jbcs2025_phi4-balanced-C4 / run_experiment.log
abarbosa's picture
Pushing fine-tuned model to Hugging Face Hub
6c824c1 verified
[2025-03-23 20:41:58,914][__main__][INFO] - cache_dir: /media/data/tmp
dataset:
name: kamel-usp/aes_enem_dataset
split: JBCS2025
training_params:
seed: 42
num_train_epochs: 20
logging_steps: 100
metric_for_best_model: QWK
bf16: true
post_training_results:
model_path: /workspace/jbcs2025/outputs/2025-03-23/18-55-03
experiments:
model:
name: microsoft/phi-4
type: phi4_classification_lora
num_labels: 6
output_dir: ./results/phi4-balanced/C4
logging_dir: ./logs/phi4-balanced/C4
best_model_dir: ./results/phi4-balanced/C4/best_model
lora_r: 8
lora_dropout: 0.05
lora_alpha: 16
lora_target_modules: all-linear
dataset:
grade_index: 3
training_id: phi4-balanced-C4
training_params:
weight_decay: 0.01
warmup_ratio: 0.1
learning_rate: 5.0e-05
train_batch_size: 1
eval_batch_size: 16
gradient_accumulation_steps: 16
gradient_checkpointing: false
[2025-03-23 20:41:58,916][__main__][INFO] - Starting the Fine Tuning training process.
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file vocab.json from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/vocab.json
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file merges.txt from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/merges.txt
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file tokenizer.json from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/tokenizer.json
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file added_tokens.json from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/added_tokens.json
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file special_tokens_map.json from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/special_tokens_map.json
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file tokenizer_config.json from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/tokenizer_config.json
[2025-03-23 20:42:03,790][transformers.tokenization_utils_base][INFO] - loading file chat_template.jinja from cache at None
[2025-03-23 20:42:03,944][__main__][INFO] - Tokenizer function parameters- Padding:longest; Truncation: False
[2025-03-23 20:42:04,348][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 20:42:04,349][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"id2label": {
"0": 0,
"1": 40,
"2": 80,
"3": 120,
"4": 160,
"5": 200
},
"initializer_range": 0.02,
"intermediate_size": 17920,
"label2id": {
"0": 0,
"40": 1,
"80": 2,
"120": 3,
"160": 4,
"200": 5
},
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 20:42:04,371][transformers.modeling_utils][INFO] - loading weights file model.safetensors from cache at /media/data/tmp/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/model.safetensors.index.json
[2025-03-23 20:42:04,372][transformers.modeling_utils][INFO] - Will use torch_dtype=torch.bfloat16 as defined in model's config object
[2025-03-23 20:42:04,372][transformers.modeling_utils][INFO] - Instantiating Phi3ForSequenceClassification model under default dtype torch.bfloat16.
[2025-03-23 20:42:25,935][transformers.modeling_utils][INFO] - Some weights of the model checkpoint at microsoft/phi-4 were not used when initializing Phi3ForSequenceClassification: ['lm_head.weight']
- This IS expected if you are initializing Phi3ForSequenceClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
- This IS NOT expected if you are initializing Phi3ForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
[2025-03-23 20:42:25,936][transformers.modeling_utils][WARNING] - Some weights of Phi3ForSequenceClassification were not initialized from the model checkpoint at microsoft/phi-4 and are newly initialized: ['score.weight']
You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.
[2025-03-23 20:42:28,026][__main__][INFO] - None
[2025-03-23 20:42:28,029][transformers.training_args][INFO] - PyTorch: setting up devices
[2025-03-23 20:42:28,068][__main__][INFO] - Total steps: 620. Number of warmup steps: 62
[2025-03-23 20:42:28,077][transformers.trainer][INFO] - You have loaded a model on multiple GPUs. `is_model_parallel` attribute will be force-set to `True` to avoid any unexpected behavior such as device placement mismatching.
[2025-03-23 20:42:28,100][transformers.trainer][INFO] - Using auto half precision backend
[2025-03-23 20:42:28,101][transformers.trainer][WARNING] - No label_names provided for model class `PeftModelForSequenceClassification`. 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.
