mradermacher/Bielik-PL-Minitron-7B-v3.0-Instruct-ungated-GGUF

Same weights, converted vs cpral/Bielik-PL-Minitron-7B-v3.0-Instruct-ungated. Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: Embeddings 0/1, MLP 0/120, Norms 0/81, Attention 0/160, lm_head 0/1

On the Hub · Downloads on Model Pulse