OpenGVLab/InternVL2_5-78B
Full fine-tune (fewer layers) vs OpenGVLab/InternViT-6B-448px-V2_5. OpenGVLab/InternViT-6B-448px-V2_5 matches part of OpenGVLab/InternVL2_5-78B (589 of its 1558 tensors; the rest of OpenGVLab/InternVL2_5-78B has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 45 of 45 layers.
Tensors changed: Embeddings 4/4, Norms 180/180, Attention 135/135, other 90/90, MLP 180/180
3 derivatives checked
- mradermacher/InternVL2_5-78B-i1-GGUF: Quantization (fewer layers)
- mradermacher/InternVL2_5-78B-GGUF: Quantization (fewer layers)
- OpenGVLab/InternVL2_5-78B-MPO: Full fine-tune
1 other repos hold byte-for-byte identical weight files.