amadeusai/Amadeus-Verbo-FI-Qwen2.5-7B-PT-BR-Instruct
Full fine-tune vs Qwen/Qwen2.5-7B-Instruct. Fine-tune: 93% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 0/1, Norms 50/57, MLP 84/84, Attention 195/196, lm_head 1/1
3 derivatives checked
- mradermacher/Amadeus-Verbo-FI-Qwen2.5-7B-PT-BR-Instruct-i1-GGUF: Quantization
- mradermacher/AV-FI-Qwen2.5-7B-PT-BR-Instruct-GGUF: Same weights, converted
- mradermacher/Amadeus-Verbo-FI-Qwen2.5-7B-PT-BR-Instruct-GGUF: Same weights, converted