Crystalcareai/CrystalMistral-14b
Same weights, repacked (fewer layers) vs eren23/dpo-binarized-NeuralTrix-7B. eren23/dpo-binarized-NeuralTrix-7B matches part of Crystalcareai/CrystalMistral-14b (291 of its 579 tensors; the rest of Crystalcareai/CrystalMistral-14b has no counterpart). On the shared part: Same weights in different files (re-saved, re-sharded or new metadata), plus 288 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 0/96, Attention 0/128, lm_head 0/1
2 derivatives checked
- mradermacher/CrystalMistral-14b-i1-GGUF: Quantization
- mradermacher/CrystalMistral-14b-GGUF: Quantization