[2025-03-23 20:42:28,121][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 20:42:28,133][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 20:42:28,134][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 20:42:28,134][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 20:42:54,626][transformers][INFO] - {'accuracy': 0.36363636363636365, 'RMSE': 36.680438185149114, 'QWK': 0.0, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.10666666666666666, 'Micro_F1': 0.36363636363636365, 'Weighted_F1': 0.19393939393939394, 'Macro_F1_(ignoring_nan)': np.float64(0.5333333333333333)}
[2025-03-23 20:42:54,629][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 20:42:54,875][transformers.trainer][INFO] - The following columns in the training set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - ***** Running training *****
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - Num examples = 500
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - Num Epochs = 20
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - Instantaneous batch size per device = 1
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - Total train batch size (w. parallel, distributed & accumulation) = 16
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - Gradient Accumulation steps = 16
[2025-03-23 20:42:54,906][transformers.trainer][INFO] - Total optimization steps = 620
[2025-03-23 20:42:54,908][transformers.trainer][INFO] - Number of trainable parameters = 27,883,520
[2025-03-23 20:49:54,343][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 20:49:54,346][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 20:49:54,346][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 20:49:54,346][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 20:50:20,840][transformers][INFO] - {'accuracy': 0.36363636363636365, 'RMSE': 36.680438185149114, 'QWK': 0.0, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.10666666666666666, 'Micro_F1': 0.36363636363636365, 'Weighted_F1': 0.19393939393939394, 'Macro_F1_(ignoring_nan)': np.float64(0.5333333333333333)}
[2025-03-23 20:50:20,840][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 20:50:20,843][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-32
[2025-03-23 20:50:21,157][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 20:50:21,158][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 20:57:21,498][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 20:57:21,501][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 20:57:21,501][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 20:57:21,502][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 20:57:47,876][transformers][INFO] - {'accuracy': 0.4621212121212121, 'RMSE': 33.93398225253192, 'QWK': 0.15453074433656944, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.18675097862089732, 'Micro_F1': 0.4621212121212121, 'Weighted_F1': 0.3838931318606115, 'Macro_F1_(ignoring_nan)': np.float64(0.4668774465522433)}
[2025-03-23 20:57:47,876][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 20:57:47,879][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-64
[2025-03-23 20:57:48,481][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 20:57:48,482][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 20:57:49,222][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-32] due to args.save_total_limit
[2025-03-23 21:04:48,693][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:04:48,695][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:04:48,695][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:04:48,695][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:05:15,051][transformers][INFO] - {'accuracy': 0.5378787878787878, 'RMSE': 30.944720996896347, 'QWK': 0.3034998664173123, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.228490990990991, 'Micro_F1': 0.5378787878787878, 'Weighted_F1': 0.47856947856947857, 'Macro_F1_(ignoring_nan)': np.float64(0.5712274774774775)}
[2025-03-23 21:05:15,051][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:05:15,054][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-96
[2025-03-23 21:05:15,393][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:05:15,394][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:05:16,040][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-64] due to args.save_total_limit
[2025-03-23 21:12:15,345][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:12:15,348][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:12:15,348][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:12:15,348][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:12:41,768][transformers][INFO] - {'accuracy': 0.5454545454545454, 'RMSE': 31.71845844395036, 'QWK': 0.3009188361408882, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.21155390793945011, 'Micro_F1': 0.5454545454545454, 'Weighted_F1': 0.46934591512904766, 'Macro_F1_(ignoring_nan)': np.float64(0.5288847698486253)}
[2025-03-23 21:12:41,769][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:12:41,772][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-128
[2025-03-23 21:12:42,252][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:12:42,252][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:19:42,312][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:19:42,315][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:19:42,315][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:19:42,315][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:20:08,584][transformers][INFO] - {'accuracy': 0.5606060606060606, 'RMSE': 28.91995221924885, 'QWK': 0.41987261146496824, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.28522175409256106, 'Micro_F1': 0.5606060606060606, 'Weighted_F1': 0.5337214163959735, 'Macro_F1_(ignoring_nan)': np.float64(0.47536959015426844)}
[2025-03-23 21:20:08,584][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:20:08,588][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-160
[2025-03-23 21:20:08,878][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:20:08,879][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:20:09,612][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-96] due to args.save_total_limit
[2025-03-23 21:20:09,650][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-128] due to args.save_total_limit
[2025-03-23 21:27:09,228][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:27:09,230][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:27:09,230][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:27:09,230][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:27:35,620][transformers][INFO] - {'accuracy': 0.5227272727272727, 'RMSE': 30.550504633038933, 'QWK': 0.3370727889381686, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.22424242424242422, 'Micro_F1': 0.5227272727272727, 'Weighted_F1': 0.4830119375573921, 'Macro_F1_(ignoring_nan)': np.float64(0.5606060606060606)}
[2025-03-23 21:27:35,620][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:27:35,623][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-192
