Text Classification
Transformers
Safetensors
Arabic
Stance Detection
Text Classification
arabic-nlp
stanceeval-2026
few-shot-learning
retrieval-augmented
Mawqif-v2
ensemble
LoRA
AraBERT
MARBERT
Instructions to use zaher-m/stanceeval2026 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use zaher-m/stanceeval2026 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="zaher-m/stanceeval2026")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("zaher-m/stanceeval2026", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| # target descriptions on for cross-target grounding; check the aug booster | |
| # separately once we have a baseline dev score | |
| model_hf: aubmindlab/bert-base-arabertv02-twitter | |
| train_csv: data/track2/train.csv | |
| dev_csv: data/track2/dev.csv | |
| out_dir: outputs/t2_arabert | |
| seed: 42 | |
| max_len: 128 | |
| batch_size: 16 | |
| eval_batch_size: 64 | |
| lr: 2.0e-5 | |
| weight_decay: 0.01 | |
| epochs: 12 | |
| patience: 3 | |
| loss: ce | |
| focal_gamma: 2.0 | |
| prep_mode: preserve | |
| use_description: true # short target description helps cross-target generalization | |
| none_bias: 0.0 | |