model checkpoint at Qwen/Qwen2.5-1.5B.
+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.
+
+[INFO|2025-02-17 12:13:43] configuration_utils.py:1095 >> loading configuration file generation_config.json from cache at /root/.cache/huggingface/hub/models--Qwen--Qwen2.5-1.5B/snapshots/8faed761d45a263340a0528343f099c05c9a4323/generation_config.json
+
+[INFO|2025-02-17 12:13:43] configuration_utils.py:1140 >> Generate config GenerationConfig {
+  "bos_token_id": 151643,
+  "eos_token_id": 151643,
+  "max_new_tokens": 2048
+}
+
+
+[INFO|2025-02-17 12:13:43] logging.py:157 >> Gradient checkpointing enabled.
+
+[INFO|2025-02-17 12:13:43] logging.py:157 >> Using torch SDPA for faster training and inference.
+
+[INFO|2025-02-17 12:13:43] logging.py:157 >> Upcasting trainable params to float32.
+
+[INFO|2025-02-17 12:13:43] logging.py:157 >> Fine-tuning method: Full
+
+[INFO|2025-02-17 12:13:43] logging.py:157 >> trainable params: 1,543,714,304 || all params: 1,543,714,304 || trainable%: 100.0000
+
+[INFO|2025-02-17 12:13:43] trainer.py:741 >> Using auto half precision backend
+
+[INFO|2025-02-17 12:13:44] trainer.py:2369 >> ***** Running training *****
+
+[INFO|2025-02-17 12:13:44] trainer.py:2370 >>   Num examples = 817
+
+[INFO|2025-02-17 12:13:44] trainer.py:2371 >>   Num Epochs = 15
+
+[INFO|2025-02-17 12:13:44] trainer.py:2372 >>   Instantaneous batch size per device = 2
+
+[INFO|2025-02-17 12:13:44] trainer.py:2375 >>   Total train batch size (w. parallel, distributed & accumulation) = 16
+
+[INFO|2025-02-17 12:13:44] trainer.py:2376 >>   Gradient Accumulation steps = 8
+
+[INFO|2025-02-17 12:13:44] trainer.py:2377 >>   Total optimization steps = 765
+
+[INFO|2025-02-17 12:13:44] trainer.py:2378 >>   Number of trainable parameters = 1,543,714,304
+
+[INFO|2025-02-17 12:13:44] integration_utils.py:817 >> Automatic Weights & Biases logging enabled, to disable set os.environ["WANDB_DISABLED"] = "true"
+
+[INFO|2025-02-17 12:14:20] logging.py:157 >> {'loss': 0.8739, 'learning_rate': 1.9998e-05, 'epoch': 0.10, 'throughput': 8024.43}
+
+[INFO|2025-02-17 12:14:40] logging.py:157 >> {'loss': 0.7957, 'learning_rate': 1.9992e-05, 'epoch': 0.20, 'throughput': 8106.29}
+
+[INFO|2025-02-17 12:14:59] logging.py:157 >> {'loss': 0.7822, 'learning_rate': 1.9981e-05, 'epoch': 0.29, 'throughput': 8219.61}
+
+[INFO|2025-02-17 12:15:19] logging.py:157 >> {'loss': 0.7815, 'learning_rate': 1.9966e-05, 'epoch': 0.39, 'throughput': 8260.20}
+
+[INFO|2025-02-17 12:15:38] logging.py:157 >> {'loss': 0.7432, 'learning_rate': 1.9947e-05, 'epoch': 0.49, 'throughput': 8269.83}
+
+[INFO|2025-02-17 12:15:58] logging.py:157 >> {'loss': 0.7389, 'learning_rate': 1.9924e-05, 'epoch': 0.59, 'throughput': 8282.10}
+
+[INFO|2025-02-17 12:16:17] logging.py:157 >> {'loss': 0.7654, 'learning_rate': 1.9897e-05, 'epoch': 0.68, 'throughput': 8305.86}
+
+[INFO|2025-02-17 12:16:37] logging.py:157 >> {'loss': 0.7753, 'learning_rate': 1.9865e-05, 'epoch': 0.78, 'throughput': 8319.21}
+
+[INFO|2025-02-17 12:16:56] logging.py:157 >> {'loss': 0.7256, 'learning_rate': 1.9830e-05, 'epoch': 0.88, 'throughput': 8321.93}
+
+[INFO|2025-02-17 12:17:16] logging.py:157 >> {'loss': 0.7270, 'learning_rate': 1.9790e-05, 'epoch': 0.98, 'throughput': 8320.84}
+
+[INFO|2025-02-17 12:17:32] logging.py:157 >> {'loss': 0.5820, 'learning_rate': 1.9746e-05, 'epoch': 1.06, 'throughput': 8319.31}
+
+[INFO|2025-02-17 12:17:52] logging.py:157 >> {'loss': 0.5628, 'learning_rate': 1.9698e-05, 'epoch': 1.16, 'throughput': 8321.68}
+
