inglebhavin98/hinglish-smooth-talker-qwen2.5-3b-experimental

Partial fine-tune vs mlx-community/Qwen2.5-3B-Instruct-4bit. Partial change: 70% of the weights changed (attention, mlp), in 28 of 36 layers.

Tensors changed: Embeddings 0/3, Norms 0/73, MLP 252/324, Attention 336/540

On the Hub · Downloads on Model Pulse