Proximile/LLaDA-8B-Tools
LoRA-style vs GSAI-ML/LLaDA-8B-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of k_proj, q_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: Attention 96/128, Norms 0/65, Embeddings 0/1, other 0/65, MLP 0/32
2 derivatives checked
- mradermacher/LLaDA-8B-Tools-i1-GGUF: Quantization
- mradermacher/LLaDA-8B-Tools-GGUF: Same weights, converted