+[INFO|2025-02-17 12:18:11] logging.py:157 >> {'loss': 0.5561, 'learning_rate': 1.9646e-05, 'epoch': 1.25, 'throughput': 8326.47}
+
+[INFO|2025-02-17 12:18:31] logging.py:157 >> {'loss': 0.5778, 'learning_rate': 1.9590e-05, 'epoch': 1.35, 'throughput': 8327.91}
+
+[INFO|2025-02-17 12:18:51] logging.py:157 >> {'loss': 0.5207, 'learning_rate': 1.9529e-05, 'epoch': 1.45, 'throughput': 8328.77}
+
+[INFO|2025-02-17 12:19:10] logging.py:157 >> {'loss': 0.5452, 'learning_rate': 1.9465e-05, 'epoch': 1.55, 'throughput': 8334.34}
+
+[INFO|2025-02-17 12:19:29] logging.py:157 >> {'loss': 0.5379, 'learning_rate': 1.9397e-05, 'epoch': 1.65, 'throughput': 8341.18}
+
+[INFO|2025-02-17 12:19:49] logging.py:157 >> {'loss': 0.5116, 'learning_rate': 1.9325e-05, 'epoch': 1.74, 'throughput': 8337.37}
+
+[INFO|2025-02-17 12:20:09] logging.py:157 >> {'loss': 0.5532, 'learning_rate': 1.9249e-05, 'epoch': 1.84, 'throughput': 8339.42}
+
+[INFO|2025-02-17 12:20:28] logging.py:157 >> {'loss': 0.5659, 'learning_rate': 1.9169e-05, 'epoch': 1.94, 'throughput': 8344.96}
+
+[INFO|2025-02-17 12:20:28] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-100
+
+[INFO|2025-02-17 12:20:28] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-100/config.json
+
+[INFO|2025-02-17 12:20:28] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-100/generation_config.json
+
+[INFO|2025-02-17 12:20:48] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-100/model.safetensors.index.json.
+
+[INFO|2025-02-17 12:20:48] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-100/tokenizer_config.json
+
+[INFO|2025-02-17 12:20:48] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-100/special_tokens_map.json
+
+[INFO|2025-02-17 12:21:37] logging.py:157 >> {'loss': 0.6028, 'learning_rate': 1.9085e-05, 'epoch': 2.02, 'throughput': 7373.76}
+
+[INFO|2025-02-17 12:21:57] logging.py:157 >> {'loss': 0.4228, 'learning_rate': 1.8997e-05, 'epoch': 2.12, 'throughput': 7412.32}
+
+[INFO|2025-02-17 12:22:17] logging.py:157 >> {'loss': 0.3648, 'learning_rate': 1.8905e-05, 'epoch': 2.22, 'throughput': 7448.87}
+
+[INFO|2025-02-17 12:22:36] logging.py:157 >> {'loss': 0.3822, 'learning_rate': 1.8810e-05, 'epoch': 2.31, 'throughput': 7486.42}
+
+[INFO|2025-02-17 12:22:56] logging.py:157 >> {'loss': 0.3907, 'learning_rate': 1.8711e-05, 'epoch': 2.41, 'throughput': 7521.14}
+
+[INFO|2025-02-17 12:23:15] logging.py:157 >> {'loss': 0.3936, 'learning_rate': 1.8608e-05, 'epoch': 2.51, 'throughput': 7551.36}
+
+[INFO|2025-02-17 12:23:35] logging.py:157 >> {'loss': 0.3760, 'learning_rate': 1.8502e-05, 'epoch': 2.61, 'throughput': 7577.31}
+
+[INFO|2025-02-17 12:23:55] logging.py:157 >> {'loss': 0.3735, 'learning_rate': 1.8392e-05, 'epoch': 2.70, 'throughput': 7602.53}
+
+[INFO|2025-02-17 12:24:14] logging.py:157 >> {'loss': 0.4114, 'learning_rate': 1.8279e-05, 'epoch': 2.80, 'throughput': 7628.26}
+
+[INFO|2025-02-17 12:24:33] logging.py:157 >> {'loss': 0.3696, 'learning_rate': 1.8162e-05, 'epoch': 2.90, 'throughput': 7652.37}
+
+[INFO|2025-02-17 12:24:53] logging.py:157 >> {'loss': 0.3888, 'learning_rate': 1.8042e-05, 'epoch': 3.00, 'throughput': 7671.13}
+
+[INFO|2025-02-17 12:25:09] logging.py:157 >> {'loss': 0.2605, 'learning_rate': 1.7918e-05, 'epoch': 3.08, 'throughput': 7688.94}
+
+[INFO|2025-02-17 12:25:28] logging.py:157 >> {'loss': 0.2475, 'learning_rate': 1.7791e-05, 'epoch': 3.18, 'throughput': 7710.27}
+
+[INFO|2025-02-17 12:25:48] logging.py:157 >> {'loss': 0.2344, 'learning_rate': 1.7660e-05, 'epoch': 3.27, 'throughput': 7726.87}
