OpenGVLab/InternVL2_5-78B-MPO

Full fine-tune vs OpenGVLab/InternVL2_5-78B. Fine-tune: 100% of the weights changed, across 80 of 80 layers.

Tensors changed: Embeddings 4/5, Norms 237/341, Attention 694/695, other 88/90, MLP 423/426, lm_head 1/1

2 derivatives checked

• mradermacher/InternVL2_5-78B-MPO-i1-GGUF: Quantization (fewer layers)
• mradermacher/InternVL2_5-78B-MPO-GGUF: Quantization (fewer layers)

On the Hub · Downloads on Model Pulse