-
### Describe the issue
During cuda graph catpure, ORT will trigger cudaStreamSynchronize, which is not allowed in CUDA graph catpure. Call stack is like the following:
```
libonnxruntime_providers_…
-
2024-08-24 13:31:04,052 - modelscope - INFO - PyTorch version 2.0.1+cu118 Found.
2024-08-24 13:31:04,052 - modelscope - INFO - Loading ast index from C:\Users\PC\.cache\modelscope\ast_indexer
2024-0…
-
I am trying to compile the official ONNX version of Efficientnet_lite4 (from [here](https://github.com/onnx/models/tree/main/vision/classification/efficientnet-lite4)) to run on the Jacinto7 but the c…
-
### System Info
tranformers: v3.0.0-alpha.15
node: v18.20.3
MacOS M1 Sonoma 14.5
### Environment/Platform
- [ ] Website/web-app
- [ ] Browser extension
- [X] Server-side (e.g., Node.js,…
-
在同样的模型加载后进行推理,GPU会比CPU慢很多,同一张图片循环遍历GPU的推理速度会提升,因为有缓存,但是不同的图片进行识别,GPU的推理速度明显慢,而且GPU的使用率已经到了100%
python3.7 使用flask提供接口
启动时先加载模型
model = ONNXPaddleOcr(use_angle_cls=True, use_gpu=True)
-
Training a model based on my experience it's just so terrible,
1. Quantization does not work
2. Running inference using onnxruntime does not work
3. Converting .onnx model to .tflite does not wor…
-
pip install -r requirements.txt
文字:
ERROR: Ignored the following versions that require a different python version: 1.14.0rc1 Requires-Python >=3.10; 1.14.0rc2 Requires-Python >=3.10
ERROR…
-
### Describe the issue
I tried building at least three version `v1.13.1`, `1.14.1` and `1.15.0` of onnxruntime, and non of them can compile with the ACL execution provider in version 20.02.
### Urge…
-
def _preprocess_images(images_folder: str, height: int, width: int, size_limit=0):
"""
Loads a batch of images and preprocess them
parameter images_folder: path to folder storing images…
-
### System Info
- `transformers` version: 4.36.2
- Platform: Windows-10-10.0.22621-SP0
- Python version: 3.10.13
- Huggingface_hub version: 0.20.1
- Safetensors version: 0.4.1
- Accelerate ver…