Upgrade the Langfuse SDK to v2 following this guide.
Apply with the Grit CLI
grit apply langfuse_node_v2
Rewrites generation parameters if there is a langfuse import
BEFORE
import { LangfuseGenerationClient } from 'langfuse'; import { messages, trace } from './messages'; const generation: LangfuseGenerationClient = trace.generation({ name: 'chat-completion', model: 'gpt-3.5-turbo', modelParameters: { temperature: 0.9, maxTokens: 2000, }, prompt: messages, completion: 'completion', });
AFTER
import { LangfuseGenerationClient } from 'langfuse'; import { messages, trace } from './messages'; const generation: LangfuseGenerationClient = trace.generation({ name: 'chat-completion', model: 'gpt-3.5-turbo', modelParameters: { temperature: 0.9, maxTokens: 2000, }, input: messages, output: 'completion', });
Does nothing if there is no langfuse import
JS
import { messages, trace } from './messages'; const generation = trace.generation({ name: 'chat-completion', model: 'gpt-3.5-turbo', modelParameters: { temperature: 0.9, maxTokens: 2000, }, prompt: messages, completion: 'completion', });