+
+[INFO|2025-02-17 12:26:08] logging.py:157 >> {'loss': 0.2665, 'learning_rate': 1.7527e-05, 'epoch': 3.37, 'throughput': 7743.40}
+
+[INFO|2025-02-17 12:26:27] logging.py:157 >> {'loss': 0.2477, 'learning_rate': 1.7390e-05, 'epoch': 3.47, 'throughput': 7760.52}
+
+[INFO|2025-02-17 12:26:47] logging.py:157 >> {'loss': 0.2679, 'learning_rate': 1.7250e-05, 'epoch': 3.57, 'throughput': 7776.01}
+
+[INFO|2025-02-17 12:27:07] logging.py:157 >> {'loss': 0.2340, 'learning_rate': 1.7107e-05, 'epoch': 3.67, 'throughput': 7790.11}
+
+[INFO|2025-02-17 12:27:26] logging.py:157 >> {'loss': 0.2587, 'learning_rate': 1.6961e-05, 'epoch': 3.76, 'throughput': 7801.88}
+
+[INFO|2025-02-17 12:27:46] logging.py:157 >> {'loss': 0.2606, 'learning_rate': 1.6812e-05, 'epoch': 3.86, 'throughput': 7815.59}
+
+[INFO|2025-02-17 12:27:46] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-200
+
+[INFO|2025-02-17 12:27:46] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-200/config.json
+
+[INFO|2025-02-17 12:27:46] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-200/generation_config.json
+
+[INFO|2025-02-17 12:28:07] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-200/model.safetensors.index.json.
+
+[INFO|2025-02-17 12:28:07] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-200/tokenizer_config.json
+
+[INFO|2025-02-17 12:28:07] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-200/special_tokens_map.json
+
+[INFO|2025-02-17 12:29:01] logging.py:157 >> {'loss': 0.2543, 'learning_rate': 1.6661e-05, 'epoch': 3.96, 'throughput': 7347.53}
+
+[INFO|2025-02-17 12:29:17] logging.py:157 >> {'loss': 0.1856, 'learning_rate': 1.6506e-05, 'epoch': 4.04, 'throughput': 7365.09}
+
+[INFO|2025-02-17 12:29:36] logging.py:157 >> {'loss': 0.1462, 'learning_rate': 1.6349e-05, 'epoch': 4.14, 'throughput': 7387.60}
+
+[INFO|2025-02-17 12:29:56] logging.py:157 >> {'loss': 0.1700, 'learning_rate': 1.6189e-05, 'epoch': 4.23, 'throughput': 7407.89}
+
+[INFO|2025-02-17 12:30:15] logging.py:157 >> {'loss': 0.1427, 'learning_rate': 1.6026e-05, 'epoch': 4.33, 'throughput': 7426.73}
+
+[INFO|2025-02-17 12:30:35] logging.py:157 >> {'loss': 0.1500, 'learning_rate': 1.5861e-05, 'epoch': 4.43, 'throughput': 7444.87}
+
+[INFO|2025-02-17 12:30:54] logging.py:157 >> {'loss': 0.1603, 'learning_rate': 1.5694e-05, 'epoch': 4.53, 'throughput': 7464.06}
+
+[INFO|2025-02-17 12:31:14] logging.py:157 >> {'loss': 0.1536, 'learning_rate': 1.5524e-05, 'epoch': 4.63, 'throughput': 7479.59}
+
+[INFO|2025-02-17 12:31:34] logging.py:157 >> {'loss': 0.1415, 'learning_rate': 1.5351e-05, 'epoch': 4.72, 'throughput': 7495.78}
+
+[INFO|2025-02-17 12:31:53] logging.py:157 >> {'loss': 0.1590, 'learning_rate': 1.5177e-05, 'epoch': 4.82, 'throughput': 7511.92}
+
+[INFO|2025-02-17 12:32:13] logging.py:157 >> {'loss': 0.1699, 'learning_rate': 1.5000e-05, 'epoch': 4.92, 'throughput': 7526.38}
+
+[INFO|2025-02-17 12:32:29] logging.py:157 >> {'loss': 0.1295, 'learning_rate': 1.4821e-05, 'epoch': 5.00, 'throughput': 7538.13}
+
+[INFO|2025-02-17 12:32:48] logging.py:157 >> {'loss': 0.1019, 'learning_rate': 1.4640e-05, 'epoch': 5.10, 'throughput': 7553.80}
+
+[INFO|2025-02-17 12:33:08] logging.py:157 >> {'loss': 0.0916, 'learning_rate': 1.4457e-05, 'epoch': 5.20, 'throughput': 7567.93}
+
+[INFO|2025-02-17 12:33:27] logging.py:157 >> {'loss': 0.0935, 'learning_rate': 1.4273e-05, 'epoch': 5.29, 'throughput': 7583.13}
+
+[INFO|2025-02-17 12:33:47] logging.py:157 >> {'loss': 0.0679, 'learning_rate': 1.4086e-05, 'epoch': 5.39, 'throughput': 7596.23}
