jbcs2025_llama31_8b-balanced-C5 / run_experiment.log
abarbosa's picture
Pushing fine-tuned model to Hugging Face Hub
6364f72 verified
raw
history blame
64.3 kB
[2025-03-17 22:47:57,195][__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/2025-03-17/18-44-41
experiments:
model:
name: meta-llama/Llama-3.1-8B
type: llama31_classification_lora
num_labels: 6
output_dir: ./results/llama31_8b-balanced/C5
logging_dir: ./logs/llama31_8b-balanced/C5
best_model_dir: ./results/llama31_8b-balanced/C5/best_model
lora_r: 8
lora_dropout: 0.05
lora_alpha: 16
lora_target_modules: all-linear
dataset:
grade_index: 4
training_id: llama31_8b-balanced-C5
training_params:
weight_decay: 0.01
warmup_ratio: 0.1
learning_rate: 5.0e-05
train_batch_size: 1
eval_batch_size: 2
gradient_accumulation_steps: 16
gradient_checkpointing: false
[2025-03-17 22:47:57,197][__main__][INFO] - Starting the Fine Tuning training process.
[2025-03-17 22:48:03,239][transformers.tokenization_utils_base][INFO] - loading file tokenizer.json from cache at /media/data/tmp/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/tokenizer.json
[2025-03-17 22:48:03,239][transformers.tokenization_utils_base][INFO] - loading file tokenizer.model from cache at None
[2025-03-17 22:48:03,239][transformers.tokenization_utils_base][INFO] - loading file added_tokens.json from cache at None
[2025-03-17 22:48:03,239][transformers.tokenization_utils_base][INFO] - loading file special_tokens_map.json from cache at /media/data/tmp/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/special_tokens_map.json
[2025-03-17 22:48:03,239][transformers.tokenization_utils_base][INFO] - loading file tokenizer_config.json from cache at /media/data/tmp/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/tokenizer_config.json
[2025-03-17 22:48:03,239][transformers.tokenization_utils_base][INFO] - loading file chat_template.jinja from cache at None
[2025-03-17 22:48:03,637][transformers.tokenization_utils_base][INFO] - Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.
[2025-03-17 22:48:03,735][__main__][INFO] - Tokenizer function parameters- Padding:longest; Truncation: False
[2025-03-17 22:48:05,848][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /media/data/tmp/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-17 22:48:05,849][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"_name_or_path": "meta-llama/Llama-3.1-8B",
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"id2label": {
"0": 0,
"1": 40,
"2": 80,
"3": 120,
"4": 160,
"5": 200
},
"initializer_range": 0.02,
"intermediate_size": 14336,
"label2id": {
"0": 0,
"40": 1,
"80": 2,
"120": 3,
"160": 4,
"200": 5
},
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-17 22:48:05,879][transformers.modeling_utils][INFO] - loading weights file model.safetensors from cache at /media/data/tmp/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/model.safetensors.index.json
[2025-03-17 22:48:05,880][transformers.modeling_utils][INFO] - Will use torch_dtype=torch.bfloat16 as defined in model's config object
[2025-03-17 22:48:05,880][transformers.modeling_utils][INFO] - Instantiating LlamaForSequenceClassification model under default dtype torch.bfloat16.
[2025-03-17 22:48:27,762][transformers.modeling_utils][INFO] - Some weights of the model checkpoint at meta-llama/Llama-3.1-8B were not used when initializing LlamaForSequenceClassification: {'lm_head.weight'}
- This IS expected if you are initializing LlamaForSequenceClassification 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 LlamaForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
[2025-03-17 22:48:27,762][transformers.modeling_utils][WARNING] - Some weights of LlamaForSequenceClassification were not initialized from the model checkpoint at meta-llama/Llama-3.1-8B 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.
[2025-03-17 22:48:30,014][__main__][INFO] - None
[2025-03-17 22:48:30,015][transformers.training_args][INFO] - PyTorch: setting up devices
[2025-03-17 22:48:30,024][__main__][INFO] - Total steps: 620. Number of warmup steps: 62
[2025-03-17 22:48:30,031][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.
[2025-03-17 22:48:30,056][transformers.trainer][INFO] - Using auto half precision backend
[2025-03-17 22:48:30,057][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.
[2025-03-17 22:48:30,104][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-17 22:48:30,128][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-17 22:48:30,128][transformers.trainer][INFO] - Num examples = 132
[2025-03-17 22:48:30,129][transformers.trainer][INFO] - Batch size = 2
[2025-03-17 22:49:44,277][transformers][INFO] - {'accuracy': 0.24242424242424243, 'RMSE': 64.8541487189571, 'QWK': 0.03236437383809365, 'HDIV': 0.12878787878787878, 'Macro_F1': 0.09051808406647116, 'Micro_F1': 0.24242424242424243, 'Weighted_F1': 0.13990935750466543}
[2025-03-17 22:49:44,505][transformers.trainer][INFO] - The following columns in the training set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - ***** Running training *****
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - Num examples = 500
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - Num Epochs = 20
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - Instantaneous batch size per device = 1
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - Total train batch size (w. parallel, distributed & accumulation) = 16
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - Gradient Accumulation steps = 16
[2025-03-17 22:49:44,569][transformers.trainer][INFO] - Total optimization steps = 620
[2025-03-17 22:49:44,574][transformers.trainer][INFO] - Number of trainable parameters = 20,996,096
[2025-03-17 23:06:28,642][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-17 23:06:28,646][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-17 23:06:28,646][transformers.trainer][INFO] - Num examples = 132
[2025-03-17 23:06:28,646][transformers.trainer][INFO] - Batch size = 2
[2025-03-17 23:07:44,336][transformers][INFO] - {'accuracy': 0.25, 'RMSE': 75.3979342213597, 'QWK': -0.16499811817839682, 'HDIV': 0.23484848484848486, 'Macro_F1': 0.09090554254488681, 'Micro_F1': 0.25, 'Weighted_F1': 0.14577389823291462}
[2025-03-17 23:07:44,341][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-32
[2025-03-17 23:07:44,833][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-17 23:07:44,834][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-17 23:24:32,338][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-17 23:24:32,341][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-17 23:24:32,341][transformers.trainer][INFO] - Num examples = 132
[2025-03-17 23:24:32,342][transformers.trainer][INFO] - Batch size = 2
[2025-03-17 23:25:47,988][transformers][INFO] - {'accuracy': 0.23484848484848486, 'RMSE': 66.14950926316519, 'QWK': 0.03725553580087271, 'HDIV': 0.12878787878787878, 'Macro_F1': 0.09814814814814815, 'Micro_F1': 0.23484848484848486, 'Weighted_F1': 0.1525252525252525}
[2025-03-17 23:25:47,991][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-64
[2025-03-17 23:25:48,457][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-17 23:25:48,457][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-17 23:25:49,958][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-32] due to args.save_total_limit
[2025-03-17 23:42:33,990][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-17 23:42:33,992][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-17 23:42:33,992][transformers.trainer][INFO] - Num examples = 132
[2025-03-17 23:42:33,993][transformers.trainer][INFO] - Batch size = 2
[2025-03-17 23:43:49,537][transformers][INFO] - {'accuracy': 0.16666666666666666, 'RMSE': 88.00826407477052, 'QWK': -0.08383018092105265, 'HDIV': 0.303030303030303, 'Macro_F1': 0.11598793363499245, 'Micro_F1': 0.16666666666666666, 'Weighted_F1': 0.13846153846153844}
[2025-03-17 23:43:49,540][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-96
[2025-03-17 23:43:50,287][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-17 23:43:50,288][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 00:00:35,785][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 00:00:35,787][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 00:00:35,787][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 00:00:35,787][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 00:01:51,406][transformers][INFO] - {'accuracy': 0.2727272727272727, 'RMSE': 65.22687678055308, 'QWK': 0.2237115474834126, 'HDIV': 0.1515151515151515, 'Macro_F1': 0.17972524968423245, 'Micro_F1': 0.2727272727272727, 'Weighted_F1': 0.2201439213416244}
[2025-03-18 00:01:51,409][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-128
[2025-03-18 00:01:51,843][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 00:01:51,844][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 00:01:53,684][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-64] due to args.save_total_limit
[2025-03-18 00:01:53,710][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-96] due to args.save_total_limit
[2025-03-18 00:18:38,011][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 00:18:38,014][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 00:18:38,014][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 00:18:38,014][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 00:19:53,683][transformers][INFO] - {'accuracy': 0.21212121212121213, 'RMSE': 70.58199615920608, 'QWK': 0.047742750824966707, 'HDIV': 0.1515151515151515, 'Macro_F1': 0.13397158397158396, 'Micro_F1': 0.21212121212121213, 'Weighted_F1': 0.17766879585061401}
[2025-03-18 00:19:53,686][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-160
