akhooli/ar-mmarco-sample
Updated • 36 • 1
How to use akhooli/arabic-colbertv2-250k with Transformers:
# Load model directly
from transformers import AutoTokenizer, HF_ColBERT
tokenizer = AutoTokenizer.from_pretrained("akhooli/arabic-colbertv2-250k")
model = HF_ColBERT.from_pretrained("akhooli/arabic-colbertv2-250k")This is an Arabic ColBERT2 model (not fully trained) using the akhooli/ar-mmarco-sample
curated dataset (see dataset card for more).
The model can be used in information retrieval (semantic similarity search) and document ranking based on a given query.
More on model and dataset: https://www.linkedin.com/posts/akhooli_arabic-mmarco-sample-dataset-and-colbert-activity-7225135682044743680-35nN