[2025-03-23 21:27:35,955][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:27:35,955][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:34:35,458][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:34:35,460][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:34:35,460][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:34:35,460][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:35:01,816][transformers][INFO] - {'accuracy': 0.5075757575757576, 'RMSE': 34.289321599553055, 'QWK': 0.47082162340882794, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.30565610859728504, 'Micro_F1': 0.5075757575757576, 'Weighted_F1': 0.5056903880433292, 'Macro_F1_(ignoring_nan)': np.float64(0.5094268476621417)}
[2025-03-23 21:35:01,816][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:35:01,819][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-224
[2025-03-23 21:35:02,175][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:35:02,175][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:35:02,951][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-160] due to args.save_total_limit
[2025-03-23 21:35:02,986][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-192] due to args.save_total_limit
[2025-03-23 21:42:02,769][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:42:02,772][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:42:02,772][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:42:02,772][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:42:29,172][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.402048319327731, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.2524590163934426, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.5384997516145057, 'Macro_F1_(ignoring_nan)': np.float64(0.6311475409836065)}
[2025-03-23 21:42:29,172][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:42:29,175][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-256
[2025-03-23 21:42:29,463][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:42:29,464][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:49:29,409][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:49:29,411][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:49:29,411][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:49:29,411][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:49:55,727][transformers][INFO] - {'accuracy': 0.5984848484848485, 'RMSE': 28.4977404739606, 'QWK': 0.4918409561020455, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3331918158567775, 'Micro_F1': 0.5984848484848485, 'Weighted_F1': 0.5808866155157715, 'Macro_F1_(ignoring_nan)': np.float64(0.5553196930946291)}
[2025-03-23 21:49:55,727][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:49:55,731][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-288
[2025-03-23 21:49:56,023][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:49:56,023][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 21:49:56,763][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-224] due to args.save_total_limit
[2025-03-23 21:49:56,810][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-256] due to args.save_total_limit
[2025-03-23 21:56:56,055][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 21:56:56,057][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 21:56:56,057][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 21:56:56,057][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 21:57:22,544][transformers][INFO] - {'accuracy': 0.5984848484848485, 'RMSE': 28.4977404739606, 'QWK': 0.4536693847294292, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3165505532462971, 'Micro_F1': 0.5984848484848485, 'Weighted_F1': 0.5739240659034582, 'Macro_F1_(ignoring_nan)': np.float64(0.5275842554104951)}
[2025-03-23 21:57:22,545][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 21:57:22,548][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-320
[2025-03-23 21:57:22,867][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 21:57:22,868][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:04:23,489][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:04:23,491][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:04:23,491][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:04:23,491][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:04:49,840][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.44558071585098613, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.30430075187969924, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.5574436090225563, 'Macro_F1_(ignoring_nan)': np.float64(0.5071679197994987)}
[2025-03-23 22:04:49,841][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:04:49,844][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-352
[2025-03-23 22:04:50,184][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:04:50,185][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:04:50,944][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-320] due to args.save_total_limit
[2025-03-23 22:11:50,060][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:11:50,062][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:11:50,062][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:11:50,062][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:12:16,556][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.40126216145148574, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.2525806451612903, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.5380254154447703, 'Macro_F1_(ignoring_nan)': np.float64(0.6314516129032258)}
[2025-03-23 22:12:16,556][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:12:16,559][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-384
[2025-03-23 22:12:16,980][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:12:16,980][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:12:17,746][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-352] due to args.save_total_limit
[2025-03-23 22:19:16,915][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:19:16,918][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:19:16,918][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:19:16,918][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:19:43,392][transformers][INFO] - {'accuracy': 0.5681818181818182, 'RMSE': 29.336088024923512, 'QWK': 0.5011709601873537, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3423423989280405, 'Micro_F1': 0.5681818181818182, 'Weighted_F1': 0.5600188811042401, 'Macro_F1_(ignoring_nan)': np.float64(0.5705706648800676)}
[2025-03-23 22:19:43,393][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:19:43,396][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-416
[2025-03-23 22:19:43,732][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:19:43,733][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:19:44,597][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-288] due to args.save_total_limit
