Instructions to use SergeySavinov/rubert-tiny-toxicity with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use SergeySavinov/rubert-tiny-toxicity with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="SergeySavinov/rubert-tiny-toxicity")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("SergeySavinov/rubert-tiny-toxicity") model = AutoModelForSequenceClassification.from_pretrained("SergeySavinov/rubert-tiny-toxicity") - Notebooks
- Google Colab
- Kaggle
| { | |
| "model_name": "cointegrated/rubert-tiny", | |
| "max_length": 512, | |
| "learning_rate": 5e-05, | |
| "batch_size": 256, | |
| "epochs": 15, | |
| "weight_decay": 0.001, | |
| "warmup_steps": 0, | |
| "warmup_ratio": 0, | |
| "lr_scheduler_type": "plateau", | |
| "lr_schedule_patience": 1, | |
| "lr_schedule_factor": 0.2, | |
| "max_grad_norm": 1.0, | |
| "freeze_encoder": false, | |
| "freeze_last_n_layers": 0, | |
| "best_score": 0.9695058551043865, | |
| "best_score_metric": "average_precision", | |
| "optimal_threshold": 0.6828358173370361, | |
| "random_state": 42, | |
| "loss_type": "focal", | |
| "focal_gamma": 2.0, | |
| "focal_alpha": null, | |
| "focal_auto_alpha": true | |
| } |