+
+[INFO|2025-02-17 12:34:06] logging.py:157 >> {'loss': 0.1010, 'learning_rate': 1.3898e-05, 'epoch': 5.49, 'throughput': 7607.87}
+
+[INFO|2025-02-17 12:34:26] logging.py:157 >> {'loss': 0.0802, 'learning_rate': 1.3708e-05, 'epoch': 5.59, 'throughput': 7621.15}
+
+[INFO|2025-02-17 12:34:46] logging.py:157 >> {'loss': 0.0812, 'learning_rate': 1.3516e-05, 'epoch': 5.68, 'throughput': 7631.54}
+
+[INFO|2025-02-17 12:35:05] logging.py:157 >> {'loss': 0.0812, 'learning_rate': 1.3324e-05, 'epoch': 5.78, 'throughput': 7643.59}
+
+[INFO|2025-02-17 12:35:05] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-300
+
+[INFO|2025-02-17 12:35:05] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-300/config.json
+
+[INFO|2025-02-17 12:35:05] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-300/generation_config.json
+
+[INFO|2025-02-17 12:35:25] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-300/model.safetensors.index.json.
+
+[INFO|2025-02-17 12:35:25] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-300/tokenizer_config.json
+
+[INFO|2025-02-17 12:35:25] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-300/special_tokens_map.json
+
+[INFO|2025-02-17 12:36:24] logging.py:157 >> {'loss': 0.0884, 'learning_rate': 1.3129e-05, 'epoch': 5.88, 'throughput': 7317.78}
+
+[INFO|2025-02-17 12:36:44] logging.py:157 >> {'loss': 0.0906, 'learning_rate': 1.2934e-05, 'epoch': 5.98, 'throughput': 7332.20}
+
+[INFO|2025-02-17 12:37:00] logging.py:157 >> {'loss': 0.0616, 'learning_rate': 1.2737e-05, 'epoch': 6.06, 'throughput': 7343.00}
+
+[INFO|2025-02-17 12:37:19] logging.py:157 >> {'loss': 0.0447, 'learning_rate': 1.2539e-05, 'epoch': 6.16, 'throughput': 7357.50}
+
+[INFO|2025-02-17 12:37:39] logging.py:157 >> {'loss': 0.0528, 'learning_rate': 1.2339e-05, 'epoch': 6.25, 'throughput': 7372.88}
+
+[INFO|2025-02-17 12:37:58] logging.py:157 >> {'loss': 0.0429, 'learning_rate': 1.2139e-05, 'epoch': 6.35, 'throughput': 7387.40}
+
+[INFO|2025-02-17 12:38:18] logging.py:157 >> {'loss': 0.0442, 'learning_rate': 1.1938e-05, 'epoch': 6.45, 'throughput': 7398.63}
+
+[INFO|2025-02-17 12:38:37] logging.py:157 >> {'loss': 0.0470, 'learning_rate': 1.1736e-05, 'epoch': 6.55, 'throughput': 7411.53}
+
+[INFO|2025-02-17 12:38:57] logging.py:157 >> {'loss': 0.0437, 'learning_rate': 1.1534e-05, 'epoch': 6.65, 'throughput': 7424.59}
+
+[INFO|2025-02-17 12:39:17] logging.py:157 >> {'loss': 0.0466, 'learning_rate': 1.1331e-05, 'epoch': 6.74, 'throughput': 7436.34}
+
+[INFO|2025-02-17 12:39:36] logging.py:157 >> {'loss': 0.0498, 'learning_rate': 1.1127e-05, 'epoch': 6.84, 'throughput': 7447.16}
+
+[INFO|2025-02-17 12:39:56] logging.py:157 >> {'loss': 0.0409, 'learning_rate': 1.0923e-05, 'epoch': 6.94, 'throughput': 7459.25}
+
+[INFO|2025-02-17 12:40:12] logging.py:157 >> {'loss': 0.0444, 'learning_rate': 1.0718e-05, 'epoch': 7.02, 'throughput': 7468.55}
+
+[INFO|2025-02-17 12:40:31] logging.py:157 >> {'loss': 0.0212, 'learning_rate': 1.0513e-05, 'epoch': 7.12, 'throughput': 7479.66}
+
+[INFO|2025-02-17 12:40:51] logging.py:157 >> {'loss': 0.0275, 'learning_rate': 1.0308e-05, 'epoch': 7.22, 'throughput': 7489.44}
+
+[INFO|2025-02-17 12:41:10] logging.py:157 >> {'loss': 0.0234, 'learning_rate': 1.0103e-05, 'epoch': 7.31, 'throughput': 7500.49}
+
+[INFO|2025-02-17 12:41:30] logging.py:157 >> {'loss': 0.0227, 'learning_rate': 9.8973e-06, 'epoch': 7.41, 'throughput': 7510.51}
+
+[INFO|2025-02-17 12:41:50] logging.py:157 >> {'loss': 0.0264, 'learning_rate': 9.6920e-06, 'epoch': 7.51, 'throughput': 7520.71}