[2025-03-23 22:19:44,648][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-384] due to args.save_total_limit
[2025-03-23 22:26:45,026][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:26:45,029][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:26:45,029][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:26:45,029][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:27:11,429][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.47181628392484354, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3308645677161419, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.56964396589584, 'Macro_F1_(ignoring_nan)': np.float64(0.5514409461935699)}
[2025-03-23 22:27:11,430][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:27:11,432][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-448
[2025-03-23 22:34:21,711][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:34:21,714][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:34:21,714][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:34:21,714][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:34:48,074][transformers][INFO] - {'accuracy': 0.5757575757575758, 'RMSE': 29.128763250176767, 'QWK': 0.4815396700706991, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3335450732209876, 'Micro_F1': 0.5757575757575758, 'Weighted_F1': 0.5647085320791996, 'Macro_F1_(ignoring_nan)': np.float64(0.5559084553683126)}
[2025-03-23 22:34:48,075][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:34:48,076][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-480
[2025-03-23 22:34:48,464][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:34:48,465][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:34:49,169][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-448] due to args.save_total_limit
[2025-03-23 22:41:49,150][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:41:49,152][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:41:49,152][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:41:49,152][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:42:15,502][transformers][INFO] - {'accuracy': 0.5984848484848485, 'RMSE': 28.4977404739606, 'QWK': 0.47344605858537747, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.31150793650793646, 'Micro_F1': 0.5984848484848485, 'Weighted_F1': 0.5776815776815778, 'Macro_F1_(ignoring_nan)': np.float64(0.5191798941798941)}
[2025-03-23 22:42:15,503][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:42:15,506][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-512
[2025-03-23 22:42:16,316][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:42:16,317][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:42:17,055][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-480] due to args.save_total_limit
[2025-03-23 22:49:17,118][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:49:17,120][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:49:17,120][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:49:17,120][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:49:43,519][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.48693105002253256, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3384561403508772, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.5721424774056353, 'Macro_F1_(ignoring_nan)': np.float64(0.564093567251462)}
[2025-03-23 22:49:43,520][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:49:43,523][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-544
[2025-03-23 22:49:43,837][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:49:43,837][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:49:44,524][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-512] due to args.save_total_limit
[2025-03-23 22:56:43,629][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 22:56:43,631][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 22:56:43,631][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 22:56:43,631][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 22:57:09,913][transformers][INFO] - {'accuracy': 0.5984848484848485, 'RMSE': 28.4977404739606, 'QWK': 0.5040376850605652, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3435825471698113, 'Micro_F1': 0.5984848484848485, 'Weighted_F1': 0.5861349342481418, 'Macro_F1_(ignoring_nan)': np.float64(0.5726375786163521)}
[2025-03-23 22:57:09,914][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 22:57:09,916][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-576
[2025-03-23 22:57:11,876][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 22:57:11,877][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 22:57:12,612][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-416] due to args.save_total_limit
[2025-03-23 22:57:12,654][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-544] due to args.save_total_limit
[2025-03-23 23:04:12,951][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 23:04:12,954][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 23:04:12,954][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 23:04:12,954][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 23:04:39,310][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.47181628392484354, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3308645677161419, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.56964396589584, 'Macro_F1_(ignoring_nan)': np.float64(0.5514409461935699)}
[2025-03-23 23:04:39,311][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 23:04:39,313][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-608
[2025-03-23 23:04:39,620][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 23:04:39,621][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 23:07:22,340][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 23:07:22,342][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 23:07:22,342][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 23:07:22,342][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 23:07:48,658][transformers][INFO] - {'accuracy': 0.5833333333333334, 'RMSE': 28.91995221924885, 'QWK': 0.47181628392484354, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3308645677161419, 'Micro_F1': 0.5833333333333334, 'Weighted_F1': 0.56964396589584, 'Macro_F1_(ignoring_nan)': np.float64(0.5514409461935699)}
[2025-03-23 23:07:48,659][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 23:07:48,661][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-620
[2025-03-23 23:07:49,135][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 23:07:49,136][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 23:07:49,714][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-608] due to args.save_total_limit
[2025-03-23 23:07:49,754][transformers.trainer][INFO] -
Training completed. Do not forget to share your model on huggingface.co/models =)
[2025-03-23 23:07:49,754][transformers.trainer][INFO] - Loading best model from /workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-576 (score: 0.5040376850605652).
