OpenGVLab/InternVL3-14B-Instruct
Full fine-tune vs OpenGVLab/InternVL3-14B-Pretrained. Fine-tune: 100% of the weights changed, across 48 of 48 layers.
Tensors changed: Embeddings 5/5, Norms 187/193, MLP 246/246, Attention 432/432, other 48/48, lm_head 1/1
6 derivatives checked
- OpenGVLab/InternVL3-14B: Full fine-tune
- OpenGVLab/InternVL3-14B-hf: Full fine-tune
- unsloth/InternVL3-14B-Instruct-GGUF: Same weights, converted (fewer layers)
- mradermacher/InternVL3-14B-Instruct-i1-GGUF: Quantization (fewer layers)
- ggml-org/InternVL3-14B-Instruct-GGUF: Same weights, converted (fewer layers)
- mradermacher/InternVL3-14B-Instruct-GGUF: Quantization (fewer layers)
2 other repos hold byte-for-byte identical weight files.