token-js / token.js

Integrate 200+ LLMs with one TypeScript SDK using OpenAI's format.
https://docs.tokenjs.ai/
MIT License
69 stars 5 forks source link

Investigate why the max temperature test takes a very long time or is frozen #71

Open sam-goldman opened 2 months ago

sam-goldman commented 2 months ago

Problem

We have a test case that ensures that a model from every integrated LLM provider can handle the user specifying a temperature value of 2, which is the max temperature allowed. However, this test either takes a significant amount of time, or is frozen altogether. (It has never completed before)

Steps to Reproduce

  1. Setup the Token.js repo
  2. pnpm vitest run test/manual/temperature.test.ts --testTimeout 100000

Expected Result

The test case passes

Actual Result

The test case eventually times out.