Jupyter Notebook: Unsloth fine tuning for Ollama (Qwen2.5 Instruct 32B, fine tuned with an Alpaca-formated json data file)
Jupyter notebook for fine tuning the Qwen2.5 Instruct 32B model for Ollama with Unsloth locally in Windows WSL. Runs successfully with an Alpaca-formatted json data file (my blog posts w/ generated prompts) on a Nvidia Geforce RTX 4090.
Ollama_+_Unsloth_+_Llama_3_+_Alpaca.ipynb
Jupyter notebook to fine tune a model locally on an Alpaca-formatted json data file. Edited from notebook provided by Unsloth to work locally.
data.json
Alpaca-formatted json file used to fine tune.
lora_model folder
Contains generated LORA model files.
model folder
Contains generated fine tuned model (4-bit quantized gguf) and Modelfile (needed to create model in Ollama).
- Downloads last month
- -
Hardware compatibility
Log In to add your hardware
4-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support