+
+[INFO|2025-02-17 12:42:09] logging.py:157 >> {'loss': 0.0251, 'learning_rate': 9.4869e-06, 'epoch': 7.61, 'throughput': 7530.64}
+
+[INFO|2025-02-17 12:42:29] logging.py:157 >> {'loss': 0.0237, 'learning_rate': 9.2820e-06, 'epoch': 7.70, 'throughput': 7541.10}
+
+[INFO|2025-02-17 12:42:29] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-400
+
+[INFO|2025-02-17 12:42:29] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-400/config.json
+
+[INFO|2025-02-17 12:42:29] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-400/generation_config.json
+
+[INFO|2025-02-17 12:42:49] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-400/model.safetensors.index.json.
+
+[INFO|2025-02-17 12:42:49] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-400/tokenizer_config.json
+
+[INFO|2025-02-17 12:42:49] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-400/special_tokens_map.json
+
+[INFO|2025-02-17 12:43:46] logging.py:157 >> {'loss': 0.0228, 'learning_rate': 9.0773e-06, 'epoch': 7.80, 'throughput': 7303.95}
+
+[INFO|2025-02-17 12:44:06] logging.py:157 >> {'loss': 0.0238, 'learning_rate': 8.8731e-06, 'epoch': 7.90, 'throughput': 7315.29}
+
+[INFO|2025-02-17 12:44:26] logging.py:157 >> {'loss': 0.0222, 'learning_rate': 8.6693e-06, 'epoch': 8.00, 'throughput': 7326.04}
+
+[INFO|2025-02-17 12:44:42] logging.py:157 >> {'loss': 0.0109, 'learning_rate': 8.4661e-06, 'epoch': 8.08, 'throughput': 7334.51}
+
+[INFO|2025-02-17 12:45:01] logging.py:157 >> {'loss': 0.0118, 'learning_rate': 8.2635e-06, 'epoch': 8.18, 'throughput': 7345.75}
+
+[INFO|2025-02-17 12:45:21] logging.py:157 >> {'loss': 0.0100, 'learning_rate': 8.0617e-06, 'epoch': 8.27, 'throughput': 7356.49}
+
+[INFO|2025-02-17 12:45:40] logging.py:157 >> {'loss': 0.0112, 'learning_rate': 7.8607e-06, 'epoch': 8.37, 'throughput': 7367.37}
+
+[INFO|2025-02-17 12:46:00] logging.py:157 >> {'loss': 0.0133, 'learning_rate': 7.6606e-06, 'epoch': 8.47, 'throughput': 7377.67}
+
+[INFO|2025-02-17 12:46:19] logging.py:157 >> {'loss': 0.0113, 'learning_rate': 7.4614e-06, 'epoch': 8.57, 'throughput': 7388.41}
+
+[INFO|2025-02-17 12:46:39] logging.py:157 >> {'loss': 0.0102, 'learning_rate': 7.2634e-06, 'epoch': 8.67, 'throughput': 7397.72}
+
+[INFO|2025-02-17 12:46:58] logging.py:157 >> {'loss': 0.0100, 'learning_rate': 7.0665e-06, 'epoch': 8.76, 'throughput': 7408.35}
+
+[INFO|2025-02-17 12:47:18] logging.py:157 >> {'loss': 0.0116, 'learning_rate': 6.8708e-06, 'epoch': 8.86, 'throughput': 7417.24}
+
+[INFO|2025-02-17 12:47:38] logging.py:157 >> {'loss': 0.0111, 'learning_rate': 6.6765e-06, 'epoch': 8.96, 'throughput': 7425.37}
+
+[INFO|2025-02-17 12:47:54] logging.py:157 >> {'loss': 0.0071, 'learning_rate': 6.4835e-06, 'epoch': 9.04, 'throughput': 7432.24}
+
+[INFO|2025-02-17 12:48:13] logging.py:157 >> {'loss': 0.0054, 'learning_rate': 6.2920e-06, 'epoch': 9.14, 'throughput': 7440.97}
+
+[INFO|2025-02-17 12:48:33] logging.py:157 >> {'loss': 0.0059, 'learning_rate': 6.1021e-06, 'epoch': 9.23, 'throughput': 7450.14}
+
+[INFO|2025-02-17 12:48:53] logging.py:157 >> {'loss': 0.0045, 'learning_rate': 5.9139e-06, 'epoch': 9.33, 'throughput': 7458.50}
+
+[INFO|2025-02-17 12:49:12] logging.py:157 >> {'loss': 0.0040, 'learning_rate': 5.7273e-06, 'epoch': 9.43, 'throughput': 7466.39}
+
+[INFO|2025-02-17 12:49:32] logging.py:157 >> {'loss': 0.0042, 'learning_rate': 5.5426e-06, 'epoch': 9.53, 'throughput': 7474.49}
+
+[INFO|2025-02-17 12:49:51] logging.py:157 >> {'loss': 0.0049, 'learning_rate': 5.3598e-06, 'epoch': 9.63, 'throughput': 7483.27}
