Introducing Multimodal Llama 3.2 -- broken Jupyter Notebook?

Short Course Q&A Introducing Multimodal Llama 3.2
1

Continuing the discussion from Error in lesson 3 multi model prompting:

The Introducing Multimodal Llama 3.2 course needs to be reviewed soon. In the Llama 3.2 Multimodal Prompting lesson, I got the same error–this Jupyter notebook, Lesson_2.ipynb, is broken in this statement:

response_32 = llama32(messages, 90)

giving:

---------------------------------------------------------------------------
Exception                                 Traceback (most recent call last)
Cell In[13], line 1
----> 1 response_32 = llama32(messages, 90)
      2 print(response_32)

File ~/work/L2/utils.py:35, in llama32(messages, model_size)
     32 res = json.loads(requests.request("POST", url, headers=headers, data=json.dumps(payload)).content)
     34 if 'error' in res:
---> 35   raise Exception(res['error'])
     37 return res['choices'][0]['message']['content']

Exception: {'message': 'Unable to access non-serverless model meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo. Please visit https://api.together.ai/models/meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo to create and start a new dedicated endpoint for the model.', 'type': 'invalid_request_error', 'param': None, 'code': 'model_not_available'}

‘model_not_available’ seems a primary reason; it seems that this notebook needs to be updated.

2

Hi Jim_Byers,

Thanks for reporting this!

@Community-Team Can someone look into this?

1 Like
3

Thank you for reporting. I’m checking it!

-- Lesly, DLAI