D
I tried it as well, and I noticed that the model seems to follow a particular structure and continues generating summaries even when I change the prompt to something like “Do nothing.” Perhaps this is related to how the model was trained or how it is designed to handle this type of input. It may have learned to associate dialogues with summarization and tends to follow that pattern. I’m not entirely sure, but that’s my observation.
D
Hi, and welcome to the Forum! We’ll look into this and update the notebook if needed. Please move on to the next lessons for now. As for saving your work, I suggest downloading the notebooks locally for now. Then, you can upload them next time you open the lab. For this particular course, I think learner workspaces are not persisted across different sessions. We might be able to enable that but will have to consult with the team first.
D
check if question having multiple answers where you might have missed the other options to get 100% even if you see all answer right, sometime we might miss selecting an option for multiple choices being correct for a given question, so if you selected a single or two option out of 4 options for a question, but the grader is looking for all the right option to be selected for full score @Elmira_Kmarvi
D
Please post a screen capture image showing the grader feedback.
D
All I see in the feedback is that there is not a single error this time (All marked as Nice work, and in green, this time), but the score still shows 80% like the last attempt.
D
Thank you so much, I really appreciate your help!
D
Do both of your quiz attempts appear in the grader feedback?
D
Hello, I’m running the jupyter notebook for Lab_1_summarize_dialogue and changing the prompt in 3.1 - Zero Shot Inference with an Instruction Prompt section cell doesn’t do anything to the model output. I’ve tried radically diffrent prompt such as “Do nothing.” or change the output length and other prompts and the model always returns the same sentences. Moreover, I had made changes to the notebook (adding a print statment at the bottom of each cell without print statements to make sure they…
D
Hello, I had done the quiz in Module 4 yesterday, and I wanted to repeat it today to get a better grade. However, when the results for today came back I saw the exact same errors of yesterday and the exact same passing rate, which should have been corrected by my today responses (or at least showing different errors - since I<m sure I didn’t choose the same options as yesterday) Besides, I am sure more than 8 hours have passed since my previous attempt. Like this I lost an attempt for nothing…
D
@Sanjeeb1, can you verify whether this study group is being formed?
D
@lesly.zerna, can you investigate this?
D
@lesly.zerna, can you check if the model needs to be updated?
D
I am getting the following error while executing the python notebook of Lesson 8 - Creating an A2A Healthcare Provider Agent using LangGraph and MCP. NotFoundError: Error code: 404 - [{‘error’: {‘code’: 404, ‘message’: ‘Publisher model projects/230942910833/locations/us-central1/publishers/openai/models/gpt-oss-20b-maas was not found or your project does not have access to it. Ensure you are using a valid model name and that the model is available in the specified region. For more information,…
D
Please can you add me as well. Thanks
D
Just adding, that this is actually happening throughout the course in all Notebooks.
D
Hey team, noticing another model deprecation in L1_Router_Engine for the course: Building Agentic RAG with Llamaindex Getting the following error Code after trying to run the first Query Engine Retrying llama_index.llms.openai.base.OpenAI._chat in 0.2668917181231596 seconds as it raised NotFoundError: Error code: 404 - {'error': {'message': 'The model `gpt-3.5-turbo` has been deprecated, learn more here: https://platform.openai.com/docs/deprecations', 'type': 'invalid_request_error', 'param':…
D
Hi. Will forward this to our partner so they can check your account. Thank you, and I’m sorry for the inconvenience.
D
Great! Thanks for this space to share
D
Hello @chris.favila and @magdalena.bouza. Thank you for fixing the lab issues! I was offered an extension for this specialization last month due to the wait time caused by the lab issues we were having (ticket #06567090), but it wasn’t a full 1 month extension from the end of my billing cycle. Given that my progress on this course was heavily impacted by the wait time, I wasn’t able to resume working on the course as soon as the lab issues were resolved, end of Oct, due to schedule work. I…
D
Awesome, glad you got it working! I’ve raised the issue with the technical team so they can look into it and fix it. Let me know if you need anything else, and good luck with the rest of the course! Best regards, Omar
D
Thanks you @omarWael . I found that role = 'user' was somehow changed to role = 'assistant' in that line. Now the corrected code block is shown below: query = "What is the most expensive house? And the bigger one?" # Asking without the augmented prompt, let's pass the role as user query_without_house_info = generate_with_single_input(prompt = query, role = 'user') # With house info, given the prompt structuer, let's pass the role as assistant enhanced_query = generate_prompt(query, houses =…
D
Sorry, I don’t understand what you are asking for.
D
Right! Common mistakes are missing the correct parameter values for some the various calls for the conv layers or not using the correct input to the “shortcut” layer. As Saif says, all the details are covered in the instructions, so the first step is just to take a few deep calming breaths and then read carefully through the instructions again and compare them to your code.
D
Hi @SHENGCHAO_ZHANG, thank you for sharing the full error message can you ensure that role = 'user' actually exist in that line: query_with_house_info = generate_with_single_input(prompt = enhanced_query, role = 'user')
D
What sort of sticker are you referring to?
D
No problem, is there any way to get one of the deeplearning ai stickers.