-
## 🐛 Bug
ValueError: The block no longer exists in the IRModule
Stack trace not available when DMLC_LOG_STACK_TRACE is disabled at compile time.
## To Reproduce
Steps to reproduce the be…
-
**Is your feature request related to a problem? Please describe.**
When using cudashm, tritonclient only support reading/writing data using numpy. Is there a way to utilize pycuda pagelocked memory o…
-
**Description**
i use model.onnx transform from yolov5s.pt(v6.1) ， deploy in triton server,
as the model name yolo, and add nms in ensemble
tritonclient.utils.InferenceServerException: [400] in e…
-
## 🐛 Bug
Using @junrushao 's https://github.com/mlc-ai/mlc-llm/issues/1530#issuecomment-1874966277, but with additional patches to enable opencl, the generation quality for TinyLlama is surprisingl…
-
### Checklist
- [X] I added a descriptive title
- [X] I searched open reports and couldn't find a duplicate
### What happened?
When I run:
conda install -c xformers xformers
I have follo…
-
I notice in pytorch.py the functionality to convert between pytorch and jax arrays goes via a conversion to the cpu. It is possible to do this without moving the data at all.
```
def j2t(x_jax, de…
-
### 问题描述 Issue Description
```
[478/2763] Building CUDA object paddle\phi\kernels\CMakeFiles\cross_entropy_kernel.dir\gpu\cross_entropy_kernel.cu.obj
FAILED: paddle/phi/kernels/CMakeFiles/cross_ent…
-
**Describe the bug**
Only when we have `timedelta64[s]` dtype for a column, the parquet writer seems to be writing it as a `timedelta64[ms]` column which is leading both cudf & pyarrow parquet reader…
-
**Describe the bug**
While trying to create an Index object of `list` dtype, we are currently getting a `RecursionError`. Having `list` dtype support in `Index` would be essential incase of `.str.spl…
-
### Description
Running (at least some) cupyx tests fails on a machine with cusparseLt 0.5.2 installed (with https://github.com/cupy/cupy/pull/8074) and passes on a different machine which does not…
ev-br updated
8 months ago