mirror of
				https://github.com/hwchase17/langchain.git
				synced 2025-10-31 16:08:59 +00:00 
			
		
		
		
	
		
			
				
	
	
		
			26 lines
		
	
	
		
			887 B
		
	
	
	
		
			Plaintext
		
	
	
	
	
	
			
		
		
	
	
			26 lines
		
	
	
		
			887 B
		
	
	
	
		
			Plaintext
		
	
	
	
	
	
| ```python
 | |
| from langchain.chains import ConversationChain
 | |
| from langchain.memory import ConversationBufferMemory
 | |
| 
 | |
| conversation = ConversationChain(
 | |
|     llm=chat,
 | |
|     memory=ConversationBufferMemory()
 | |
| )
 | |
| 
 | |
| conversation.run("Answer briefly. What are the first 3 colors of a rainbow?")
 | |
| # -> The first three colors of a rainbow are red, orange, and yellow.
 | |
| conversation.run("And the next 4?")
 | |
| # -> The next four colors of a rainbow are green, blue, indigo, and violet.
 | |
| ```
 | |
| 
 | |
| <CodeOutputBlock lang="python">
 | |
| 
 | |
| ```
 | |
|     'The next four colors of a rainbow are green, blue, indigo, and violet.'
 | |
| ```
 | |
| 
 | |
| </CodeOutputBlock>
 | |
| 
 | |
| Essentially, `BaseMemory` defines an interface of how `langchain` stores memory. It allows reading of stored data through `load_memory_variables` method and storing new data through `save_context` method. You can learn more about it in the [Memory](/docs/modules/memory/) section.
 | |
| 
 |