YuxinJiang
/

debate-alpaca-lora_7b_001

Model card Files Files and versions Community

YuxinJiang commited on Apr 6, 2023

Commit

e70964d

·

1 Parent(s): 30c605c

Update README.md

Files changed (1) hide show

README.md +2 -1

README.md CHANGED Viewed

@@ -34,7 +34,8 @@ Response: 1. AI will replace teachers and thus reduce the number of jobs in this
 ## Data Collection
 We collected 1,560 discussion threads before January 2020 of Kialo. Each discussion forms an argument tree as the following figure shows. Except the thesis, every claim in the argument tree either opposes or supports its parent claim. Moreover, each claim has impact votes assigned by the users of the platform. The impact votes evaluate how impactful a claim is.
-<img src="https://github.com/YJiangcm/Debate-alpaca-lora/blob/master/pics/kialo_example.png" width="600" height="300">
 We extract each argument with its supported children claims or opposed children claims to form a training example. The children claims are ranked by the descending order of impact. We delete children claims with votes less than 5 or average impact score less than 2. Finally, we obtain 20,998 [training data](https://github.com/YJiangcm/Debate-alpaca-lora/blob/master/kialo_debate.json) formuated as the following examples:
 ```

 ## Data Collection
 We collected 1,560 discussion threads before January 2020 of Kialo. Each discussion forms an argument tree as the following figure shows. Except the thesis, every claim in the argument tree either opposes or supports its parent claim. Moreover, each claim has impact votes assigned by the users of the platform. The impact votes evaluate how impactful a claim is.
+![Debate-alpaca-lora](https://github.com/YJiangcm/Debate-alpaca-lora/raw/master/pics/kialo_example.png)
 We extract each argument with its supported children claims or opposed children claims to form a training example. The children claims are ranked by the descending order of impact. We delete children claims with votes less than 5 or average impact score less than 2. Finally, we obtain 20,998 [training data](https://github.com/YJiangcm/Debate-alpaca-lora/blob/master/kialo_debate.json) formuated as the following examples:
 ```