metadata
license: apache-2.0
base_model: facebook/bart-base
tags:
- generated_from_trainer
datasets:
- datasets/all_binary_and_xe_ey_fae_counterfactual
metrics:
- accuracy
model-index:
- name: bart-adapter-finetuned-xe_ey_fae
results:
- task:
name: Masked Language Modeling
type: fill-mask
dataset:
name: datasets/all_binary_and_xe_ey_fae_counterfactual
type: datasets/all_binary_and_xe_ey_fae_counterfactual
metrics:
- name: Accuracy
type: accuracy
value: 0.3096946377787028
bart-adapter-finetuned-xe_ey_fae
This model is a fine-tuned version of facebook/bart-base on the datasets/all_binary_and_xe_ey_fae_counterfactual dataset. It achieves the following results on the evaluation set:
- Loss: 4.2302
- Accuracy: 0.3097
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 100
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 7.6974 | 0.06 | 500 | 6.7246 | 0.0649 |
| 6.8017 | 0.12 | 1000 | 6.4067 | 0.0762 |
| 6.5894 | 0.18 | 1500 | 6.2661 | 0.0821 |
| 6.443 | 0.24 | 2000 | 6.1350 | 0.0905 |
| 6.3245 | 0.3 | 2500 | 6.0024 | 0.1008 |
| 6.2208 | 0.35 | 3000 | 5.8518 | 0.1145 |
| 6.097 | 0.41 | 3500 | 5.6588 | 0.1330 |
| 5.9862 | 0.47 | 4000 | 5.4641 | 0.1543 |
| 5.8742 | 0.53 | 4500 | 5.3200 | 0.1707 |
| 5.7716 | 0.59 | 5000 | 5.2044 | 0.1840 |
| 5.6952 | 0.65 | 5500 | 5.1154 | 0.1952 |
| 5.6209 | 0.71 | 6000 | 5.0428 | 0.2044 |
| 5.5752 | 0.77 | 6500 | 4.9711 | 0.2136 |
| 5.5091 | 0.83 | 7000 | 4.9078 | 0.2212 |
| 5.4657 | 0.89 | 7500 | 4.8495 | 0.2287 |
| 5.4245 | 0.95 | 8000 | 4.8012 | 0.2360 |
| 5.3813 | 1.0 | 8500 | 4.7563 | 0.2409 |
| 5.3501 | 1.06 | 9000 | 4.7166 | 0.2464 |
| 5.3098 | 1.12 | 9500 | 4.6838 | 0.2501 |
| 5.2856 | 1.18 | 10000 | 4.6515 | 0.2551 |
| 5.2549 | 1.24 | 10500 | 4.6121 | 0.2602 |
| 5.2217 | 1.3 | 11000 | 4.5841 | 0.2637 |
| 5.1997 | 1.36 | 11500 | 4.5588 | 0.2674 |
| 5.1844 | 1.42 | 12000 | 4.5309 | 0.2708 |
| 5.1491 | 1.48 | 12500 | 4.4999 | 0.2748 |
| 5.1244 | 1.54 | 13000 | 4.4783 | 0.2780 |
| 5.1047 | 1.6 | 13500 | 4.4561 | 0.2812 |
| 5.0917 | 1.66 | 14000 | 4.4409 | 0.2826 |
| 5.0631 | 1.71 | 14500 | 4.4198 | 0.2851 |
| 5.0537 | 1.77 | 15000 | 4.4003 | 0.2881 |
| 5.0339 | 1.83 | 15500 | 4.3855 | 0.2899 |
| 5.0235 | 1.89 | 16000 | 4.3650 | 0.2921 |
| 5.0074 | 1.95 | 16500 | 4.3496 | 0.2942 |
| 4.9927 | 2.01 | 17000 | 4.3361 | 0.2965 |
| 4.9797 | 2.07 | 17500 | 4.3203 | 0.2981 |
| 4.9725 | 2.13 | 18000 | 4.3118 | 0.2995 |
| 4.9552 | 2.19 | 18500 | 4.2977 | 0.3012 |
| 4.956 | 2.25 | 19000 | 4.2894 | 0.3019 |
| 4.9427 | 2.31 | 19500 | 4.2781 | 0.3036 |
| 4.9337 | 2.36 | 20000 | 4.2773 | 0.3038 |
| 4.9333 | 2.42 | 20500 | 4.2624 | 0.3056 |
| 4.9173 | 2.48 | 21000 | 4.2643 | 0.3059 |
| 4.915 | 2.54 | 21500 | 4.2537 | 0.3069 |
| 4.9092 | 2.6 | 22000 | 4.2457 | 0.3084 |
| 4.9043 | 2.66 | 22500 | 4.2456 | 0.3081 |
| 4.9014 | 2.72 | 23000 | 4.2424 | 0.3087 |
| 4.8889 | 2.78 | 23500 | 4.2347 | 0.3104 |
| 4.8898 | 2.84 | 24000 | 4.2340 | 0.3095 |
| 4.8814 | 2.9 | 24500 | 4.2297 | 0.3100 |
| 4.8804 | 2.96 | 25000 | 4.2290 | 0.3095 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.2.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2