OpenGVLab/InternVL2_5-1B
Full fine-tune (fewer layers) vs OpenGVLab/InternViT-300M-448px-V2_5. OpenGVLab/InternViT-300M-448px-V2_5 matches part of OpenGVLab/InternVL2_5-1B (340 of its 637 tensors; the rest of OpenGVLab/InternVL2_5-1B has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 24 of 24 layers.
Tensors changed: Embeddings 4/4, Attention 96/96, other 48/48, MLP 96/96, Norms 96/96
6 derivatives checked
- 5CD-AI/Vintern-1B-v3_5: Partial fine-tune
- mradermacher/InternVL2_5-1B-i1-GGUF: Quantization (fewer layers)
- ggml-org/InternVL2_5-1B-GGUF: Same weights, converted (fewer layers)
- OpenGVLab/InternVL2_5-1B-MPO: Full fine-tune
- mradermacher/InternVL2_5-1B-GGUF: Same weights, converted (fewer layers)
- D-Robotics/InternVL2_5-1B-GGUF-BPU: Same weights, converted (fewer layers)
2 other repos hold byte-for-byte identical weight files.