ClaudioItaly/intelligence-cod-rag-7b-v3-2

Partial fine-tune vs ATH-MaaS/Marco-o1. Partial change: 39% of the weights changed (attention, lm_head, mlp, norms), in 26 of 28 layers.

Tensors changed: Embeddings 0/1, Norms 6/57, MLP 31/84, Attention 49/196, lm_head 1/1

1 derivatives checked

• mradermacher/intelligence-cod-rag-7b-v3-2-i1-GGUF: Quantization

On the Hub · Downloads on Model Pulse