Instructions to use danush99/Model_LightOnOCR-Sin-Handwritten-Text with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use danush99/Model_LightOnOCR-Sin-Handwritten-Text with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("lightonai/LightOnOCR-2-1B") model = PeftModel.from_pretrained(base_model, "danush99/Model_LightOnOCR-Sin-Handwritten-Text") - Notebooks
- Google Colab
- Kaggle
File size: 637 Bytes
3609b75 | 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 | {
"image_processor": {
"data_format": "channels_first",
"default_to_square": true,
"do_convert_rgb": true,
"do_normalize": true,
"do_rescale": true,
"do_resize": true,
"image_mean": [
0.48145466,
0.4578275,
0.40821073
],
"image_processor_type": "PixtralImageProcessorFast",
"image_std": [
0.26862954,
0.26130258,
0.27577711
],
"patch_size": 14,
"resample": 3,
"rescale_factor": 0.00392156862745098,
"size": {
"longest_edge": 1540
}
},
"patch_size": 14,
"processor_class": "LightOnOcrProcessor",
"spatial_merge_size": 2
}
|