#cognitivecomputations/dolphin-2.9.1-qwen-110b
11 messages · Page 1 of 1 (latest)
It's really looking promising and I think would be a great replacement of the older dolphin mixtral
Qwen is very censored. But dolphin variants should be uncensored?
Looks cool 👍 don't tell me it has a 4 k context window again 😟 oh no it's 32 k 🤣
if this is good, I would love this
I am curious about the speed though
that is my only concern
a smaller dolphin model might be more ideal idk
not exactly a dolphin model, but this one is a great uncensored smaller model: https://huggingface.co/cognitivecomputations/Llama-3-8B-Instruct-abliterated-v2
Together.ai added this model to the request Page, you can vote for it here: https://together-ai.canny.io/model-requests/p/cognitivecomputations-dolphin-291-qwen-110b
would love to see it supported by OR