Hugging Face's logo

kamel-usp
/
jbcs2025_Llama-3.1-8B-llama31_classification_lora-C1-essay_only-r8

PEFT
Safetensors
Portuguese
English
aes
Eval Results
Model card Files Community
jbcs2025_Llama-3.1-8B-llama31_classification_lora-C1-essay_only-r8 / README.md
abarbosa's picture
abarbosa
Pushing fine-tuned model to Hugging Face Hub
3532be2 verified
preview code
|
raw
history blame contribute delete
1.19 kB
metadata
language:
  - pt
  - en
tags:
  - aes
datasets:
  - kamel-usp/aes_enem_dataset
base_model: meta-llama/Llama-3.1-8B
metrics:
  - accuracy
  - qwk
library_name: peft
model-index:
  - name: Llama-3.1-8B-llama31_classification_lora-C1-essay_only-r8
    results:
      - task:
          type: text-classification
          name: Automated Essay Score
        dataset:
          name: Automated Essay Score ENEM Dataset
          type: kamel-usp/aes_enem_dataset
          config: JBCS2025
          split: test
        metrics:
          - name: Macro F1
            type: f1
            value: 0.4490739809600428
          - name: QWK
            type: qwk
            value: 0.6316725978647688
          - name: Weighted Macro F1
            type: f1
            value: 0.6710525197142647

Model ID: Llama-3.1-8B-llama31_classification_lora-C1-essay_only-r8

Results

test_data
eval_accuracy 0.688406
eval_RMSE 25.0217
eval_QWK 0.631673
eval_Macro_F1 0.449074
eval_Weighted_F1 0.671053
eval_Micro_F1 0.688406
eval_HDIV 0.00724638