Incorrect Model ID for DeepSeek Distill R1 Llama 70B in Supported Models Documentation

SambaNova Devs Discussion
1

https://community.sambanova.ai/t/supported-models/193

The Model ID for DeepSeek Distill R1 Llama 70B in the Supported Models documentation is incorrect as DeepSeek-Distill-R1-Llama-70B, making it unusable.

In the Playground on SambaNova Cloud, the code uses DeepSeek-R1-Distill-Llama-70B, which works.

2 Likes
2

Thank you I will correct it in a bit.

Coby

1 Like
3

I don’t see DeepSeek Distill R1 Llama 70B in the playground? Where is it?

Also got this error when executed it using API @Coby

AI response error; aborting request: 400 Unknown model: DeepSeek-Distill-R1-Llama-70B
1 Like
4

@febryanvaldo089
I can’t see it now either.
You can try using DeepSeek-R1-Distill-Llama-70B.
I can currently use this to call the API.

1 Like
5

I think this is a mistake in the documentation, as the model name on Hugging Face is DeepSeek-R1-Distill-Llama-70B.

The name means that the DeepSeek-R1 model was used to generate data, which was then used to distilled into the Llama 3.3 - 70B model.

2 Likes
6

Hello @pcchen

Yeah i grabbed the Model IDs using python script, and it’s DeepSeek-Distill-R1-Llama-70B, looking at the log surely it was the wrong name. It still not fixed till now. I will try your suggestion though.

Edit:
Indeed, the correct one is DeepSeek-R1-Distill-Llama-70B

612×181 5.26 KB

2 Likes
7

@febryanvaldo089 @pcchen and others Thank you for catching this. I have corrected it .
-Coby

1 Like
8

Hello,

Thanks, it seemed fixed now.

1 Like