Personal project collection to understand how LLMs work. Will draw from various open-source project collections to get a clear picture of the various concepts that are becoming increasingly popular in the field.
If there is time, I'll create slides or one-pagers that can elaborate on the concepts.
-
NLP basics
a. Text preprocessing, tokenization, POS, NER, etc.
-
NLP tasks
a. Sentiment Analysis, Text classification, etc.
-
LLM concepts
a. Transformer architecture
b. BERT, GPT
c. Understand how LLMs are pretrained.
d. Fine Tuning - example notebook Text Classification