jiazhengli commited on
Commit
8f97e33
1 Parent(s): 0ffb2d3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +39 -21
README.md CHANGED
@@ -5,43 +5,61 @@ tags:
5
  - llama-factory
6
  - lora
7
  - generated_from_trainer
8
- base_model: mistralai/Mixtral-8x7B-Instruct-v0.1
 
 
9
  model-index:
10
  - name: sft_trained_woaqa_mixtral_dpo
11
  results: []
 
 
 
 
 
 
 
 
12
  ---
13
 
14
- <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
- should probably proofread and complete it, then remove this comment. -->
16
 
17
- # sft_trained_woaqa_mixtral_dpo
18
 
19
- This model is a fine-tuned version of [mistralai/Mixtral-8x7B-Instruct-v0.1](https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1) on the sft_wo_aqa_pref dataset.
20
- It achieves the following results on the evaluation set:
21
- - Loss: 2.6468
22
- - Rewards/chosen: 14.3798
23
- - Rewards/rejected: 12.6853
24
- - Rewards/accuracies: 0.6426
25
- - Rewards/margins: 1.6945
26
- - Logps/rejected: -209.4146
27
- - Logps/chosen: -167.4447
28
- - Logits/rejected: -0.8532
29
- - Logits/chosen: -0.8500
30
-
31
- ## Model description
32
-
33
- More information needed
34
 
35
  ## Intended uses & limitations
36
 
37
- More information needed
38
 
39
  ## Training and evaluation data
40
 
41
- More information needed
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
42
 
43
  ## Training procedure
44
 
 
 
45
  ### Training hyperparameters
46
 
47
  The following hyperparameters were used during training:
 
5
  - llama-factory
6
  - lora
7
  - generated_from_trainer
8
+ base_model:
9
+ - jiazhengli/Mixtral-8x7B-Instruct-v0.1-QLoRA-Assessment-Rationale-sft
10
+ - mistralai/Mixtral-8x7B-Instruct-v0.1
11
  model-index:
12
  - name: sft_trained_woaqa_mixtral_dpo
13
  results: []
14
+ datasets:
15
+ - jiazhengli/Rationale_MCTS
16
+ - jiazhengli/Synthetic_Rationale
17
+ language:
18
+ - en
19
+ metrics:
20
+ - accuracy
21
+ - f1
22
  ---
23
 
24
+ # Mixtral-8x7B-Instruct-v0.1-QLoRA-Assessment-Rationale-dpo
 
25
 
26
+ The model trained with w/o private data from the EMNLP 2024 Paper: Calibrating LLMs with Preference Optimization on Thought Trees for Generating Rationale in Science Question Scoring.
27
 
28
+ - **Paper:** [Calibrating LLMs with Preference Optimization on Thought Trees for Generating Rationale in Science Question Scoring](https://arxiv.org/abs/2406.19949) (EMNLP 2024 Findings)
29
+ - **GitHub Repository:** [Thought Tree Assessment Repository](https://github.com/lijiazheng99/thought_tree_assessment)
 
 
 
 
 
 
 
 
 
 
 
 
 
30
 
31
  ## Intended uses & limitations
32
 
33
+ This model offers a valuable resource for research in explainable AI within educational technology. The model is trained with **noisy** response-level rationales. This makes them **unsuitable** for direct application in high-stakes assessments without additional verification.
34
 
35
  ## Training and evaluation data
36
 
37
+ We trained and evaluated the model on the [Synthetic Rationale data](https://huggingface.co/datasets/jiazhengli/Synthetic_Rationale), which was generated from the [Rationale MCTS data](https://huggingface.co/datasets/jiazhengli/Rationale_MCTS).
38
+
39
+ To extract scores from rationales, please use the [jiazhengli/deberta-v3-large-Rationale-to-Score](https://huggingface.co/jiazhengli/deberta-v3-large-Rationale-to-Score).
40
+
41
+ ## Citation
42
+
43
+ Please cite the following work if you utilize this model:
44
+
45
+ **BibTeX:**
46
+
47
+ ```bibtex
48
+ @misc{li2024calibratingllmspreferenceoptimization,
49
+ title={Calibrating LLMs with Preference Optimization on Thought Trees for Generating Rationale in Science Question Scoring},
50
+ author={Jiazheng Li and Hainiu Xu and Zhaoyue Sun and Yuxiang Zhou and David West and Cesare Aloisi and Yulan He},
51
+ year={2024},
52
+ eprint={2406.19949},
53
+ archivePrefix={arXiv},
54
+ primaryClass={cs.CL},
55
+ url={https://arxiv.org/abs/2406.19949},
56
+ }
57
+ ```
58
 
59
  ## Training procedure
60
 
61
+ Please refer to our [paper](https://arxiv.org/abs/2406.19949).
62
+
63
  ### Training hyperparameters
64
 
65
  The following hyperparameters were used during training: