-
Hello, after I finished training t1, I then wanted to train t2, and found an error
```python
[06/20 09:19:56 fvcore.common.checkpoint]: Loading trainer from output/M-OWODB/model_final.pth ...
[06…
-
### KVCache usage
it seems like the KVCache is only allocated on the first device in CUDA_VISIBLE_DEVICES, even if using `tensor-parallel-size` > 1. Is there any plan to support full KVCache alloca…
-
The T4.BuildTools package seems to compile t4 files against a different version of .net depending on if the build is run from visual studio vs dotnet build.
My Project file looks like:
```
…
-
nvfuser commit: d75fc93
cuda: 12.5
```python
import torch
from nvfuser import FusionDefinition, DataType
def nvfuser_fusion_id1(fd : FusionDefinition) -> None :
T0 = fd.define_tensor(sha…
-
In the "Known Issues" of the release note of TensorRT 10.0, it is mentioned that "UNets with tensors containing >2^31 elements may fail during the engine building step." This problem brings great rest…
-
### Bug description
Upon enabling the `relationJoins` feature, we noticed significant performance degradation when join on our higher-volume tables.
### How to reproduce
With a query of:
…
-
I'm running a single T4 node on GKE. Nodes are properly labeled as shown below and `sky show-gpus --cloud kubernetes` is also correct but fails to launch.
```bash
(sky) gcpuser@gfd-ebd1-head-…
-
Hi! I met a question when I generated the engine for yolov7 in the GPU Tesla T4. The environment I configured is DeepStream 6.0, TensorRT 8.4.2, CUDA 11.4, and a version of cuDNN compatible with CUDA …
-
[`TextTemplatePreprocessor `](https://github.com/ufcpp/TextTemplateSourceGenerator/blob/main/src/TextTemplateSourceGenerator/TextTemplatePreprocessor.cs) is something like TextTemplatingFile**Preproce…
-
WORK-IN-PROGRESS: At the moment where are not all engines in this repo. Feel free to leave me a message / feedback or any hints in this issue. More engines will follow.
### HINT:
If the POGO Pin …