randomsubmit/GraphForge-Qwen3.6-35B-A3B-SFT

Full fine-tune vs Qwen/Qwen3.6-35B-A3B. Fine-tune: 97% of the weights changed, across 40 of 40 layers.

Tensors changed: Embeddings 0/4, Attention 245/388, MLP 280/392, Norms 100/241, lm_head 1/1

2 derivatives checked

• mradermacher/GraphForge-Qwen3.6-35B-A3B-SFT-i1-GGUF: Quantization
• mradermacher/GraphForge-Qwen3.6-35B-A3B-SFT-GGUF: Quantization

On the Hub · Downloads on Model Pulse