NousResearch/Genstruct-7B
Full fine-tune vs mistralai/Mistral-7B-v0.1. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Norms 13/65, MLP 96/96, Attention 128/128
5 derivatives checked
- mradermacher/Genstruct-7B-GGUF: Same weights, converted
- mradermacher/Genstruct-7B-i1-GGUF: Quantization
- tensorblock/Genstruct-7B-GGUF: Quantization
- ilevytate/Fusion-7B-Quintessence: Full fine-tune (fewer layers)
- artificialguybr/GenStructDolphin-7B-Slerp: Full fine-tune
2 other repos hold byte-for-byte identical weight files.