mradermacher/MSH-Lite-7B-v1-Bielik-v2.3-Instruct-Llama-Prune-GGUF

Same weights, converted vs meditsolutions/MSH-Lite-7B-v1-Bielik-v2.3-Instruct-Llama-Prune. Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/101, MLP 0/150, Attention 0/200, lm_head 0/1

On the Hub · Downloads on Model Pulse