DQN-Labs/dqnMath-v0.1-3B-HF
LoRA-style vs LakoMoor/Ministral-3-3B-Text-Only. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of q_proj, v_proj changed, in 8 of 26 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/53, MLP 0/78, Attention 16/104
1 derivatives checked
- Flexan/DQN-Labs-dqnMath-v0.1-3B-HF-GGUF: Same weights, converted