RUC-NLPIR/OmniAtlas-Qwen3-30B-A3B
Partial fine-tune vs Qwen/Qwen3-Omni-30B-A3B-Thinking. Partial change: 72% of the weights changed (attention, embeddings, lm_head, mlp, norms), in 48 of 48 layers.
Tensors changed: other 0/11, MLP 1737/1989, Norms 186/439, Attention 192/556, Embeddings 1/4, lm_head 1/1
2 derivatives checked
- mradermacher/OmniAtlas-Qwen3-30B-A3B-i1-GGUF: Quantization
- mradermacher/OmniAtlas-Qwen3-30B-A3B-GGUF: Quantization