Visual Document Retrieval
PEFT
Safetensors
vidore
multimodal_embedding
multilingual_embedding
Text-to-Visual Document (T→VD) retrieval
Instructions to use Metric-AI/ColQwenStella-2b-multilingual with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Metric-AI/ColQwenStella-2b-multilingual with PEFT:
from peft import PeftModel from transformers import AutoModel base_model = AutoModel.from_pretrained("Metric-AI/ColQwenStella-base-2b") model = PeftModel.from_pretrained(base_model, "Metric-AI/ColQwenStella-2b-multilingual") - Notebooks
- Google Colab
- Kaggle
File size: 670 Bytes
a19ef03 57b3dd5 a19ef03 f96bf68 a55cbd2 a19ef03 | 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 33 | {
"do_convert_rgb": true,
"do_normalize": true,
"do_rescale": true,
"do_resize": true,
"image_mean": [
0.48145466,
0.4578275,
0.40821073
],
"image_processor_type": "Qwen2VLImageProcessor",
"image_std": [
0.26862954,
0.26130258,
0.27577711
],
"max_pixels": 12845056,
"merge_size": 2,
"min_pixels": 3136,
"patch_size": 14,
"processor_class": "ColQwenStellaProcessor",
"resample": 3,
"rescale_factor": 0.00392156862745098,
"size": {
"longest_edge": 12845056,
"shortest_edge": 3136
},
"temporal_patch_size": 2,
"auto_map": {
"AutoProcessor": "processing_colqwenstella.ColQwenStellaProcessor"
}
}
|