+
+[INFO|2025-02-17 12:49:51] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-500
+
+[INFO|2025-02-17 12:49:51] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-500/config.json
+
+[INFO|2025-02-17 12:49:51] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-500/generation_config.json
+
+[INFO|2025-02-17 12:50:12] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-500/model.safetensors.index.json.
+
+[INFO|2025-02-17 12:50:12] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-500/tokenizer_config.json
+
+[INFO|2025-02-17 12:50:12] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-500/special_tokens_map.json
+
+[INFO|2025-02-17 12:51:10] logging.py:157 >> {'loss': 0.0045, 'learning_rate': 5.1789e-06, 'epoch': 9.72, 'throughput': 7292.37}
+
+[INFO|2025-02-17 12:51:30] logging.py:157 >> {'loss': 0.0040, 'learning_rate': 5.0000e-06, 'epoch': 9.82, 'throughput': 7300.78}
+
+[INFO|2025-02-17 12:51:49] logging.py:157 >> {'loss': 0.0042, 'learning_rate': 4.8232e-06, 'epoch': 9.92, 'throughput': 7310.63}
+
+[INFO|2025-02-17 12:52:05] logging.py:157 >> {'loss': 0.0037, 'learning_rate': 4.6487e-06, 'epoch': 10.00, 'throughput': 7317.82}
+
+[INFO|2025-02-17 12:52:25] logging.py:157 >> {'loss': 0.0021, 'learning_rate': 4.4764e-06, 'epoch': 10.10, 'throughput': 7327.07}
+
+[INFO|2025-02-17 12:52:44] logging.py:157 >> {'loss': 0.0020, 'learning_rate': 4.3064e-06, 'epoch': 10.20, 'throughput': 7335.94}
+
+[INFO|2025-02-17 12:53:04] logging.py:157 >> {'loss': 0.0023, 'learning_rate': 4.1388e-06, 'epoch': 10.29, 'throughput': 7344.64}
+
+[INFO|2025-02-17 12:53:24] logging.py:157 >> {'loss': 0.0020, 'learning_rate': 3.9737e-06, 'epoch': 10.39, 'throughput': 7352.70}
+
+[INFO|2025-02-17 12:53:43] logging.py:157 >> {'loss': 0.0021, 'learning_rate': 3.8111e-06, 'epoch': 10.49, 'throughput': 7360.36}
+
+[INFO|2025-02-17 12:54:03] logging.py:157 >> {'loss': 0.0025, 'learning_rate': 3.6511e-06, 'epoch': 10.59, 'throughput': 7369.18}
+
+[INFO|2025-02-17 12:54:22] logging.py:157 >> {'loss': 0.0021, 'learning_rate': 3.4938e-06, 'epoch': 10.68, 'throughput': 7377.11}
+
+[INFO|2025-02-17 12:54:42] logging.py:157 >> {'loss': 0.0017, 'learning_rate': 3.3393e-06, 'epoch': 10.78, 'throughput': 7384.90}
+
+[INFO|2025-02-17 12:55:02] logging.py:157 >> {'loss': 0.0018, 'learning_rate': 3.1875e-06, 'epoch': 10.88, 'throughput': 7392.59}
+
+[INFO|2025-02-17 12:55:21] logging.py:157 >> {'loss': 0.0018, 'learning_rate': 3.0387e-06, 'epoch': 10.98, 'throughput': 7400.64}
+
+[INFO|2025-02-17 12:55:37] logging.py:157 >> {'loss': 0.0014, 'learning_rate': 2.8927e-06, 'epoch': 11.06, 'throughput': 7406.36}
+
+[INFO|2025-02-17 12:55:57] logging.py:157 >> {'loss': 0.0013, 'learning_rate': 2.7498e-06, 'epoch': 11.16, 'throughput': 7413.44}
+
+[INFO|2025-02-17 12:56:16] logging.py:157 >> {'loss': 0.0014, 'learning_rate': 2.6099e-06, 'epoch': 11.25, 'throughput': 7420.91}
+
+[INFO|2025-02-17 12:56:36] logging.py:157 >> {'loss': 0.0013, 'learning_rate': 2.4731e-06, 'epoch': 11.35, 'throughput': 7427.77}
+
+[INFO|2025-02-17 12:56:56] logging.py:157 >> {'loss': 0.0013, 'learning_rate': 2.3396e-06, 'epoch': 11.45, 'throughput': 7434.56}
+
+[INFO|2025-02-17 12:57:15] logging.py:157 >> {'loss': 0.0012, 'learning_rate': 2.2092e-06, 'epoch': 11.55, 'throughput': 7441.93}
+
+[INFO|2025-02-17 12:57:15] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-600
+
+[INFO|2025-02-17 12:57:15] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-600/config.json
+
