mlabonne/AlphaMonarch-7B
LoRA-style vs mlabonne/NeuralMonarch-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
12 derivatives checked
- mradermacher/AlphaMonarch-7B-i1-GGUF: Quantization
- dagbs/AlphaHitchhiker-7B-GGUF: Full fine-tune
- adamksn/NeuralMaxime-7B-slerp-GGUF: Quantization
- mradermacher/AlphaMonarch-7B-GGUF: Same weights, converted
- bartowski/AlphaHitchhiker-7B-v2-GGUF: Full fine-tune
- featherless-ai-quants/mlabonne-AlphaMonarch-7B-GGUF: Quantization
- jsfs11/MixtureofMerges-MoE-2x7b-v6: MoE built from dense models
- Kukedlc/NeuralMaxime-7B-slerp: LoRA-style
- mlabonne/FrankenMonarch-7B: Partial fine-tune (fewer layers)
- macadeliccc/MonarchLake-7B: Full fine-tune
- macadeliccc/AlphaHitchhiker-7B: LoRA-style
- mlabonne/FrankenMonarch-11b: Full fine-tune (fewer layers)
1 other repos hold byte-for-byte identical weight files.