MadeAgents/Hammer2.1-3b
LoRA-style vs Qwen/Qwen2.5-Coder-3B-Instruct. 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 36 of 36 layers; norms, biases and embeddings are untouched.
Tensors changed: Norms 0/73, MLP 108/108, Attention 144/252
3 derivatives checked
- mradermacher/Hammer2.1-3b-GGUF: Same weights, converted
- mradermacher/Hammer2.1-3b-i1-GGUF: Quantization
- Melvin56/Hammer2.1-3b-GGUF: Same weights, converted