VishwamAI/vishwamai-stage1
LoRA-style vs VishwamAI/vishwamai-merged-gemma. 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 42 of 42 layers; norms, biases and embeddings are untouched.
Tensors changed: MLP 126/606, Norms 0/502, other 0/261, Attention 132/755, Embeddings 0/6
1 derivatives checked
- VishwamAI/vishwamai-distilled-gguf: Quantization (fewer layers)