OpenGVLab/InternVL3_5-2B
Partial fine-tune vs OpenGVLab/InternVL3_5-2B-MPO. Partial change: 87% of the weights changed (attention, embeddings, lm_head, mlp, norms), in 28 of 28 layers.
Tensors changed: lm_head 1/1, Embeddings 1/5, Norms 98/209, MLP 89/186, Attention 112/208, other 0/48
6 derivatives checked
- QiZishi/OphVLM-R1: Partial fine-tune
- bartowski/OpenGVLab_InternVL3_5-2B-GGUF: Same weights, converted (fewer layers)
- lmstudio-community/InternVL3_5-2B-GGUF: Quantization (fewer layers)
- mradermacher/InternVL3_5-2B-i1-GGUF: Quantization (fewer layers)
- mradermacher/InternVL3_5-2B-GGUF: Same weights, converted (fewer layers)
- OpenGVLab/InternVL3_5-2B-Flash: Full fine-tune
2 other repos hold byte-for-byte identical weight files.