jsbeaudry/Oswald-ASR-0.6-V1
LoRA-style vs Qwen/Qwen3-ASR-0.6B-hf. 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: other 0/11, MLP 84/156, Norms 0/187, Attention 166/256, Embeddings 0/1
1 derivatives checked
- jsbeaudry/Oswald-ASR-0.6-V1-GGUF: Quantization (fewer layers)