+[INFO|2025-02-17 12:57:15] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-600/generation_config.json
+
+[INFO|2025-02-17 12:57:37] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-600/model.safetensors.index.json.
+
+[INFO|2025-02-17 12:57:37] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-600/tokenizer_config.json
+
+[INFO|2025-02-17 12:57:37] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-600/special_tokens_map.json
+
+[INFO|2025-02-17 12:58:39] logging.py:157 >> {'loss': 0.0013, 'learning_rate': 2.0821e-06, 'epoch': 11.65, 'throughput': 7271.40}
+
+[INFO|2025-02-17 12:58:58] logging.py:157 >> {'loss': 0.0013, 'learning_rate': 1.9584e-06, 'epoch': 11.74, 'throughput': 7279.67}
+
+[INFO|2025-02-17 12:59:17] logging.py:157 >> {'loss': 0.0014, 'learning_rate': 1.8380e-06, 'epoch': 11.84, 'throughput': 7287.84}
+
+[INFO|2025-02-17 12:59:37] logging.py:157 >> {'loss': 0.0012, 'learning_rate': 1.7211e-06, 'epoch': 11.94, 'throughput': 7295.63}
+
+[INFO|2025-02-17 12:59:53] logging.py:157 >> {'loss': 0.0014, 'learning_rate': 1.6077e-06, 'epoch': 12.02, 'throughput': 7301.92}
+
+[INFO|2025-02-17 13:00:13] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 1.4978e-06, 'epoch': 12.12, 'throughput': 7308.89}
+
+[INFO|2025-02-17 13:00:32] logging.py:157 >> {'loss': 0.0012, 'learning_rate': 1.3915e-06, 'epoch': 12.22, 'throughput': 7316.86}
+
+[INFO|2025-02-17 13:00:52] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 1.2889e-06, 'epoch': 12.31, 'throughput': 7324.06}
+
+[INFO|2025-02-17 13:01:11] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 1.1899e-06, 'epoch': 12.41, 'throughput': 7331.13}
+
+[INFO|2025-02-17 13:01:31] logging.py:157 >> {'loss': 0.0012, 'learning_rate': 1.0946e-06, 'epoch': 12.51, 'throughput': 7338.53}
+
+[INFO|2025-02-17 13:01:50] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 1.0031e-06, 'epoch': 12.61, 'throughput': 7345.34}
+
+[INFO|2025-02-17 13:02:10] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 9.1535e-07, 'epoch': 12.70, 'throughput': 7352.25}
+
+[INFO|2025-02-17 13:02:30] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 8.3145e-07, 'epoch': 12.80, 'throughput': 7359.27}
+
+[INFO|2025-02-17 13:02:49] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 7.5141e-07, 'epoch': 12.90, 'throughput': 7365.77}
+
+[INFO|2025-02-17 13:03:09] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 6.7528e-07, 'epoch': 13.00, 'throughput': 7372.62}
+
+[INFO|2025-02-17 13:03:25] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 6.0307e-07, 'epoch': 13.08, 'throughput': 7377.30}
+
+[INFO|2025-02-17 13:03:44] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 5.3483e-07, 'epoch': 13.18, 'throughput': 7383.80}
+
+[INFO|2025-02-17 13:04:04] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 4.7058e-07, 'epoch': 13.27, 'throughput': 7390.60}
+
+[INFO|2025-02-17 13:04:23] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 4.1035e-07, 'epoch': 13.37, 'throughput': 7396.91}
+
+[INFO|2025-02-17 13:04:43] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 3.5415e-07, 'epoch': 13.47, 'throughput': 7402.92}
+
+[INFO|2025-02-17 13:04:43] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-700
+
+[INFO|2025-02-17 13:04:43] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-700/config.json
+
+[INFO|2025-02-17 13:04:43] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-700/generation_config.json
+
+[INFO|2025-02-17 13:05:09] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-700/model.safetensors.index.json.
