I want to create a model which will help me generate test cases after looking at a piece of requirement.
The requirement can be extracted from any source and by seeing the requirements one by one my model should be able to generate test cases.
Can anyone let me know how can I achieve the same ? And what all would I need for the same. I don’t want to use any LLM. I want to build this from scratch.
not much…ML specialisation (all courses), Deep learning specialisation(couple of them), some gen AI courses, some independent LLM research and built few apps with it…
The courses here which teach you the mechanisms you need to build your own “mini” LLM from scratch would be DLS Course 5 (Sequence Models) and the last two courses of the NLP Specialization. NLP Course 3 covers Sequence Models in general and NLP Course 4 covers Attention Models. DLS C5 covers both Sequence Models and Attention Models (which are a subtype of Sequence Models). There may also be some of the TF Specializations that cover this, but the correct order would be DLS C5 and then NLP and then see how far that gets you towards the solution you want to build.