[2025-03-18 00:19:54,138][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 00:19:54,138][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 00:36:40,307][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 00:36:40,310][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 00:36:40,311][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 00:36:40,311][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 00:37:55,888][transformers][INFO] - {'accuracy': 0.24242424242424243, 'RMSE': 65.59748701764651, 'QWK': 0.20500814332247552, 'HDIV': 0.11363636363636365, 'Macro_F1': 0.19603237333577708, 'Micro_F1': 0.24242424242424243, 'Weighted_F1': 0.24172658841086073}
[2025-03-18 00:37:55,892][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-192
[2025-03-18 00:37:56,578][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 00:37:56,579][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 00:37:58,359][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-160] due to args.save_total_limit
[2025-03-18 00:54:43,001][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 00:54:43,004][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 00:54:43,004][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 00:54:43,004][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 00:55:58,683][transformers][INFO] - {'accuracy': 0.3333333333333333, 'RMSE': 61.20011883529758, 'QWK': 0.3801592608352077, 'HDIV': 0.09090909090909094, 'Macro_F1': 0.29144402966866734, 'Micro_F1': 0.3333333333333333, 'Weighted_F1': 0.32806062697367044}
[2025-03-18 00:55:58,685][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-224
[2025-03-18 00:56:00,042][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 00:56:00,043][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 00:56:01,976][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-128] due to args.save_total_limit
[2025-03-18 00:56:01,996][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-192] due to args.save_total_limit
[2025-03-18 01:12:47,695][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 01:12:47,698][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 01:12:47,699][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 01:12:47,699][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 01:14:03,433][transformers][INFO] - {'accuracy': 0.1893939393939394, 'RMSE': 76.8311805390904, 'QWK': 0.05107463391591882, 'HDIV': 0.21969696969696972, 'Macro_F1': 0.13419994690332612, 'Micro_F1': 0.1893939393939394, 'Weighted_F1': 0.16065685872262273}
[2025-03-18 01:14:03,436][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-256
[2025-03-18 01:14:04,398][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 01:14:04,399][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 01:30:51,890][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 01:30:51,893][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 01:30:51,893][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 01:30:51,893][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 01:32:07,555][transformers][INFO] - {'accuracy': 0.32575757575757575, 'RMSE': 62.1825270205921, 'QWK': 0.36911183027687877, 'HDIV': 0.11363636363636365, 'Macro_F1': 0.22144078750334958, 'Micro_F1': 0.32575757575757575, 'Weighted_F1': 0.27548695925061367}
[2025-03-18 01:32:07,559][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-288
[2025-03-18 01:32:08,212][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 01:32:08,213][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 01:32:09,335][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-256] due to args.save_total_limit
[2025-03-18 01:48:53,732][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 01:48:53,737][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 01:48:53,737][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 01:48:53,737][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 01:50:09,451][transformers][INFO] - {'accuracy': 0.30303030303030304, 'RMSE': 64.10219114110616, 'QWK': 0.2570233114166167, 'HDIV': 0.13636363636363635, 'Macro_F1': 0.22320469355717623, 'Micro_F1': 0.30303030303030304, 'Weighted_F1': 0.27005319566718955}
[2025-03-18 01:50:09,454][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-320
[2025-03-18 01:50:10,446][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 01:50:10,447][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 01:50:12,396][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-288] due to args.save_total_limit
[2025-03-18 02:06:58,515][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 02:06:58,518][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 02:06:58,518][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 02:06:58,518][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 02:08:14,181][transformers][INFO] - {'accuracy': 0.3181818181818182, 'RMSE': 60.0, 'QWK': 0.3805069211807093, 'HDIV': 0.10606060606060608, 'Macro_F1': 0.22958520739630187, 'Micro_F1': 0.3181818181818182, 'Weighted_F1': 0.28138657943755396}
[2025-03-18 02:08:14,185][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-352
[2025-03-18 02:08:14,930][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 02:08:14,931][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 02:08:16,742][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-224] due to args.save_total_limit
[2025-03-18 02:08:16,766][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-320] due to args.save_total_limit