+
+[INFO|2025-02-17 13:05:09] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-700/tokenizer_config.json
+
+[INFO|2025-02-17 13:05:09] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-700/special_tokens_map.json
+
+[INFO|2025-02-17 13:06:06] logging.py:157 >> {'loss': 0.0009, 'learning_rate': 3.0203e-07, 'epoch': 13.57, 'throughput': 7259.08}
+
+[INFO|2025-02-17 13:06:25] logging.py:157 >> {'loss': 0.0011, 'learning_rate': 2.5399e-07, 'epoch': 13.67, 'throughput': 7266.36}
+
+[INFO|2025-02-17 13:06:45] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 2.1007e-07, 'epoch': 13.76, 'throughput': 7273.17}
+
+[INFO|2025-02-17 13:07:05] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 1.7027e-07, 'epoch': 13.86, 'throughput': 7280.07}
+
+[INFO|2025-02-17 13:07:24] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 1.3461e-07, 'epoch': 13.96, 'throughput': 7286.52}
+
+[INFO|2025-02-17 13:07:40] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 1.0312e-07, 'epoch': 14.04, 'throughput': 7291.00}
+
+[INFO|2025-02-17 13:08:00] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 7.5795e-08, 'epoch': 14.14, 'throughput': 7297.68}
+
+[INFO|2025-02-17 13:08:19] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 5.2656e-08, 'epoch': 14.23, 'throughput': 7304.61}
+
+[INFO|2025-02-17 13:08:39] logging.py:157 >> {'loss': 0.0009, 'learning_rate': 3.3710e-08, 'epoch': 14.33, 'throughput': 7311.13}
+
+[INFO|2025-02-17 13:08:58] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 1.8967e-08, 'epoch': 14.43, 'throughput': 7317.57}
+
+[INFO|2025-02-17 13:09:18] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 8.4311e-09, 'epoch': 14.53, 'throughput': 7322.85}
+
+[INFO|2025-02-17 13:09:38] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 2.1080e-09, 'epoch': 14.63, 'throughput': 7328.91}
+
+[INFO|2025-02-17 13:09:57] logging.py:157 >> {'loss': 0.0010, 'learning_rate': 0.0000e+00, 'epoch': 14.72, 'throughput': 7335.46}
+
+[INFO|2025-02-17 13:09:57] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-765
+
+[INFO|2025-02-17 13:09:57] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-765/config.json
+
+[INFO|2025-02-17 13:09:57] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-765/generation_config.json
+
+[INFO|2025-02-17 13:10:18] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-765/model.safetensors.index.json.
+
+[INFO|2025-02-17 13:10:18] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-765/tokenizer_config.json
+
+[INFO|2025-02-17 13:10:18] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/checkpoint-765/special_tokens_map.json
+
+[INFO|2025-02-17 13:10:50] trainer.py:2643 >> 
+
+Training completed. Do not forget to share your model on huggingface.co/models =)
+
+
+
+[INFO|2025-02-17 13:10:50] trainer.py:3910 >> Saving model checkpoint to saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10
+
+[INFO|2025-02-17 13:10:50] configuration_utils.py:420 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/config.json
+
+[INFO|2025-02-17 13:10:50] configuration_utils.py:909 >> Configuration saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/generation_config.json
+
+[INFO|2025-02-17 13:11:12] modeling_utils.py:2996 >> The model is bigger than the maximum size per checkpoint (5GB) and is going to be split in 2 checkpoint shards. You can find where each parameters has been saved in the index located at saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/model.safetensors.index.json.
+
+[INFO|2025-02-17 13:11:12] tokenization_utils_base.py:2491 >> tokenizer config file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/tokenizer_config.json
+
+[INFO|2025-02-17 13:11:12] tokenization_utils_base.py:2500 >> Special tokens file saved in saves/Qwen2.5-1.5B/full/train_2025-02-17-12-11-10/special_tokens_map.json
+
+[WARNING|2025-02-17 13:11:13] logging.py:162 >> No metric eval_loss to plot.
+
+[WARNING|2025-02-17 13:11:13] logging.py:162 >> No metric eval_accuracy to plot.
+
+[INFO|2025-02-17 13:11:13] modelcard.py:449 >> Dropping the following result as it does not have all the necessary fields:
+{'task': {'name': 'Causal Language Modeling', 'type': 'text-generation'}}
+