Update README.md
Browse files
README.md
CHANGED
@@ -14,17 +14,19 @@ license: cc-by-sa-4.0
|
|
14 |
2. And the version without "Original text paraphrasing" is added.
|
15 |
3. The old version has been moved to "legacy" branch.
|
16 |
|
|
|
|
|
|
|
17 |
# Long Context Instruction-tuning dataset with "Original text paraphrasing"
|
18 |
* consist of multiple tasks
|
19 |
* Chinese and English
|
20 |
* sample length ranging from 4k to 32k
|
21 |
* the answer contains "Original text paraphrasing" part
|
22 |
-
|
23 |
|
24 |
# 长文本指令微调数据
|
25 |
* 此数据集由多种长文本任务数据集组合而成。
|
26 |
* 包含中文和英文两部分
|
27 |
-
* [Paper](https://arxiv.org/abs/2312.11193)
|
28 |
|
29 |
![](The_Proportion_of_Different_Tasks.png)
|
30 |
|
@@ -53,4 +55,5 @@ license: cc-by-sa-4.0
|
|
53 |
|
54 |
2. [从ShareGPT中筛选的长对话(中英)](https://huggingface.co/datasets/yuyijiong/Sharegpt-long-conversation)
|
55 |
|
56 |
-
3. 预训练长文本语料库(中英)[LongData-Corpus
|
|
|
|
14 |
2. And the version without "Original text paraphrasing" is added.
|
15 |
3. The old version has been moved to "legacy" branch.
|
16 |
|
17 |
+
* [Paper](https://arxiv.org/abs/2312.11193)
|
18 |
+
* [Github](https://github.com/yuyijiong/train_with_paraphrasing)
|
19 |
+
|
20 |
# Long Context Instruction-tuning dataset with "Original text paraphrasing"
|
21 |
* consist of multiple tasks
|
22 |
* Chinese and English
|
23 |
* sample length ranging from 4k to 32k
|
24 |
* the answer contains "Original text paraphrasing" part
|
25 |
+
|
26 |
|
27 |
# 长文本指令微调数据
|
28 |
* 此数据集由多种长文本任务数据集组合而成。
|
29 |
* 包含中文和英文两部分
|
|
|
30 |
|
31 |
![](The_Proportion_of_Different_Tasks.png)
|
32 |
|
|
|
55 |
|
56 |
2. [从ShareGPT中筛选的长对话(中英)](https://huggingface.co/datasets/yuyijiong/Sharegpt-long-conversation)
|
57 |
|
58 |
+
3. 预训练长文本语料库(中英)[LongData-Corpus
|
59 |
+
](https://huggingface.co/datasets/yuyijiong/LongData-Corpus)
|