smolLM-sft-gec / README.md
coreprinciple's picture
Create README.md
6346563 verified
metadata
datasets:
  - grammarly/coedit
metrics:
  - bleu
base_model:
  - HuggingFaceTB/SmolLM-135M

Task: he task is to Fine-tune the SmolLM-135M model using the CoEdIT dataset which includes input sentences with grammatical errors and their corrected versions.

This model is a finetuned version with bleu score of 0.47