[2025-03-18 02:25:01,393][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 02:25:01,396][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 02:25:01,396][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 02:25:01,396][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 02:26:16,975][transformers][INFO] - {'accuracy': 0.26515151515151514, 'RMSE': 60.201681240995455, 'QWK': 0.39525619024270653, 'HDIV': 0.08333333333333337, 'Macro_F1': 0.24763387289624172, 'Micro_F1': 0.26515151515151514, 'Weighted_F1': 0.25899349085256135}
[2025-03-18 02:26:16,979][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-384
[2025-03-18 02:26:17,987][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 02:26:17,988][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 02:26:19,080][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-352] due to args.save_total_limit
[2025-03-18 02:43:05,751][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 02:43:05,754][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 02:43:05,754][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 02:43:05,754][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 02:44:21,436][transformers][INFO] - {'accuracy': 0.2878787878787879, 'RMSE': 58.981250230796896, 'QWK': 0.38916478555304734, 'HDIV': 0.0757575757575758, 'Macro_F1': 0.23379159344505882, 'Micro_F1': 0.2878787878787879, 'Weighted_F1': 0.23514724849108284}
[2025-03-18 02:44:21,440][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-416
[2025-03-18 02:44:22,161][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 02:44:22,161][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 03:01:10,574][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 03:01:10,578][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 03:01:10,578][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 03:01:10,578][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 03:02:26,292][transformers][INFO] - {'accuracy': 0.30303030303030304, 'RMSE': 66.51497894641795, 'QWK': 0.34061422236820504, 'HDIV': 0.13636363636363635, 'Macro_F1': 0.22355989731978676, 'Micro_F1': 0.30303030303030304, 'Weighted_F1': 0.2666766786880561}
[2025-03-18 03:02:26,296][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-448
[2025-03-18 03:02:27,111][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 03:02:27,112][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 03:02:28,990][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-416] due to args.save_total_limit
[2025-03-18 03:19:16,010][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 03:19:16,013][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 03:19:16,013][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 03:19:16,013][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 03:20:31,669][transformers][INFO] - {'accuracy': 0.32575757575757575, 'RMSE': 59.39084716749482, 'QWK': 0.4298860128235573, 'HDIV': 0.08333333333333337, 'Macro_F1': 0.26813891813891816, 'Micro_F1': 0.32575757575757575, 'Weighted_F1': 0.31763971763971766}
[2025-03-18 03:20:31,673][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-480
[2025-03-18 03:20:32,623][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 03:20:32,624][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 03:20:34,492][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-384] due to args.save_total_limit
[2025-03-18 03:20:34,513][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-448] due to args.save_total_limit
[2025-03-18 03:37:21,087][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 03:37:21,090][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 03:37:21,090][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 03:37:21,090][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 03:38:36,834][transformers][INFO] - {'accuracy': 0.2878787878787879, 'RMSE': 58.15340215170718, 'QWK': 0.42696209622207004, 'HDIV': 0.05303030303030298, 'Macro_F1': 0.20220652686992957, 'Micro_F1': 0.2878787878787879, 'Weighted_F1': 0.23529808964325488}
[2025-03-18 03:38:36,837][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-512
[2025-03-18 03:38:37,400][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 03:38:37,401][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 03:55:25,879][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 03:55:25,882][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 03:55:25,883][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 03:55:25,883][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 03:56:41,664][transformers][INFO] - {'accuracy': 0.2727272727272727, 'RMSE': 59.797638546664835, 'QWK': 0.41984505363528, 'HDIV': 0.08333333333333337, 'Macro_F1': 0.2372466413177874, 'Micro_F1': 0.2727272727272727, 'Weighted_F1': 0.2692383741516407}
[2025-03-18 03:56:41,668][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-544
[2025-03-18 03:56:42,430][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 03:56:42,431][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 03:56:44,288][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-512] due to args.save_total_limit
[2025-03-18 04:13:31,366][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 04:13:31,371][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 04:13:31,371][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 04:13:31,371][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 04:14:47,009][transformers][INFO] - {'accuracy': 0.2727272727272727, 'RMSE': 60.0, 'QWK': 0.3971953994710622, 'HDIV': 0.06818181818181823, 'Macro_F1': 0.2294526863492381, 'Micro_F1': 0.2727272727272727, 'Weighted_F1': 0.2691151703690889}
