TomPei commited on
Commit
32af9ba
1 Parent(s): f7e2888

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +14 -0
README.md CHANGED
@@ -84,6 +84,20 @@ We utilized OpenCSG's enterprise-grade large language model, csg-wukong-enterpri
84
 
85
  We recorded 100,000 data samples along with their scores, creating the dataset `fineweb_edu_classifier_chinese_data`. Using the scores from this dataset as labels, we trained a Chinese BERT model, `fineweb_edu_classifier_chinese`, which can assign a score of 0-5 to each input text. We plan to further optimize this scoring model, and in the future, the OpenCSG algorithm team will open-source the `fineweb_edu_classifier_chinese_data` and the `fineweb_edu_classifier_chinese scoring model` to further promote community development and collaboration. This dataset contains meticulously annotated and scored educational text data, providing high-quality training data for researchers and developers.
86
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
87
  **We warmly invite developers and researchers interested in this field to follow and engage with the community, working together to advance the technology. Stay tuned for the open-source release of the dataset!**
88
 
89
  ## License Agreement
 
84
 
85
  We recorded 100,000 data samples along with their scores, creating the dataset `fineweb_edu_classifier_chinese_data`. Using the scores from this dataset as labels, we trained a Chinese BERT model, `fineweb_edu_classifier_chinese`, which can assign a score of 0-5 to each input text. We plan to further optimize this scoring model, and in the future, the OpenCSG algorithm team will open-source the `fineweb_edu_classifier_chinese_data` and the `fineweb_edu_classifier_chinese scoring model` to further promote community development and collaboration. This dataset contains meticulously annotated and scored educational text data, providing high-quality training data for researchers and developers.
86
 
87
+
88
+
89
+ # abaltion experiments
90
+ After meticulously designed ablation studies, we aimed to contrast the effects between the Chinese-fineweb-edu dataset and traditional Chinese pre-training corpora.
91
+ For this purpose, we randomly selected samples from five datasets—CCI2-Data, SkyPile-150B, TeleChat-PTD, IndustryCorpus, and MAP-CC—proportional to the Chinese-fineweb-edu dataset, constructing a comparison dataset named chinese-random-select.
92
+ In our experiments, we utilized a model with 2.1 billion parameters, training it for 65k steps on both datasets respectively.
93
+ Throughout the training, we periodically saved checkpoints of the model and conducted validations on Chinese evaluation benchmarks CEval and CMMLU.
94
+ The graph below displays the performance trends of these two datasets in evaluation tasks.
95
+ The results distinctly show that the dataset trained on Chinese-fineweb-edu significantly outperforms the chinese-random-select dataset in both evaluation tasks, especially demonstrating considerable advantages in the later stages of training. This underscores the effectiveness and adaptability of Chinese-fineweb-edu in Chinese language tasks. Furthermore, these experimental outcomes also highlight the critical impact of dataset selection and construction on the ultimate performance of models.
96
+ <p align="center">
97
+ <img width="900px" alt="experiment" src="./chinese-fineweb-benchmark.png">
98
+ </p>
99
+
100
+
101
  **We warmly invite developers and researchers interested in this field to follow and engage with the community, working together to advance the technology. Stay tuned for the open-source release of the dataset!**
102
 
103
  ## License Agreement