mlabonne/NeuralHermes-2.5-Mistral-7B
LoRA-style vs teknium/OpenHermes-2.5-Mistral-7B. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/65, MLP 96/96, Attention 128/128, lm_head 0/1
5 derivatives checked
- TheBloke/NeuralHermes-2.5-Mistral-7B-GGUF: Quantization
- MaziyarPanahi/NeuralHermes-2.5-Mistral-7B-GGUF: Quantization
- tensorblock/NeuralHermes-2.5-Mistral-7B-GGUF: Quantization
- FredrikBL/HermesFlashback-7B.1: Full fine-tune
- FredrikBL/NeuralFlashback-ties: Full fine-tune
1 other repos hold byte-for-byte identical weight files.