Accelerating smooth inferences with transformers
AI & ML interests
Fix issues and experiment with new things.
Recent Activity
Organization Card
Stranger Vision — Here, we modify, repair, and experiment with things.
As this is an early stage of the community, we have made some marginal fixes to ensure certain models work with the latest updates. We fixed the issue in Dots.OCR inference caused by a NoneType argument in the latest Transformers versions, resolved the problem with num_hidden_layers in Nanonets-1.5B, and exposed the path for PaddleOCR to enable smoother Transformers inference. Additionally, we fixed the compatibility issues in DeepSeek OCR with the latest Transformers version, ensuring that inference now runs smoothly without any errors.
Accelerating smooth inferences with transformers
-
strangervisionhf/dots.ocr-base-fix
Image-Text-to-Text • 3B • Updated • 1 -
strangervisionhf/deepseek-ocr-latest-transformers
Image-Text-to-Text • 3B • Updated • 26 • 3 -
strangervisionhf/paddle.ocr_path_expose
Image-Text-to-Text • 1.0B • Updated • 10 • 2 -
strangervisionhf/excess_layer_pruned-nanonets-1.5b
Image-Text-to-Text • 2B • Updated • 1
Accelerating smooth inferences with transformers
Accelerating smooth inferences with transformers
-
strangervisionhf/dots.ocr-base-fix
Image-Text-to-Text • 3B • Updated • 1 -
strangervisionhf/deepseek-ocr-latest-transformers
Image-Text-to-Text • 3B • Updated • 26 • 3 -
strangervisionhf/paddle.ocr_path_expose
Image-Text-to-Text • 1.0B • Updated • 10 • 2 -
strangervisionhf/excess_layer_pruned-nanonets-1.5b
Image-Text-to-Text • 2B • Updated • 1
models
4
strangervisionhf/deepseek-ocr-latest-transformers
Image-Text-to-Text
•
3B
•
Updated
•
26
•
3
strangervisionhf/dots.ocr-base-fix
Image-Text-to-Text
•
3B
•
Updated
•
1
strangervisionhf/excess_layer_pruned-nanonets-1.5b
Image-Text-to-Text
•
2B
•
Updated
•
1
strangervisionhf/paddle.ocr_path_expose
Image-Text-to-Text
•
1.0B
•
Updated
•
10
•
2