[2025-03-18 04:14:47,012][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-576
[2025-03-18 04:14:47,483][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 04:14:47,484][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 04:14:48,675][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-544] due to args.save_total_limit
[2025-03-18 04:31:35,446][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 04:31:35,449][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 04:31:35,449][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 04:31:35,449][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 04:32:51,159][transformers][INFO] - {'accuracy': 0.3181818181818182, 'RMSE': 58.981250230796896, 'QWK': 0.4287696019300361, 'HDIV': 0.0757575757575758, 'Macro_F1': 0.2680758090649689, 'Micro_F1': 0.3181818181818182, 'Weighted_F1': 0.31353410215198835}
[2025-03-18 04:32:51,162][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-608
[2025-03-18 04:32:51,600][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 04:32:51,601][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 04:32:53,465][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-576] due to args.save_total_limit
[2025-03-18 04:39:20,353][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-620
[2025-03-18 04:39:20,897][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 04:39:20,898][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 04:39:22,802][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-608] due to args.save_total_limit
[2025-03-18 04:39:22,821][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 04:39:22,824][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 04:39:22,824][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 04:39:22,824][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 04:40:38,446][transformers][INFO] - {'accuracy': 0.3409090909090909, 'RMSE': 57.94459213400123, 'QWK': 0.4536489151873767, 'HDIV': 0.08333333333333337, 'Macro_F1': 0.28543976348854394, 'Micro_F1': 0.3409090909090909, 'Weighted_F1': 0.3336524894174562}
[2025-03-18 04:40:38,449][transformers.trainer][INFO] - Saving model checkpoint to /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-620
[2025-03-18 04:40:39,117][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 04:40:39,118][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 04:40:41,068][transformers.trainer][INFO] - Deleting older checkpoint [/workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-480] due to args.save_total_limit
[2025-03-18 04:40:41,091][transformers.trainer][INFO] -
Training completed. Do not forget to share your model on huggingface.co/models =)
[2025-03-18 04:40:41,092][transformers.trainer][INFO] - Loading best model from /workspace/jbcs2025/outputs/2025-03-17/22-47-57/results/llama31_8b-balanced/C5/checkpoint-620 (score: 0.4536489151873767).
[2025-03-18 04:40:47,380][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 04:40:47,383][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 04:40:47,383][transformers.trainer][INFO] - Num examples = 132
[2025-03-18 04:40:47,383][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 04:42:02,861][transformers][INFO] - {'accuracy': 0.3409090909090909, 'RMSE': 57.94459213400123, 'QWK': 0.4536489151873767, 'HDIV': 0.08333333333333337, 'Macro_F1': 0.28543976348854394, 'Micro_F1': 0.3409090909090909, 'Weighted_F1': 0.3336524894174562}
[2025-03-18 04:42:02,865][__main__][INFO] - Training completed successfully.
[2025-03-18 04:42:02,865][__main__][INFO] - Running on Test
[2025-03-18 04:42:02,865][transformers.trainer][INFO] - The following columns in the evaluation set don't have a corresponding argument in `PeftModelForSequenceClassification.forward` and have been ignored: supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt. If supporting_text, essay_text, id_prompt, id, reference, grades, essay_year, prompt are not expected by `PeftModelForSequenceClassification.forward`, you can safely ignore this message.
[2025-03-18 04:42:02,868][transformers.trainer][INFO] -
***** Running Evaluation *****
[2025-03-18 04:42:02,868][transformers.trainer][INFO] - Num examples = 138
[2025-03-18 04:42:02,868][transformers.trainer][INFO] - Batch size = 2
[2025-03-18 04:43:25,959][transformers][INFO] - {'accuracy': 0.2898550724637681, 'RMSE': 62.60064360459999, 'QWK': 0.400524367674275, 'HDIV': 0.10144927536231885, 'Macro_F1': 0.2873977873977874, 'Micro_F1': 0.2898550724637681, 'Weighted_F1': 0.27963119810945897}
[2025-03-18 04:43:25,962][transformers.trainer][INFO] - Saving model checkpoint to ./results/llama31_8b-balanced/C5/best_model
[2025-03-18 04:43:26,364][transformers.configuration_utils][INFO] - loading configuration file config.json from cache at /root/.cache/huggingface/hub/models--meta-llama--Llama-3.1-8B/snapshots/d04e592bb4f6aa9cfee91e2e20afa771667e1d4b/config.json
[2025-03-18 04:43:26,364][transformers.configuration_utils][INFO] - Model config LlamaConfig {
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 128000,
"eos_token_id": 128001,
"head_dim": 128,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 131072,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 8.0,
"high_freq_factor": 4.0,
"low_freq_factor": 1.0,
"original_max_position_embeddings": 8192,
"rope_type": "llama3"
},
"rope_theta": 500000.0,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.49.0",
"use_cache": true,
"vocab_size": 128256
}
[2025-03-18 04:43:27,995][__main__][INFO] - Fine Tuning Finished.