-
### What happened?
When I try to compile the sharded Llama 3.1 8b fp16 [IR](https://gist.github.com/aviator19941/bab5886f53f2fd0b3b8458519148542c) for CPU or GPU:
I get this error for CPU:
https://g…
-
when creating a memory allocation with `MemoryPointer::new()` , the call to `sys::hipMalloc` should allocate on the default device.
The observation is that even if `hipGetDevice()` returns `0`, the m…
-
When using the Low Pass filter `TabOpLowPassFilter` the time column is also filtered/shifted.
A simple reproduction is available [here](https://github.com/gateway240/opensim-core-examples) in the `…
-
### What is the issue?
I have 8 AMD 7900XTX cards in llama.cpp to limit access to certain GPUs, I use the HIP_VISIBLE_DEVICES command and it works correctly. However, if I want to limit GPU access fo…
-
### Search before asking
- [X] I have searched the HUB [issues](https://github.com/ultralytics/hub/issues) and [discussions](https://github.com/ultralytics/hub/discussions) and found no similar quest…
-
### Problem Description
as `device_functions.h` is located at `/opt/rocm/include/hip` , so expecting to use it as:
```c++
#include
#include
// Buggy: somehow setuptool will add prefix to
…
-
Following the build instructions in the readme,
```
cmake .. -G "Ninja" -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DSD_HIPBLAS=ON -DCMAKE_BUILD_TYPE=Release -DAMDGPU_TARGETS=gfx1100
…
-
### Problem Description
Hi,
As seen in https://godbolt.org/z/ehxeqG4sj, `clang --stdlib=libc++` works with clang-17 and does not work with newer versions (checked with clang version 19.1.0 and …
-
Similar with https://github.com/llvm/llvm-project/issues/72261
```cpp
struct A {
A(){}
~A() {}
int x;
};
#define __device__ __attribute__((device))
#define __shared__ __attribute__((…
-
### Problem
Currently no HIP exists to define Block streams
### Solution
Create Block Stream HIP
- [x] Draft HIP
- [x] Early socialization w DAs
- [x] Early socialization w PMs
- [x] Early socializ…