luzimu/FullStack-Learn-LM-30B-A3B

Full fine-tune vs Qwen/Qwen3-Coder-30B-A3B-Instruct. Fine-tune: 100% of the weights changed, across 48 of 48 layers.

Tensors changed: Embeddings 1/1, Norms 173/193, MLP 2612/2613, Attention 192/192, lm_head 1/1

2 derivatives checked

• mradermacher/FullStack-Learn-LM-30B-A3B-i1-GGUF: Quantization
• mradermacher/FullStack-Learn-LM-30B-A3B-GGUF: Quantization

On the Hub · Downloads on Model Pulse