ManniX-ITA/Qwen3.5-4B-MicroCoder

Partial fine-tune vs Qwen/Qwen3.5-4B. Partial change: 63% of the weights changed (attention, mlp, norms), in 32 of 32 layers.

Tensors changed: Embeddings 0/5, MLP 96/199, Attention 94/324, Norms 27/209, other 0/1

1 derivatives checked

• mradermacher/Qwen3.5-4B-MicroCoder-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse