Spaces:

huggingchat

/

chat-ui

like 1.1k

Running

what are limits of using these? how many api calls can i send them per month?

How can I know which model am using

How can I know which model am using

at the bottom of your screen:

Out of all these models, Gemma, which was recently released, has the newest information about .NET. However, I don't know which one has the most accurate answers regarding coding

Gemma seems really biased. With web search on, it says that it doesn't have access to recent information asking it almost anything about recent events. But when I ask it about recent events with Google, I get responses with the recent events.

apparently gemma cannot code?

Gemma is just like Google's Gemini series models, it have a very strong moral limit put on, any operation that may related to file operation, access that might be deep, would be censored and refused to reply.
So even there are solution for such things in its training data, it will just be filtered and ignored.
But still didn't test the coding accuracy that doesn't related to these kind of "dangerous" operations

@nsarrazin why didn't you guys add Deepseek-R1-Distill-Llama-70B? It outperforms the one you added in most benchmarks.

1. can we have a control option for reasoning? to turn off/on or to tweak it. sometimes it takes ~a minute just to overthink (or multiple rounds) simple things.
   https://hf.co/chat/r/5YOOIkK?leafId=b105d900-8461-4244-83f4-ddbf26acdb59
2. LeafID works well even if we refresh/restart the tab but reset back to branch 1 when traversing between different chats. E.g. Re-selecting a chat after selecting another chat. (Sounds like a feature but it'd be the best to be stored in the chats instead)

Have you guys had any issues with the chat interface lagging? it freezes for me when i ask for an error info

Hi @nsarrazin is it possible that you guys can remove nemo and phi3.5 mini and replace both of them with Phi4-14b??

Have you guys had any issues with the chat interface lagging? it freezes for me when i ask for an error info

noo and i actually deployed locally on raspberry pi with models running on inference endpoint.

@nsarrazin why didn't you guys add Deepseek-R1-Distill-Llama-70B? It outperforms the one you added in most benchmarks.

Please add this model to Huggingchat, it's nice and kinda revolutionary, many people think it's good

Also, can you please let us set repetition penalty and temperature to 0? At least in API playground it is possible to have temperature set to 0
And another thing to add, or rather bring back, is to able to see token usage in system prompt text field and in assistant editor

I noticed that some answers from the model "deepseek-ai/DeepSeek-R1-Distill-Qwen-32B" are similar to the ones given by the model "CohereForAI/c4ai-command-r-plus-08-2024".