mradermacher/Mixtral-GQA-400m-v2-GGUF

Same weights, converted vs BEE-spoke-data/Mixtral-GQA-400m-v2. Same weights stored in another dtype (BF16/F32 → F16/F32), plus 18 new tensors.

Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/13, other 0/6, Attention 0/24

On the Hub · Downloads on Model Pulse