[2025-03-23 23:07:49,908][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-23/20-41-58/results/phi4-balanced/C4/checkpoint-620] due to args.save_total_limit
[2025-03-23 23:07:49,954][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 23:07:49,957][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 23:07:49,957][transformers.trainer][INFO] - Num examples = 132
[2025-03-23 23:07:49,957][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 23:08:16,365][transformers][INFO] - {'accuracy': 0.5984848484848485, 'RMSE': 28.4977404739606, 'QWK': 0.5040376850605652, 'HDIV': 0.007575757575757569, 'Macro_F1': 0.3435825471698113, 'Micro_F1': 0.5984848484848485, 'Weighted_F1': 0.5861349342481418, 'Macro_F1_(ignoring_nan)': np.float64(0.5726375786163521)}
[2025-03-23 23:08:16,368][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 23:08:16,369][__main__][INFO] - Training completed successfully.
[2025-03-23 23:08:16,369][__main__][INFO] - Running on Test
[2025-03-23 23:08:16,369][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id. If essay_text, prompt, id_prompt, reference, grades, supporting_text, essay_year, id are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-23 23:08:16,371][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-23 23:08:16,371][transformers.trainer][INFO] - Num examples = 138
[2025-03-23 23:08:16,371][transformers.trainer][INFO] - Batch size = 16
[2025-03-23 23:08:44,397][transformers][INFO] - {'accuracy': 0.7028985507246377, 'RMSE': 24.55399256179405, 'QWK': 0.579465541490858, 'HDIV': 0.007246376811594235, 'Macro_F1': 0.27468966776195697, 'Micro_F1': 0.7028985507246377, 'Weighted_F1': 0.6761155293109196, 'Macro_F1_(ignoring_nan)': np.float64(0.3296276013143483)}
[2025-03-23 23:08:44,398][tensorboardX.summary][INFO] - Summary name eval/Macro_F1_(ignoring_nan) is illegal; using eval/Macro_F1__ignoring_nan_ instead.
[2025-03-23 23:08:44,400][transformers.trainer][INFO] - Saving model checkpoint to ./results/phi4-balanced/C4/best_model
[2025-03-23 23:08:44,680][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--microsoft--phi-4/snapshots/187ef0342fff0eb3333be9f00389385e95ef0b61/config.json
[2025-03-23 23:08:44,681][transformers.configuration_utils][INFO] - Model config Phi3Config {
"architectures": [
"Phi3ForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"embd_pdrop": 0.0,
"eos_token_id": 100265,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 17920,
"max_position_embeddings": 16384,
"model_type": "phi3",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 10,
"original_max_position_embeddings": 16384,
"pad_token_id": 100349,
"partial_rotary_factor": 1.0,
"resid_pdrop": 0.0,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 250000,
"sliding_window": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.50.0",
"use_cache": true,
"vocab_size": 100352
}
[2025-03-23 23:08:45,141][__main__][INFO] - Fine Tuning Finished.