-
When I super-resolution a 1024 * 1024 image into 4096 * 4096, I used 70GB of GPU memory and spent 18 minutes, which seems to contradict the advantages stated in the paper. I wonder if this is normal?
-
It seems there are memory leaks and bad CPU performance.
Maybe this helps us to fix that:
https://github.com/grafana/crossplane-provider-grafana/issues/107
https://github.com/grafana/crossplan…
-
my inference size is 640 x 480, tested in 3090, when i set if_local as False, the pipe time is 1.22s, memory costing is large to 22G. however, when setting if_local to True, the pipe time is 2s, memeo…
-
I am running into an issue with JAX where after successive vmaps one gigantic matrix is created (or at least pre-allocated). In my case I am computing kernel matrices:
```python
import jax
import…
-
**Description**
The [docs](https://docs.nvidia.com/deeplearning/triton-inference-server/user-guide/docs/protocol/extension_shared_memory.html) state that:
`Using shared memory instead of sending t…
-
**Describe the bug**
The app can take nearly 3 GB of memory at full capacity, and this may be problematic for hosting online. We can try packages like profvis to figure out where the app is spending …
-
This is related to #4591; when forcing `spv::ZeroInitializeWorkgroupMemoryMode::Polyfill` in `device_from_raw()`, we observe very slow (but correct!) behavior for zeroing the workgroup shared array - …
-
Are the [benchmarks](https://github.com/yandex/YaFSDP?tab=readme-ov-file#benchmarks) conducted against FSDP or FSDP2?
see [speed/memory differences](https://github.com/pytorch/torchtitan/pull/165)
-
In the Readme it says: "The module eeprom must be loaded to display info about your currently installed memory. Load with modprobe eeprom and refresh the module screen."
-
Presumably Nevegrad performs excellent on computationally expensive objective functions, because it is good at choosing an informative next iterative. On the other hand, it is sometimes slow for choos…