rinna/qwen2.5-bakeneko-32b

Full fine-tune vs Qwen/Qwen2.5-32B. Fine-tune: 100% of the weights changed, across 64 of 64 layers.

Tensors changed: Embeddings 1/1, Norms 122/129, MLP 192/192, Attention 447/448, lm_head 1/1

2 derivatives checked

• mmnga/qwen2.5-bakeneko-32b-gguf: Quantization
• csabakecskemeti/rinna.qwen2.5-bakeneko-32b-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse