Fine-Tuning of Small Language Model (10$) for study purpose.
Budget: $10 – $30 USD
Low Budget - Study purpose
Total budget 10$.
-------------------------------------------
Select One LLM model - can be small LLM (CPU based - like TinyLlama or DistilBert etc) :
- able to run locally
- fine tunable on local or Colab.
Diverse types of Question Answer generation from a generic passage eg Python textbook.
(True/False, Multiple Choice Questions, Match the following, Fill in the blanks etc)
Prompt Engineering - 0 shot, 1 shot, few shot
Prepare document.
-----------------------------------------
Prepare dataset for Fine Tuning.
Fine tuning (graphs of loss - training, evaluation, test) & other details
(Also provide links to official pages where this details are mentioned)
Evaluation of Question/Answers generated before and after the Fine tuning. There should be noticable difference between the two (before and after FT).
Prepare document.
The final result should be good enough, consistent & usable, saved to db (any db).
Total budget 10$.
-------------------------------------------
Select One LLM model - can be small LLM (CPU based - like TinyLlama or DistilBert etc) :
- able to run locally
- fine tunable on local or Colab.
Diverse types of Question Answer generation from a generic passage eg Python textbook.
(True/False, Multiple Choice Questions, Match the following, Fill in the blanks etc)
Prompt Engineering - 0 shot, 1 shot, few shot
Prepare document.
-----------------------------------------
Prepare dataset for Fine Tuning.
Fine tuning (graphs of loss - training, evaluation, test) & other details
(Also provide links to official pages where this details are mentioned)
Evaluation of Question/Answers generated before and after the Fine tuning. There should be noticable difference between the two (before and after FT).
Prepare document.
The final result should be good enough, consistent & usable, saved to db (any db).