diff --git a/AI/Generative AI & LLM's/W3-ReinforcementLearning=LLM-poweredApplications.md b/AI/Generative AI & LLM's/W3-ReinforcementLearning=LLM-poweredApplications.md index 1aa7178..8931366 100644 --- a/AI/Generative AI & LLM's/W3-ReinforcementLearning=LLM-poweredApplications.md +++ b/AI/Generative AI & LLM's/W3-ReinforcementLearning=LLM-poweredApplications.md @@ -44,4 +44,31 @@ The clarity of your instructions can make a big difference on the quality of the [Constitutional AI: Harmlessness from AI Feedback paper](images/ConstitutionalAI.pdf) +## LLM Powered applications + +### Introduction Model optimizations for deployment + +Increase performance -> reduce LLM size, which reduces inference latency +The challenge is to reduce the size of the model while still maintaining +model performance. + +![Model Optimizations Techniques](images/optimizationsTechniques1.png)] + +[Video has a lot of information](images/ModelOptimizationsfordeployment.mp4) + +![Generative AI Project Lifecycle Cheat Sheet](images/GenerativeAIProjectLifecycleCheatSheet.png) + +~[LLM-Powered Appplication](images/PowerApplications1.png) + +Langchain is an example of Orchestration Library + +Retrieval Augmented Generation (**RAG**) is a great way to overcome the knowledge cutoff (because the world has changes since the model was trained with data current to that date) issue and help the model update its understanding of the world. + +![RAG](images/RAG1.png) + +[Facebook RAG paper](images/RAG_Paper.pdf) + +The external data store could be a vector store,a SQL database, CSV files, Wikis or other data storage format. + +![RAG](images/RAG2.png) diff --git a/AI/Generative AI & LLM's/images/GenerativeAIProjectLifecycleCheatSheet.png b/AI/Generative AI & LLM's/images/GenerativeAIProjectLifecycleCheatSheet.png new file mode 100644 index 0000000..3487d2c Binary files /dev/null and b/AI/Generative AI & LLM's/images/GenerativeAIProjectLifecycleCheatSheet.png differ diff --git a/AI/Generative AI & LLM's/images/ModelOptimizationsfordeployment.mp4 b/AI/Generative AI & LLM's/images/ModelOptimizationsfordeployment.mp4 new file mode 100644 index 0000000..35da1f6 Binary files /dev/null and b/AI/Generative AI & LLM's/images/ModelOptimizationsfordeployment.mp4 differ diff --git a/AI/Generative AI & LLM's/images/PowerApplications1.png b/AI/Generative AI & LLM's/images/PowerApplications1.png new file mode 100644 index 0000000..7920722 Binary files /dev/null and b/AI/Generative AI & LLM's/images/PowerApplications1.png differ diff --git a/AI/Generative AI & LLM's/images/RAG1.png b/AI/Generative AI & LLM's/images/RAG1.png new file mode 100644 index 0000000..2dab1dd Binary files /dev/null and b/AI/Generative AI & LLM's/images/RAG1.png differ diff --git a/AI/Generative AI & LLM's/images/RAG2.png b/AI/Generative AI & LLM's/images/RAG2.png new file mode 100644 index 0000000..1c72a40 Binary files /dev/null and b/AI/Generative AI & LLM's/images/RAG2.png differ diff --git a/AI/Generative AI & LLM's/images/RAG_Paper.pdf b/AI/Generative AI & LLM's/images/RAG_Paper.pdf new file mode 100644 index 0000000..3921837 Binary files /dev/null and b/AI/Generative AI & LLM's/images/RAG_Paper.pdf differ diff --git a/AI/Generative AI & LLM's/images/optimizationsTechniques1.png b/AI/Generative AI & LLM's/images/optimizationsTechniques1.png new file mode 100644 index 0000000..4339bd3 Binary files /dev/null and b/AI/Generative AI & LLM's/images/optimizationsTechniques1.png differ