Text Classification
Transformers
PyTorch
Arabic
bert
hate-speech
gender-based-violence
arabic
binary-classification
pilot
Eval Results (legacy)
text-embeddings-inference
Instructions to use thejosango/nuha-binary with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use thejosango/nuha-binary with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="thejosango/nuha-binary")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("thejosango/nuha-binary") model = AutoModelForSequenceClassification.from_pretrained("thejosango/nuha-binary") - Notebooks
- Google Colab
- Kaggle
File size: 671 Bytes
3c6e625 b13ab70 3c6e625 aa7b772 152141c 3c6e625 cc3ed62 f54519f 4a30404 b13ab70 3c6e625 4c790e4 2dfdcee bf5de58 4c790e4 6aaa15d f54519f aca5e41 2dfdcee ee0cc92 cc3ed62 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 | [experiment]
name = "binary-56"
type = "binary"
[dataset]
path = "thejosango/nuha-dataset"
dataset_revision = "main"
augment_ratio = 0.75
undersampling_strategy = false
[model]
pretrained_model_name_or_path = "thejosango/nuha-mlm"
revision = "ce20f497544665775129f9ff5b3cd2a3e350dce8"
num_hidden_layers = 4
classifier_dropout = 0.50
[training]
num_train_epochs = 5
warmup_steps = 0
lr_scheduler_type = "linear"
learning_rate = 5e-5
per_device_train_batch_size = 64
per_device_eval_batch_size = 64
gradient_accumulation_steps = 1
weight_decay = 1e-3
label_smoothing_factor = 0.1
weighted_loss = true
early_stopping_patience = 5
early_stopping_threshold = 0.005
|