OpenGVLab/InternViT-6B-448px-V1-5
Full fine-tune vs OpenGVLab/InternViT-6B-448px-V1-2. 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
- OpenGVLab/InternVL-Chat-V1-5: Full fine-tune (fewer layers)
- OpenGVLab/InternVL2-26B: Full fine-tune (fewer layers)
- OpenGVLab/InternViT-6B-448px-V2_5: Full fine-tune
1 other repos hold byte-for-byte identical weight files.