continuedev / continue

⏩ Continue is the leading open-source AI code assistant. You can connect any models and any context to build custom autocomplete and chat experiences inside VS Code and JetBrains
https://docs.continue.dev/
Apache License 2.0
15.81k stars 1.2k forks source link

index wait forever to complete #1964

Open wensimin opened 1 month ago

wensimin commented 1 month ago

Before submitting your bug report

Relevant environment info

- OS: windows 11
- Continue: 0.0.56
- IDE: IntelliJ IDEA 2024.1.3 (Ultimate Edition)
- Model: shaw/dmeta-embedding-zh
- config.json:

  "embeddingsProvider": {
    "provider": "ollama",
    "model": "shaw/dmeta-embedding-zh"
  },

Description

Index the continue code base. index wait forever to complete. every time I restart the idea, it will stagnate at about 20%-40%

1723102063018

To reproduce

use ollama embedding model ,this is ollama logs

1723101191058

continue log no error

Log output

[info] Starting Continue core...
[2024-08-08T07:06:02] [info] Starting Continue core... 
[2024-08-08T07:06:02] Setup 
[2024-08-08T07:06:02] Core started 
[2024-08-08T07:06:12] Chunk with more than 500 tokens constructed:  C:/Users/hukai/IdeaProjects/continue\core\llm\llamaTokenizer.js 526
[2024-08-08T07:06:12] Chunk with more than 500 tokens constructed:  C:/Users/hukai/IdeaProjects/continue\core\llm\llms\Replicate.ts 1103
[2024-08-08T07:07:23] Chunk with more than 500 tokens constructed:  C:/Users/hukai/IdeaProjects/continue\core\llm\llamaTokenizer.js 526
[2024-08-08T07:07:28] Chunk with more than 500 tokens constructed:  C:/Users/hukai/IdeaProjects/continue\core\llm\llms\Replicate.ts 1103
Cheizr commented 1 month ago

To add to this issue, everytime i save a file, continue indexes all the files again.

fry69 commented 1 month ago

To add to this issue, every time i save a file, continue indexes all the files again

This is a bug/regression, already fixed here -> https://github.com/continuedev/continue/pull/1960 it will be included in the next Pre-Release.