KJøretid
#SBATCH --nodes=2
#SBATCH --ntasks-per-node=3
#SBATCH --gres=gpu:A100m40:1
{'train_runtime': 60.0918, 'train_samples_per_second': 41.603, 'train_steps_per_second': 0.166, 'train_loss': 0.6561894416809082, 'epoch': 5.0}
Time: 60.09
Samples/second: 41.60
<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
norbert2_sentiment_norec_en_gpu_500_rader_max_noder_task
This model is a fine-tuned version of bert-large-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6280
- Compute Metrics: :
- Accuracy: 0.678
- Balanced Accuracy: 0.4889
- F1 Score: 0.8076
- Recall: 0.9713
- Precision: 0.6912
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: 2e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 256
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Compute Metrics | Accuracy | Balanced Accuracy | F1 Score | Recall | Precision |
|---|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 2 | 0.6324 | : | 0.696 | 0.5 | 0.8208 | 1.0 | 0.696 |
| No log | 2.0 | 4 | 0.6264 | : | 0.692 | 0.4971 | 0.8180 | 0.9943 | 0.6948 |
| No log | 3.0 | 6 | 0.6180 | : | 0.696 | 0.5 | 0.8208 | 1.0 | 0.696 |
| No log | 4.0 | 8 | 0.6236 | : | 0.694 | 0.5023 | 0.8185 | 0.9914 | 0.6970 |
| 0.6562 | 5.0 | 10 | 0.6280 | : | 0.678 | 0.4889 | 0.8076 | 0.9713 | 0.6912 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1+cu117
- Datasets 2.9.0
- Tokenizers 0.13.2
