amos1088's picture
Upload README.md with huggingface_hub
310f636 verified
|
Raw
History Blame Contribute Delete
1.36 kB
metadata
tags:
  - document-relevance
  - dpo
  - gpt-oss-20b
datasets:
  - custom-relevance-dataset
metrics:
  - accuracy
model-index:
  - name: gpt-oss-20b-relevance-ft-20250811_213108
    results:
      - task:
          type: text-classification
          name: Document Relevance Classification
        metrics:
          - type: accuracy
            value: 0.575
            name: Validation Accuracy
          - type: yes_ratio
            value: 0.475
            name: Yes Prediction Ratio
          - type: no_ratio
            value: 0.525
            name: No Prediction Ratio

gpt-oss-20b Document Relevance Classifier

This model was trained using standard fine-tuning for document relevance classification.

Training Configuration

  • Base Model: openai/gpt-oss-20b
  • Training Type: Standard Fine-tuning
  • Learning Rate: 5e-06
  • Batch Size: 32
  • Epochs: 5
  • Training Samples: 2000
  • Validation Samples: 400

Performance Metrics

  • Accuracy: 57.50%
  • Yes Predictions: 47.5%
  • No Predictions: 52.5%

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

# Load base model
model = AutoModelForCausalLM.from_pretrained("openai/gpt-oss-20b")
tokenizer = AutoTokenizer.from_pretrained("openai/gpt-oss-20b")

# Load adapter
model = PeftModel.from_pretrained(model, "amos1088/gpt-oss-20b-relevance-ft-20250811_213108")

Training Date

2025-08-11 21:31:08 UTC