orbcom-pedroferreira/AMALIA-VL-DPO-GGUF

Same weights, converted (fewer layers) vs amalia-llm/AMALIA-VL-DPO. orbcom-pedroferreira/AMALIA-VL-DPO-GGUF matches part of amalia-llm/AMALIA-VL-DPO (381 of its 823 tensors; the rest of amalia-llm/AMALIA-VL-DPO has no counterpart). On the shared part: Same weights stored in another dtype (BF16 → F16/F32).

Tensors changed: Embeddings 0/1, Norms 0/85, MLP 0/126, Attention 0/168, lm_head 0/1

On the Hub · Downloads on Model Pulse