gpt22gpt2-gpt2-xl-cnn-dailymail-seed42
This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.7421
- Rouge1: 0.3138
- Rouge2: 0.1176
- Rougel: 0.1915
- Rougelsum: 0.2853
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- num_epochs: 3.0
Training results
Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum |
---|---|---|---|---|---|---|---|
2.5134 | 0.2229 | 2000 | 2.3245 | 0.2387 | 0.0579 | 0.1507 | 0.2246 |
2.1313 | 0.4458 | 4000 | 1.9736 | 0.2856 | 0.0953 | 0.1779 | 0.2701 |
1.9255 | 0.6687 | 6000 | 1.8098 | 0.2633 | 0.0883 | 0.1681 | 0.2467 |
1.8411 | 0.8916 | 8000 | 1.7341 | 0.3111 | 0.1184 | 0.1924 | 0.2935 |
1.4487 | 1.1145 | 10000 | 1.7213 | 0.2883 | 0.1042 | 0.1836 | 0.2725 |
1.4506 | 1.3374 | 12000 | 1.7006 | 0.2987 | 0.1074 | 0.1863 | 0.2792 |
1.4432 | 1.5603 | 14000 | 1.6813 | 0.3236 | 0.1303 | 0.2032 | 0.3014 |
1.4367 | 1.7832 | 16000 | 1.6605 | 0.3371 | 0.1379 | 0.2086 | 0.3155 |
1.3804 | 2.0061 | 18000 | 1.7213 | 0.3080 | 0.1142 | 0.1899 | 0.2840 |
1.0104 | 2.2290 | 20000 | 1.7530 | 0.3192 | 0.1230 | 0.1970 | 0.2941 |
1.0021 | 2.4519 | 22000 | 1.7514 | 0.3046 | 0.1130 | 0.1868 | 0.2726 |
0.9945 | 2.6748 | 24000 | 1.7402 | 0.3137 | 0.1186 | 0.1922 | 0.2841 |
0.9898 | 2.8977 | 26000 | 1.7421 | 0.3138 | 0.1176 | 0.1915 | 0.2853 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0
- Datasets 2.21.0
- Tokenizers 0.19.1
- Downloads last month
- 233
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.