[SOLVED] M5_UGL_1_R: Authentication error received

Course Q&A Agentic AI
,
1

When running the notebook M5_UGL_1_R on the DeepLearning.AI platform, I always receive the following authentication error when running the cell with the command “full_content_round = generate_llm_code(…)”.

AuthenticationError: Error code: 401 - {‘error’: {‘message’: ‘Your authentication token is not from a valid issuer.’, ‘type’: ‘invalid_request_error’, ‘code’: ‘invalid_issuer’, ‘param’: None}}

I have tried to log out/in, ran the notebook again, but always the same error.

2

are you getting error on running codes in the deep learning.ai platform?

also did you try clearing cache and browsing history and then try running the codes?

3

Im seeing quite a few posts with this error, my take is that there is something wrong with the LLM access key that DLAI uses for these Labs. Perhaps if someone knows the QA for these courses can let them know.

4

@gent.spah

Based on this link, it’s @lesly.zerna

2 Likes
5

@gent.spah

during testing of this course, some of the learner faced this issue but it was addressed then, with update by developer team.

But even I am surprise that this error is being thrown. I feel it might be llm version use too, as instruction in labs particularly mentions to use only only two llms with their respective version mentioned but what if there is an overall change in the version or llm update could be causing.

Balaji has tagged lesly, she will have a look again.

2 Likes
6

Follow-up:
I just ran the notebook again (without changing any code in the notebook or any notebook settings), and this time, the entire notebook ran without any issues.

Thank you for all your feedback.

1 Like
7

Thank you for the updated! I’ve tested again today and this code cell:

Andrew’s prompt from the lecture

prompt_round = “Do you have any round sunglasses in stock that are under $100?”

Generate the plan-as-code (FULL content; may include <execute_python> tags)

full_content_round = generate_llm_code(
prompt_round,
inventory_tbl=inventory_tbl,
transactions_tbl=transactions_tbl,
model=“o4-mini”,
temperature=1.0,
)

Inspect the LLM’s plan + code (no execution here)

utils.print_html(full_content_round, title=“Plan with Code (Full Response)”)

Indeed takes a bit longer to execute and it is properly working :rocket: