commit files to HF hub

README.md

@@ -31,25 +31,25 @@ model-index:
     metrics:
     - name: BLEU4 (Question Answering)
       type: bleu4_question_answering
-      value: 37.41
+      value: 31.34
     - name: ROUGE-L (Question Answering)
       type: rouge_l_question_answering
-      value: 75.9
+      value: 70.66
     - name: METEOR (Question Answering)
       type: meteor_question_answering
-      value: 54.68
+      value: 50.53
     - name: BERTScore (Question Answering)
       type: bertscore_question_answering
-      value: 97.07
+      value: 96.27
     - name: MoverScore (Question Answering)
       type: moverscore_question_answering
-      value: 91.88
+      value: 90.04
     - name: AnswerF1Score (Question Answering)
       type: answer_f1_score__question_answering
-      value: 80.37
+      value: 74.71
     - name: AnswerExactMatch (Question Answering)
       type: answer_exact_match_question_answering
-      value: 73.69
+      value: 68.16
 ---
 
 # Model Card of `vocabtrimmer/mt5-small-trimmed-ko-30000-koquad-qa`
@@ -93,16 +93,16 @@ output = pipe("question: 매드 클라운이 참가해 큰 화제를 모았던
 
 |                  |   Score | Type    | Dataset                                                          |
 |:-----------------|--------:|:--------|:-----------------------------------------------------------------|
-| AnswerExactMatch |   73.69 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| AnswerF1Score    |   80.37 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| BERTScore        |   97.07 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| Bleu_1           |   70.24 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| Bleu_2           |   61.81 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| Bleu_3           |   51.41 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| Bleu_4           |   37.41 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| METEOR           |   54.68 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| MoverScore       |   91.88 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
-| ROUGE_L          |   75.9  | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| AnswerExactMatch |   68.16 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| AnswerF1Score    |   74.71 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| BERTScore        |   96.27 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| Bleu_1           |   64.73 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| Bleu_2           |   56.16 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| Bleu_3           |   45.24 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| Bleu_4           |   31.34 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| METEOR           |   50.53 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| MoverScore       |   90.04 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
+| ROUGE_L          |   70.66 | default | [lmqg/qg_koquad](https://huggingface.co/datasets/lmqg/qg_koquad) |
 
 
 
@@ -118,11 +118,11 @@ The following hyperparameters were used during fine-tuning:
  - max_length: 512
  - max_length_output: 32
  - epoch: 5
- - batch: 32
+ - batch: 64
  - lr: 0.001
  - fp16: False
  - random_seed: 1
- - gradient_accumulation_steps: 2
+ - gradient_accumulation_steps: 1
  - label_smoothing: 0.15
 
 The full configuration can be found at [fine-tuning config file](https://huggingface.co/vocabtrimmer/mt5-small-trimmed-ko-30000-koquad-qa/raw/main/trainer_config.json).

eval/metric.first.answer.paragraph_question.answer.lmqg_qg_koquad.default.json

@@ -1 +1 @@
-{"validation": {"Bleu_1": 0.6992903294141829, "Bleu_2": 0.6127222228962876, "Bleu_3": 0.5083515309130512, "Bleu_4": 0.3635940594823205, "METEOR": 0.5462332421668936, "ROUGE_L": 0.7541467112803363, "BERTScore": 0.9710218722812839, "MoverScore": 0.9181111629442422, "AnswerF1Score": 79.72134499367644, "AnswerExactMatch": 73.06625043357613}, "test": {"Bleu_1": 0.7024203558129186, "Bleu_2": 0.6180556794660192, "Bleu_3": 0.5140638613931836, "Bleu_4": 0.37413409334289116, "METEOR": 0.5468391612861541, "ROUGE_L": 0.7589999956094918, "BERTScore": 0.970729525474637, "MoverScore": 0.9188335398437054, "AnswerF1Score": 80.36954056464982, "AnswerExactMatch": 73.69060006937218}}
+{"validation": {"Bleu_1": 0.6537011097941328, "Bleu_2": 0.5735758572983447, "Bleu_3": 0.4767258223922961, "Bleu_4": 0.3453850019905082, "METEOR": 0.5026406437140047, "ROUGE_L": 0.6970146393881299, "BERTScore": 0.9623686538113102, "MoverScore": 0.8978806396080128, "AnswerF1Score": 73.82546289414114, "AnswerExactMatch": 66.89212625737079}, "test": {"Bleu_1": 0.6472518457751592, "Bleu_2": 0.5615554084794614, "Bleu_3": 0.4523617968173102, "Bleu_4": 0.31339085747018447, "METEOR": 0.5052834772276389, "ROUGE_L": 0.7066270889175926, "BERTScore": 0.9626775134407808, "MoverScore": 0.9004137617009108, "AnswerF1Score": 74.70867716705372, "AnswerExactMatch": 68.15816857440167}}