Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/da03/o1-reasoning-tokens
https://github.com/da03/o1-reasoning-tokens
Last synced: 23 days ago
JSON representation
- Host: GitHub
- URL: https://github.com/da03/o1-reasoning-tokens
- Owner: da03
- Created: 2024-10-21T16:01:06.000Z (about 1 month ago)
- Default Branch: main
- Last Pushed: 2024-10-21T19:46:44.000Z (30 days ago)
- Last Synced: 2024-10-22T06:58:09.623Z (29 days ago)
- Language: Python
- Homepage: https://x.com/yuntiandeng/status/1848421766093255027
- Size: 264 KB
- Stars: 4
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# o1-reasoning-tokens
## Overview
This repository contains code for analyzing the reasoning token counts returned by OpenAI's o1-mini model and experiments that test the effect of setting `max_completion_tokens` on reasoning tokens.
## Usage
### 1. Running the `max_completion_tokens` Experiment
To run the experiment that tests how `max_completion_tokens` affects reasoning token counts, use the following command:
```bash
python src/main.py
```### 2. Analyzing o1-mini Reasoning Tokens from WildChat
To analyze reasoning token counts for OpenAI's o1-mini model from conversations collected by the WildChat chatbot, use the following command:
```bash
python src/analyze_wildchat_reasoning_tokens_o1mini.py
```### 3. Analyzing o1-preview Reasoning Tokens from WildChat
To analyze reasoning token counts for OpenAI's o1-preview model from conversations collected by the WildChat chatbot, use the following command:
```bash
python src/analyze_wildchat_reasoning_tokens_o1preview.py
```