MegaScience/Qwen2.5-1.5B-MegaScience
Full fine-tune vs Qwen/Qwen2.5-1.5B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 57/57, MLP 84/84, Attention 196/196
3 derivatives checked
- prithivMLmods/MegaScience-Qwen-GGUF: Partial fine-tune
- mradermacher/Qwen2.5-1.5B-MegaScience-i1-GGUF: Quantization
- mradermacher/Qwen2.5-1.5B-MegaScience-GGUF: Same weights, converted