Skywork/Skywork-OR1-Math-7B
Full fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Qwen-7B. Fine-tune: 93% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 0/1, Attention 196/196, MLP 84/84, Norms 48/57, lm_head 1/1
4 derivatives checked
- bartowski/Skywork_Skywork-OR1-Math-7B-GGUF: Same weights, converted
- tinyopsec/Skywork-OR1-Math-7B-GGUF: Same weights, converted
- lmstudio-community/Skywork-OR1-Math-7B-GGUF: Quantization
- Salesforce/E1-Math-7B: Partial fine-tune