eulogik/Bharat-Tiny-LLM-v3
LoRA-style vs Qwen/Qwen3-1.7B-Base. 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 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/113, MLP 84/84, Attention 112/112
1 derivatives checked
- eulogik/Bharat-Tiny-LLM-v3.4: LoRA-style