alamios/DeepSeek-R1-DRAFT-Qwen2.5-0.5B
Full fine-tune vs Qwen/Qwen2.5-0.5B. Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: Norms 11/49, MLP 72/72, Attention 166/168
2 derivatives checked
- tensorblock/DeepSeek-R1-DRAFT-Qwen2.5-0.5B-GGUF: Quantization
- alamios/DeepSeek-R1-DRAFT-Qwen2.5-0.5B-GGUF: Same weights, converted