unsloth/Pixtral-12B-2409
Same weights, repacked vs mistralai/Pixtral-12B-2409. Same weights in different files (re-saved, re-sharded or new metadata), plus 222 new tensors.
Tensors changed: Attention 0/160, Norms 0/81, MLP 0/120, lm_head 0/1, Embeddings 0/1
1 derivatives checked
- TikFesku/Pixtral-12B-2409-Q8_0-GGUF: Quantization (fewer layers)
13 other repos hold byte-for-byte identical weight files.