34 Photos and videos
𧡠Called Google Gemini API via Python LangChain for the first time.
Here's how (thread π)
#LangChain #GoogleGemini #Python #GenAI #LearningInPublic #Agents
1
38
Step 2: Call the model
As shown in follwowing image you can call model:
#LangChain #GoogleGemini #Python #GenAI #LearningInPublic
1
29
πOutput:
πThis same pattern works for Groq, OpenAI, Anthropic and Open source models like HugginFace.
β¨just swap the importn and initialization of the model
#LangChain #GoogleGemini #Python #GenAI #LearningInPublic
34
π΄ ASSISTANT role β the reply
β’ The AI's generated response
β’ Shaped by BOTH system user
β’ Previous assistant messages are fed back into the next request
This is the secret behind "memory":
The model reads ALL of this every single time.
#LLM #GenerativeAI #AgenticAI
1
30
The LLM has zero persistent memory.
Your app manually builds and passes this array on every call.
That growing array = the "memory" of your chatbot.
For Agentic AI, managing this array across 10 steps is a core engineering challenge.
#LLM #GenerativeAI #AgenticAI #python
33
That's it. 3 concepts every AI builder must know:
β
LLM β the brain
β
Tokens β what it reads
β
Context Window β how much it remembers
Follow for more beginner-friendly AI breakdowns π
#AI #LLM #AIBasics #GenerativeAI #MachineLearning
30
Most people use AI every day but don't know AI fundamentals! π
π§ What is an LLM?
π’ Tokens vs. Words
π Why Context Windows matter
Perfect for beginners or devs starting their AI journey.
My Medium article: tinyurl.com/5yb4f5ma
#AI #GenAI #BuildInPublic #AgenticAI
1
31