Text Classification
Transformers
Safetensors
Chinese
qwen3
text-generation
cross-encoder
reranker
event-coreference
event-deduplication
chinese
news-clustering
knowledge-distillation
calibrated-probabilities
text-embeddings-inference
Instructions to use MaYiding/EventTwin with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use MaYiding/EventTwin with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="MaYiding/EventTwin")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("MaYiding/EventTwin") model = AutoModelForCausalLM.from_pretrained("MaYiding/EventTwin", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Commit History
EventTwin v2.3 weights: calibration.json 6f16377 verified
EventTwin v2.3 weights: model.safetensors ff73e39 verified
EventTwin v2.2 model card: cross-granularity records + dual-traffic profile aba6cdd verified
EventTwin v2.2 weights: calibration.json 740eb1d verified
EventTwin v2.2 weights: model.safetensors 8675d87 verified
EventTwin v2.1 model card: single model beats ensemble (AUROC 0.983) 69ef178 verified
EventTwin v2.1 weights: calibration.json 16028dc verified
EventTwin v2.1 weights: model.safetensors 4758a5f verified
Upload folder using huggingface_hub 0667a77 verified
Upload README.md with huggingface_hub 9291335 verified
Upload folder using huggingface_hub 09f3932 verified
Upload folder using huggingface_hub 455cbb8 verified
EventTwin v1.2 (internal v10a): LS=0.2 + R-Drop + EMA — strongest pos layer (0.847), zero high-confidence false merges fce54b5
MaYiding commited on
EventTwin v1.1 (internal v9): +7pt pos-layer AUROC via EMA + synth-v2 + counterfactual filtering (T=0.625) b82f40f
MaYiding commited on
EventTwin v1.0: calibrated Chinese same-event cross-encoder (568M, T=0.824) d7204f3
MaYiding commited on