thu-ml/STAIR-Qwen2-7B-SFT

Full fine-tune vs Qwen/Qwen2-7B-Instruct. Fine-tune: 93% of the weights changed, across 28 of 28 layers.

Tensors changed: Embeddings 0/1, Norms 54/57, MLP 84/84, Attention 196/196, lm_head 1/1

2 derivatives checked

• mradermacher/STAIR-Qwen2-7B-SFT-i1-GGUF: Quantization
• mradermacher/STAIR-Qwen2-7B-SFT-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse