github/owl
1
0
Fork 0
mirror of https://github.com/camel-ai/owl.git synced 2026-03-22 05:57:17 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
lazychih114
2025-06-10 00:15:36 +08:00
parent 2fae29d101
commit 599c8bafcb
1 changed files with 5 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
README.md
View File

@@ -603,19 +603,19 @@ The web interface is built using Gradio and runs locally on your machine. No dat
# 🧪 Experiments
To reproduce OWL's GAIA benchmark score of 58.18:
Furthermore, to ensure optimal performance on the GAIA benchmark, please note that our `gaia58.18` branch includes a customized version of the CAMEL framework in the `owl/camel` directory. This version contains enhanced toolkits with improved stability for gaia benchmark compared to the standard CAMEL installation.
To reproduce OWL's GAIA benchmark score:
Furthermore, to ensure optimal performance on the GAIA benchmark, please note that our `gaia69` branch includes a customized version of the CAMEL framework in the `owl/camel` directory. This version contains enhanced toolkits with improved stability for gaia benchmark compared to the standard CAMEL installation.
When running the benchmark evaluation:
1. Switch to the `gaia58.18` branch:
1. Switch to the `gaia69` branch:
```bash
git checkout gaia58.18
git checkout gaia69
```
2. Run the evaluation script:
```bash
python run_gaia_roleplaying.py
python run_gaia_workforce_claude.py
```
This will execute the same configuration that achieved our top-ranking performance on the GAIA benchmark.