Update README.md

README.md

@@ -10,37 +10,37 @@ dataset_info:
     - name: content
       dtype: string
   splits:
-  - name: train_douban_sft
+  - name: train_douban
     num_bytes: 5567696
     num_examples: 3086
-  - name: train_human_value_sft
+  - name: train_human_value
     num_bytes: 806635
     num_examples: 1007
-  - name: train_logi_qa_sft
+  - name: train_logi_qa
     num_bytes: 666517
     num_examples: 421
-  - name: train_ruozhiba_sft
+  - name: train_ruozhiba
     num_bytes: 228494
     num_examples: 240
-  - name: train_segmentfault_sft
+  - name: train_segmentfault
     num_bytes: 1068526
     num_examples: 458
-  - name: train_wiki_sft
+  - name: train_wiki
     num_bytes: 27611061
     num_examples: 10603
-  - name: train_wikihow_sft
+  - name: train_wikihow
     num_bytes: 11069103
     num_examples: 1485
-  - name: train_xhs_sft
+  - name: train_xhs
     num_bytes: 2551884
     num_examples: 1508
-  - name: train_zhihu_sft
+  - name: train_zhihu
     num_bytes: 13986060
     num_examples: 5631
-  - name: train_sft
+  - name: train
     num_bytes: 63555976
     num_examples: 24439
-  - name: test_sft
+  - name: test
     num_bytes: 228494
     num_examples: 240
   download_size: 67916523
@@ -48,27 +48,27 @@ dataset_info:
 configs:
 - config_name: default
   data_files:
-  - split: train_douban_sft
+  - split: train_douban
     path: data/train_douban_sft-*
-  - split: train_human_value_sft
+  - split: train_human_value
     path: data/train_human_value_sft-*
-  - split: train_logi_qa_sft
+  - split: train_logi_qa
     path: data/train_logi_qa_sft-*
-  - split: train_ruozhiba_sft
+  - split: train_ruozhiba
     path: data/train_ruozhiba_sft-*
-  - split: train_segmentfault_sft
+  - split: train_segmentfault
     path: data/train_segmentfault_sft-*
-  - split: train_wiki_sft
+  - split: train_wiki
     path: data/train_wiki_sft-*
-  - split: train_wikihow_sft
+  - split: train_wikihow
     path: data/train_wikihow_sft-*
-  - split: train_xhs_sft
+  - split: train_xhs
     path: data/train_xhs_sft-*
-  - split: train_zhihu_sft
+  - split: train_zhihu
     path: data/train_zhihu_sft-*
-  - split: train_sft
+  - split: train
     path: data/train_sft-*
-  - split: test_sft
+  - split: test
     path: data/test_sft-*
 task_categories:
 - question-answering
@@ -83,7 +83,7 @@ size_categories:
 数据完全来自于[COIG-CQIA](https://huggingface.co/datasets/m-a-p/COIG-CQIA)。
 
 暂时忽略了chinese_traditional，coig_pc，exam，finance这些转换麻烦或者语义上不适合当QA数据集的subset。
-其中train_sft是全集，test_sft是ruozhiba，以便代码能够跑通。
+其中train是全集，test是ruozhiba，以便代码能够跑通。
 
 经过reformat的本数据集可以直接使用[alignment-handbook](https://github.com/huggingface/alignment-handbook) 进行sft。