In the LLM Bootcamp, we covered 14 different modules, including from Recurrent Neural Network, Transformers, Fine Tuning LLMs and Open Source LLM framework.
Assessment Questions:
Question-1:
- Use Transformers model and build a Code Generation application on the given data.
- Data: https://huggingface.co/datasets/theblackcat102/evol-codealpaca-v1
Question-2:
- Use GenAI Stack and build an ChatBot application like ChatGPT on your own data
- In case you don't have an OpenAI API Key, you are free to directly choose Langchain and HuggingFace model to work on a use case to build an ChatBot like ChatGPT on your own data.
- You are free to use any default data type(CSV/PDF/JSON)
- Documentation: https://genaistack.aiplanet.com/
- Since, we have already added Chat with Webpage in Module 14, we will not accept notebook with Webpage for assessment submission. Submission with Webpage as data type will only be accepted, if it has certain application use case.
Guidelines:
- Add both the code in same notebook.
- Register to this challenge with the same mail id as register in LLM Bootcamp.
- Click on Make Submission.
- Enter-
_assessment as the title. - Summary you can enter what approach you took.
- Use Tag: deep-learning
Evaluation:
The Notebook will be evaluated based:
- Step 1 : Import required modules
- Step 2 : Model Name - You can use any model or pipeline
- Step 3 : Usage of data [For question 1 given HuggingFace dataset and For question 2- your own data CSV/PDF/Web page]
- Step 4: Evaluation
- Step 5: Use 2-3 prompt and generate results
- Refer to Module 6,7, and 8 for Question 1
- Refer to module 14 for Question 2
Note: For question 2, please attach the link of the data used
Reminder: Add both Question 1 and Question 2 code in same notebook and Submit. Also for Question 2 avoid using Webpage as the data.