_assessment

For the text to code application, I use the orca mini 3b model and fine tuned it using the SFT trainer. I used the dataset for first 1000 rows, as the allowance limit in Collab prohibited above it. The losses are at acceptable numbers and the output given out is correct. For the ChatBot application, I used the GenStack AI and its associated libraries to read a finance related document and asked questions to the bot accordingly. The ETL reads the document and passes it through the OpenAI engine to generate output.