Text Classification
Transformers
Safetensors
English
roberta
toxicity
llada
distillation
custom_code
text-embeddings-inference
Instructions to use kl1/roberta_toxicity_classifier_LLaDA with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use kl1/roberta_toxicity_classifier_LLaDA with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="kl1/roberta_toxicity_classifier_LLaDA", trust_remote_code=True)# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("kl1/roberta_toxicity_classifier_LLaDA", trust_remote_code=True) model = AutoModelForSequenceClassification.from_pretrained("kl1/roberta_toxicity_classifier_LLaDA", trust_remote_code=True) - Notebooks
- Google Colab
- Kaggle
| { | |
| "step": 20000, | |
| "metrics": { | |
| "supervised_ce": 0.22887812554836273, | |
| "labeled_fraction": 1.0, | |
| "label_accuracy": 0.9453125, | |
| "label_mae": 0.0588543675839901, | |
| "label_precision": 0.7674418687820435, | |
| "label_recall": 0.6470588445663452, | |
| "label_f1": 0.7021276354789734, | |
| "label_gold_positive_rate": 0.099609375, | |
| "label_pred_positive_rate": 0.083984375, | |
| "loss": 0.27317333221435547, | |
| "kl": 0.22147610783576965, | |
| "ce": 0.07613664120435715, | |
| "teacher_ce": 0.07613664120435715, | |
| "toxic_mae": 0.05239273980259895, | |
| "toxic_rmse": 0.12058839946985245, | |
| "toxic_max_abs": 0.7615423202514648, | |
| "label_match": 0.986328125, | |
| "teacher_label_match": 0.986328125, | |
| "lr": 0.0, | |
| "epoch": 6.0, | |
| "step": 20000.0, | |
| "val_supervised_ce": 0.24059099189263375, | |
| "val_labeled_fraction": 1.0, | |
| "val_label_accuracy": 0.9560078190392064, | |
| "val_label_mae": 0.06301387556438685, | |
| "val_label_precision": 0.7140320075305637, | |
| "val_label_recall": 0.7808826945852421, | |
| "val_label_f1": 0.743371362151996, | |
| "val_label_gold_positive_rate": 0.08225244853388948, | |
| "val_label_pred_positive_rate": 0.08995947956150471, | |
| "val_loss": 0.2812121282663035, | |
| "val_kl": 0.20310567812750818, | |
| "val_ce": 0.07181113709131223, | |
| "val_teacher_ce": 0.07181113709131223, | |
| "val_toxic_mae": 0.04778627983770471, | |
| "val_toxic_rmse": 0.10417259909648449, | |
| "val_toxic_max_abs": 0.625468818201666, | |
| "val_label_match": 0.9860112805199612, | |
| "val_teacher_label_match": 0.9860112805199612, | |
| "val_label_global_accuracy": 0.956007819022215, | |
| "val_label_global_f1": 0.7445462607153414, | |
| "val_label_global_precision": 0.7126527840651878, | |
| "val_label_global_recall": 0.779428147047902, | |
| "val_label_global_pred_positive_rate": 0.08995947986841202, | |
| "val_label_best_threshold": 0.5377818942070007, | |
| "val_label_best_accuracy": 0.9578811264278878, | |
| "val_label_best_f1": 0.7477285200317092, | |
| "val_label_best_precision": 0.7368990384615385, | |
| "val_label_best_recall": 0.7588810496348558, | |
| "val_label_best_pred_positive_rate": 0.08470607548952103, | |
| "val_label_roc_auc": 0.9761634605252926, | |
| "val_label_pr_auc": 0.8327706031677701 | |
| } | |
| } |