akshatkot/linus-coder

Partial fine-tune vs unsloth/Qwen2.5-Coder-7B-Instruct-bnb-4bit. Partial change: 6% of the weights changed (attention, mlp), in 28 of 28 layers.

Tensors changed: MLP 199/445, Attention 247/664, lm_head 0/1, Embeddings 0/1, Norms 0/57

On the Hub · Downloads on Model Pulse