ljb121002 commited on
Commit
7e29626
1 Parent(s): ac29e72

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -0
README.md ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Dataset description
2
+
3
+ This release incorporates the complete data sequence used in our CrystalCoder training, covering data sequences from the three pre-training stages. The data is a combination of two previous works: [SlimPajama dataset](https://huggingface.co/datasets/cerebras/SlimPajama-627B) and [StarCoder](https://huggingface.co/datasets/bigcode/starcoderdata). We divide them across three stages with different weights.
4
+
5
+ ## Stage 1
6
+ During this stage, we utilize half of the [SlimPajama data](https://huggingface.co/datasets/cerebras/SlimPajama-627B).
7
+
8
+ ## Stage 2
9
+ In the second stage, the remaining half of the [SlimPajama data](https://huggingface.co/datasets/cerebras/SlimPajama-627B) is employed, along with two epochs of [StarCoder data](https://huggingface.co/datasets/bigcode/starcoderdata). For StarCoder data, we implement [FIM augmentation](https://arxiv.org/abs/2207.14255) with an FIM rate of 0.9.
10
+
11
+ ## Stage 3
12
+ The third stage involves the reuse of Python, HTML, CSS, and JavaScript data from the [StarCoder data](https://huggingface.co/datasets/bigcode/starcoderdata) for training over three epochs. In this stage, FIM with a rate of 0.3 is applied. Additionally, a small portion of the SlimPajama dataset except the Github part is also reused.
13
+
14
+ # Primary usage
15
+
16
+ This dataset is used for our training of our CrystalCoder, and also for further reproduction.
17
+
18
+ # License: