mradermacher/OptiMind-SFT-GGUF

Quantization vs microsoft/OptiMind-SFT. mradermacher/OptiMind-SFT-GGUF is a GGUF quantization (Q8_0) of the same weights: 193 tensors match exactly and 98 within quantization error.

Tensors changed: MLP 0/48, Attention 0/192, Norms 0/49, lm_head 0/1, Embeddings 0/1

On the Hub · Downloads on Model Pulse