Bootstrap knowledge of LLMs ASAP. With a bias/focus to GPT.
Avoid being a link dump. Try to provide only valuable well tuned information.
Neural network links before starting with transformers.
Maybe you've heard about this technique but you haven't completely understood it, especially the PPO part. This explanation might help.
We will focus on text-to-text language models 📝, such as GPT-3, BLOOM, and T5. Models like BERT, which are encoder-only, are not addressed.
Reinforcement Learning from Human Feedback (RLHF) has been successfully applied in ChatGPT, hence its major increase in popularity. 📈
RLHF is especially useful in two scenarios 🌟:
People
:bowtie: |
😄 :smile: |
😆 :laughing: |
---|---|---|
😊 :blush: |
😃 :smiley: |
:relaxed: |
😏 :smirk: |
😍 :heart_eyes: |
😘 :kissing_heart: |
😚 :kissing_closed_eyes: |
😳 :flushed: |
😌 :relieved: |
😆 :satisfied: |
😁 :grin: |
😉 :wink: |
😜 :stuck_out_tongue_winking_eye: |
😝 :stuck_out_tongue_closed_eyes: |
😀 :grinning: |
😗 :kissing: |
😙 :kissing_smiling_eyes: |
😛 :stuck_out_tongue: |
Reference: https://gist.github.com/justlaputa/a6da84981eca963817e652b5f2452cfc
Generate authorization token
for unstar.
https://github.com/settings/tokens
Export variable
export APIKEY=xxxxxxxxx
pip install streamlit
pip install streamlit-chat