-
When I try to compile the test suite on Debain 12 with gcc (Debian 12.2.0-14) 12.2.0 I get the following error
```
CXX=g++ make
gnu++23: test extensions (-Wall -Wextra -Wpedantic -Wno-attributes …
-
When training the model, both lightning and huggingface both try to parelize the code and a warning pops up.
-
At the moment the number of parallel workers in meta-modelling is limited to 10. Mainly to not overload the system.
This ticket exists to investigate what are the bottlenecks and investigate in how fa…
-
In the 2022 branch, nested parallellism is done through `OMP_NESTED`. This is deprecated and maybe even was in 2022. Use `OMP_ACTIVE_LEVELS` insteadl.
-
Documentation on how to best exploit parallelism in CASTEP. Including:
- Available parallelism (g-vector, k-point, band, farm, etc.)
- Meaning of key terms
- Use cases for each
- OMP and MPI…
-
torchxla spmd whether expert parallelism is supported?
If it is a moe model, how should it be computed in xla?
## ❓ Questions and Help
-
Hi,
I was running Flan-t5 XXL with ctranslate2 and observed completely different results when run with tensor parallelism.
**To convert from HF to CT2:**
```bash
ct2-transformers-converter -…
-
### Feature request
Being able to split models into multiple GPUs, as with vllm/aphrodite engine for LLMs.
### Motivation
It would be extremely helpful to be able to split larger models into multip…
-
There are 2 separate problems with this plugin regarding parallelism:
1. This plugin assumes that it is the ONLY instance running on a build machine, and doesn't work for teams using multiple runne…
-
### OpenTofu Version
```shell
1.6.2
```
### Use Cases
Callling API limited providers in the same module/run.
I.e. mixing grafana and aws or status page.
### Attempted Solutions
Using terragrunt…