|
--- |
|
license: apache-2.0 |
|
library_name: transformers |
|
tags: [] |
|
--- |
|
|
|
# Nox |
|
|
|
![image/jpeg](https://cdn-uploads.huggingface.co/production/uploads/64241c3d774cc340797429fc/u60xmpVM1K0AtPXZtw8b4.jpeg) |
|
|
|
The nox project is a set of tools that make it easy to use various fine tuning technologies using solar models. |
|
We constructed ko data using grammatically accurate data.(It's not perfect, but I tried my best.) |
|
And we created nox-solar model using a fine-tuning technique(sft,dpo) Our model, the nox-solar model, ranked first on the [Open Ko-LLM Leaderboard](https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard). |
|
|
|
Currently, we are planning to make all code and datasets public. |
|
|
|
Through this, users are expected to be able to freely conduct research and development using Nox. |
|
|
|
|
|
|
|
## Model Details |
|
|
|
* **Model Developers** : davidkim(changyeon kim) |
|
* **Repository** : https://github.com/davidkim205/nox(will be updated soon.) |
|
* **base mode** : Edentns/DataVortexS-10.7B-dpo-v1.11 |
|
* **sft dataset** : komt-124k(will be updated soon.) |
|
* **dpo dataset** : comparison_v2_289k(will be updated soon.) |
|
* **evalution** : [kollm_evalution](https://github.com/davidkim205/kollm_evaluation) |
|
* **evalution dataset** : [open-ko-llm-leaderboard datasets](https://huggingface.co/collections/davidkim205/open-ko-llm-leaderboard-datasets-65eea9e87fc3ae80787ee15a) |
|
|
|
## Evaluation |
|
### [The Open Ko-LLM Leaderboard](https://huggingface.co/spaces/upstage/open-ko-llm-leaderboard) |
|
| Model | Average | Ko-ARC | Ko-HellaSwag | Ko-MMLU | Ko-TruthfulQA | Ko-CommonGen V2 | |
|
| ------------------------------ | ------- | ------ | ------------ | ------- | ------------- | --------------- | |
|
| davidkim205/nox-solar-10.7b-v4 | 67.77 | 73.55 | 72.07 | 57.93 | 79.32 | 55.96 | |
|
|
|
|
|
|
|
### [kollm_evalution](https://github.com/davidkim205/kollm_evalution) |
|
(will be updated soon.) |
|
|