metadata
language:
- ko
- en
license: mit
Model Card for free-evo-qwen72b-v0.8
1st place : 2024 4th May - avg. 81.28
Open Llm Leaderboard but this kicked away. maybe the explanation was not enough. i don't care.
Method
- We were inspired by this Sakana project
Process
1. two models with the same architecture are needed so fine-tune a model to create a gap between the two of them.
2. merge original one and fine-tuned one
3. evaluate the merged model
4. merge again it with original model
5. evaluate again
6. keep going until evaluate avg is higher then original one
that's it. simple.
Base Architecture
- QWEN2