# Load model directly
from transformers import AutoTokenizer, AutoModelForMaskedLM
tokenizer = AutoTokenizer.from_pretrained("Billwzl/roberta-base-IMDB_roberta")
model = AutoModelForMaskedLM.from_pretrained("Billwzl/roberta-base-IMDB_roberta", device_map="auto")Quick Links
roberta-base-IMDB_roberta
This model is a fine-tuned version of roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.1897
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 16
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.7882 | 1.0 | 1250 | 2.4751 |
| 2.5749 | 2.0 | 2500 | 2.4183 |
| 2.4501 | 3.0 | 3750 | 2.3799 |
| 2.3697 | 4.0 | 5000 | 2.3792 |
| 2.3187 | 5.0 | 6250 | 2.3622 |
| 2.24 | 6.0 | 7500 | 2.3491 |
| 2.164 | 7.0 | 8750 | 2.3146 |
| 2.1187 | 8.0 | 10000 | 2.2804 |
| 2.0552 | 9.0 | 11250 | 2.2629 |
| 2.0285 | 10.0 | 12500 | 2.2088 |
| 1.9807 | 11.0 | 13750 | 2.2061 |
| 1.9597 | 12.0 | 15000 | 2.2094 |
| 1.9062 | 13.0 | 16250 | 2.1486 |
| 1.8766 | 14.0 | 17500 | 2.1348 |
| 1.8528 | 15.0 | 18750 | 2.1665 |
| 1.8425 | 16.0 | 20000 | 2.1897 |
Framework versions
- Transformers 4.21.1
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1
- Downloads last month
- 4
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("fill-mask", model="Billwzl/roberta-base-IMDB_roberta")