OpenLLM-Korea/GME-VARCO-VISION-Embedding

Partial fine-tune vs Qwen/Qwen2-VL-7B-Instruct. Partial change: 79% of the weights changed (attention, mlp, norms), in 28 of 32 layers.

Tensors changed: Embeddings 0/2, Norms 55/187, MLP 84/216, Attention 196/324, lm_head 0/1

2 derivatives checked

• mradermacher/GME-VARCO-VISION-Embedding-i1-GGUF: Quantization (fewer layers)
• mradermacher/GME-VARCO-VISION-Embedding-GGUF: Same weights, converted (fewer layers)

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse