mradermacher

/

model_requests

like 140

Let's hope I can =)

You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#Ovis2.6-30B-A3B-GGUF for quants to appear.

Let's hope I can =)

You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#Ovis2.6-30B-A3B-GGUF for quants to appear.

I noticed in the status that '+vision' means for models that can process images, right? But the Ovis2.6-30B-A3B has vision capability, yet I don't see '+vision' next to it in the status.

I noticed in the status that '+vision' means for models that can process images, right? But the Ovis2.6-30B-A3B has vision capability, yet I don't see '+vision' next to it in the status.

In that case vision is not yet supported in the llama.cpp. Often llama.cpp developer omit vision when implementing a new architecture and then add it somewhere later. You could request the mmproj containing the vision stack once llama.cpp supports vision for the Ovis2_6_MoeForCausalLM architecture.

Are we even sure Ovis2_6_MoeForCausalLM is supported by llama.cpp at all? I only see Ovis 2.5 support to ever have been added so I assume the model will likely fail but we will see.

Are we even sure Ovis2_6_MoeForCausalLM is supported by llama.cpp at all?

Let's hope I can =)

goodmorning nico =)