amadeusai/Amadeus-Verbo-BI-Qwen-2.5-72B-PT-BR-Instruct-Experimental

Full fine-tune vs Qwen/Qwen2.5-72B. Fine-tune: 98% of the weights changed, across 80 of 80 layers.

Tensors changed: Embeddings 0/1, MLP 240/240, Attention 559/560, Norms 113/161, lm_head 1/1

2 derivatives checked

• mradermacher/AV-BI-Qwen2.5-72B-PT-BR-Instruct-i1-GGUF: Quantization
• mradermacher/AV-BI-Qwen2.5-72B-PT-BR-Instruct-GGUF: Quantization

On the Hub · Downloads on Model Pulse