prithivMLmods/Sombrero-QwQ-32B-Elite10-Fixed

Full fine-tune vs Qwen/QwQ-32B. Fine-tune: 99% of the weights changed, across 64 of 64 layers.

Tensors changed: Embeddings 1/1, Norms 128/129, MLP 189/192, Attention 441/448, lm_head 1/1

2 derivatives checked

• mradermacher/Sombrero-QwQ-32B-Elite10-Fixed-i1-GGUF: Quantization
• mradermacher/Sombrero-QwQ-32B-Elite10-Fixed-GGUF: Quantization

On the Hub · Downloads on Model Pulse