-
When I train the model of dtype fp16 or mixed precision on V100 sxm2 16G, loss.backward() return error:
python: /project/lib/Analysis/Allocation.cpp:40: std::pair mlir::triton::getCvtOrder(mlir::A…
-
This is a continuation of our discussion about precision input of device parameters such as integration time. Post closing of #23.
We are targeting the full range of (1, 2000ms) for integration tim…
-
TLDR: gps-sdr-sim -> HackRF -> ublox GPS: drifts for dozens of meters in static and also dynamic mode, sometimes looses fix
## Setup
I tried multiple configurations. But always similar scenario:…
-
I tested LyCORIS/LoCon preset for SDXL and it works fine, but if the box "DoRA Weight Decompose" is checked (with or without extra algorithms) then it will not continue. I tried default preset with 'G…
-
Implement analysis to set the flag if command-line option -enable-16bit-types is specified and Shader Model is 6.2 or later
NOTE: If this is set, min-precision mode (UseMinPrecision) is not set.
-
hello,
I'm trying to test the model based on the ReadMe file you posted, but I'm getting the following error.
"precision": torchmetrics.Precision(num_classes=2, average=None).to(device),
Ty…
-
```python
#-- coding:utf8 --
import MNN
import MNN.numpy as np
import MNN.cv as cv2
def inference(model, img, precision, backend, thread):
config = {}
config['precision'] = precision
…
-
Adding the atomic baseline should be made optional, as it can lead to loss of precision in large systems because of the finite range of float32
-
We transferred the model to onnx and found that the accuracy was not consistent with the original model
-
### Search before asking
- [X] I have searched the YOLOv8 [issues](https://github.com/ultralytics/ultralytics/issues) and [discussions](https://github.com/ultralytics/ultralytics/discussions) and f…