Code used to fine-tune this model: abacaj/mistral-7b-sft. Add your data in the data folder as train.jsonl
and validation.jsonl
.
Install dependencies:
python -m venv env \
&& source env/bin/activate \
&& pip install -r requirements.txt
Run training code:
torchrun --nnodes=1 --nproc-per-node=<REPLACE_WITH_NUMBER_OF_GPUS> train.py