FreedomIntelligence/HuatuoGPT-o1-7B
Full fine-tune vs Qwen/Qwen2.5-7B-Instruct. Fine-tune: 93% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 0/1, Norms 53/57, MLP 84/84, Attention 196/196, lm_head 1/1
8 derivatives checked
- bartowski/HuatuoGPT-o1-7B-GGUF: Full fine-tune
- bartowski/HuatuoGPT-o1-7B-v0.1-GGUF: Same weights, converted
- mradermacher/HuatuoGPT-o1-7B-i1-GGUF: Quantization
- mradermacher/HuatuoGPT-o1-7B-GGUF: Same weights, converted
- tensorblock/HuatuoGPT-o1-7B-GGUF: Full fine-tune
- DevQuasar-6/FreedomIntelligence.HuatuoGPT-o1-7B-GGUF: Full fine-tune
- Xiaojian9992024/Qwen2.5-THREADRIPPER-Small: Full fine-tune
- TheBlueObserver/HuatuoGPT-o1-7B-MLX: Same weights, converted
2 other repos hold byte-for-byte identical weight files.