Merge branch 'main' of https://huggingface.co/spaces/BramVanroy/open_dutch_llm_leaderboard
Browse files- README.md +15 -0
- content.py +18 -7
- evals/mmlu/mmlu_nl_GEITje-7B-chat.json +23 -0
- evals/mmlu/mmlu_nl_GEITje-7B.json +23 -0
README.md
CHANGED
|
@@ -8,3 +8,18 @@ sdk_version: 4.11.0
|
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
app_file: app.py
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
+
|
| 12 |
+
If you use the results of this leaderboard or refer to the leaderboard as a whole, please cite:
|
| 13 |
+
|
| 14 |
+
Vanroy, B. (2023). *Language Resources for Dutch Large Language Modelling*. [https://arxiv.org/abs/2312.12852](https://arxiv.org/abs/2312.12852)
|
| 15 |
+
|
| 16 |
+
```bibtext
|
| 17 |
+
@misc{vanroy2023language,
|
| 18 |
+
title={Language Resources for {Dutch} Large Language Modelling},
|
| 19 |
+
author={Bram Vanroy},
|
| 20 |
+
year={2023},
|
| 21 |
+
eprint={2312.12852},
|
| 22 |
+
archivePrefix={arXiv},
|
| 23 |
+
primaryClass={cs.CL}
|
| 24 |
+
}
|
| 25 |
+
```
|
content.py
CHANGED
|
@@ -39,21 +39,32 @@ This leaderboard has borrowed heavily from the following sources:
|
|
| 39 |
"""
|
| 40 |
|
| 41 |
|
| 42 |
-
CITATION =
|
| 43 |
|
| 44 |
|
| 45 |
If you use or cite the Dutch benchmark results or this specific leaderboard page, please cite the following paper:
|
| 46 |
|
| 47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 48 |
|
| 49 |
|
| 50 |
If you use the multilingual benchmarks, please cite the following paper:
|
| 51 |
|
| 52 |
```bibtex
|
| 53 |
-
@misc{
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
year={
|
| 57 |
-
}
|
| 58 |
```
|
| 59 |
"""
|
|
|
|
| 39 |
"""
|
| 40 |
|
| 41 |
|
| 42 |
+
CITATION = """## Citation
|
| 43 |
|
| 44 |
|
| 45 |
If you use or cite the Dutch benchmark results or this specific leaderboard page, please cite the following paper:
|
| 46 |
|
| 47 |
+
Vanroy, B. (2023). *Language Resources for Dutch Large Language Modelling*. [https://arxiv.org/abs/2312.12852](https://arxiv.org/abs/2312.12852)
|
| 48 |
+
|
| 49 |
+
```bibtext
|
| 50 |
+
@misc{vanroy2023language,
|
| 51 |
+
title={Language Resources for {Dutch} Large Language Modelling},
|
| 52 |
+
author={Bram Vanroy},
|
| 53 |
+
year={2023},
|
| 54 |
+
eprint={2312.12852},
|
| 55 |
+
archivePrefix={arXiv},
|
| 56 |
+
primaryClass={cs.CL}
|
| 57 |
+
}
|
| 58 |
+
```
|
| 59 |
|
| 60 |
|
| 61 |
If you use the multilingual benchmarks, please cite the following paper:
|
| 62 |
|
| 63 |
```bibtex
|
| 64 |
+
@misc{lai2023openllmbenchmark,
|
| 65 |
+
title={Open Multilingual {LLM} Evaluation Leaderboard},
|
| 66 |
+
author={Viet Lai and Nghia Trung Ngo and Amir Pouran Ben Veyseh and Franck Dernoncourt and Thien Huu Nguyen},
|
| 67 |
+
year={2023}
|
| 68 |
+
}
|
| 69 |
```
|
| 70 |
"""
|
evals/mmlu/mmlu_nl_GEITje-7B-chat.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"mmlu_nl": {
|
| 4 |
+
"acc": 0.38362297943386203,
|
| 5 |
+
"acc_stderr": 0.004236270988211309,
|
| 6 |
+
"acc_norm": 0.3342187144266525,
|
| 7 |
+
"acc_norm_stderr": 0.004109503747402399
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
"versions": {
|
| 11 |
+
"mmlu_nl": 0
|
| 12 |
+
},
|
| 13 |
+
"config": {
|
| 14 |
+
"model": "hf-auto",
|
| 15 |
+
"model_args": "pretrained=Rijgersberg/GEITje-7B-chat,use_accelerate=True,device_map_option=auto,dtype=auto,load_in_8bit=True",
|
| 16 |
+
"batch_size": 2,
|
| 17 |
+
"device": "cuda",
|
| 18 |
+
"no_cache": false,
|
| 19 |
+
"limit": null,
|
| 20 |
+
"bootstrap_iters": 100000,
|
| 21 |
+
"description_dict": {}
|
| 22 |
+
}
|
| 23 |
+
}
|
evals/mmlu/mmlu_nl_GEITje-7B.json
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"results": {
|
| 3 |
+
"mmlu_nl": {
|
| 4 |
+
"acc": 0.38202929346588754,
|
| 5 |
+
"acc_stderr": 0.004232924126485795,
|
| 6 |
+
"acc_norm": 0.32488426804280185,
|
| 7 |
+
"acc_norm_stderr": 0.004080014081775866
|
| 8 |
+
}
|
| 9 |
+
},
|
| 10 |
+
"versions": {
|
| 11 |
+
"mmlu_nl": 0
|
| 12 |
+
},
|
| 13 |
+
"config": {
|
| 14 |
+
"model": "hf-auto",
|
| 15 |
+
"model_args": "pretrained=Rijgersberg/GEITje-7B,use_accelerate=True,device_map_option=auto,dtype=auto,load_in_8bit=True",
|
| 16 |
+
"batch_size": 2,
|
| 17 |
+
"device": "cuda",
|
| 18 |
+
"no_cache": false,
|
| 19 |
+
"limit": null,
|
| 20 |
+
"bootstrap_iters": 100000,
|
| 21 |
+
"description_dict": {}
|
| 22 |
+
}
|
| 23 |
+
}
|