prithivMLmods/docscopeOCR-7B-050425-exp
Full fine-tune vs Qwen/Qwen2.5-VL-7B-Instruct. Fine-tune: 93% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/2, Norms 61/122, MLP 280/280, Attention 324/324, lm_head 1/1
4 derivatives checked
- mradermacher/docscopeOCR-7B-050425-exp-i1-GGUF: Quantization (fewer layers)
- mradermacher/docscopeOCR-7B-050425-exp-GGUF: Same weights, converted (fewer layers)
- prithivMLmods/docscopeOCR-7B-050425-exp-GGUF: Quantization (fewer layers)
- prithivMLmods/DREX-062225-exp: LoRA-style