-
## 🐛 Bug
When I deploy my own 2B model using MLC on Android, the model interface initializes successfully and displays the "Ready to chat" prompt after opening. However, the app crashes after sendi…
-
hello I have cupy cuda array and I want to pass it into julia as is.
CUDA arrays are just list of pointers so it should be possible
from CUDA.jl side I know is possible as I have a comment
"""
fo…
-
### 请提出你的问题 Please ask your question
当使用cupy和paddle相结合是，在每个batch中，我都会将cupy的数据通过from_dlpack(a.toDlpack())转换，其中a是cupy.ndarray。
但是这样做会导致内存直接占用爆炸，即在batch结束后，不会自动的删除未使用的部分（我认为是这样的），请问，如何避免出现这种情况，谢谢.
注…
-
JAX has a lot of powerful building blocks for vectorizing and parallelizing code, but there are quite a few different APIs and using them well takes skill.
It would be nice to have a comprehensive …
-
For doing computations on tensors with `pycuda`, I can do this
for going from `Tensor` to `pycuda.gpuarray.GPUArray`:
```
def tensor_to_gpuarray(tensor):
if not tensor.is_cuda:
raise…
-
The motivating example is the NumPy API that is exposed to other Python extensions such that they can work with arrays natively/directly without a round-trip through Python code/abstractions.
How t…
-
Current fails for:
Calls `internal_new_from_data`: (#65)
- [x] `__getitem__`
- [x] `__rpow__` (straight up calls torch.tensor)
- [x] torch.tensor
- [x] Tensor.new()
[Data pointer accessed by…
-
### Description
When trying to build from source on Windows 11 with CUDA 11.8 the build failed as the symbolic link to the third party dlpack library folder did not work as intended.
The releva…
-
Note: Providing complete information in the most concise form is the best way to get help. This issue template serves as the checklist for essential information to most of the technical issues and bug…
-
## 🐛 Bug
When I compile Phi-2 (https://huggingface.co/microsoft/phi-2) with `tvm.relax.transform.FuseOps()` and `tvm.relax.transform.FuseTIR()` transformations commented out (https://github.com/mlc-a…