Multithreaded CI builds #3311
Multithreaded CI builds #3311
Conversation
|
Is there any reason we shouldn't use |
I second this, since adding self hosted runners on forks for testing PRs is trivial, detecting local CPU count can be very useful. While we're at it, GitHub appears to automatically prioritize self hosted runners, and the only thing you need to do is to give them the same tag As soon as you register a self hosted runner in your forked repo, CI jobs will run on it automatically. |
Done, and added for Windows too. That's a much better way of doing it. |
.github/workflows/build.yml
Outdated
| @@ -472,7 +472,7 @@ jobs: | |||
| run: | | |||
| sudo pkg update | |||
| sudo pkg install -y gmake automake autoconf pkgconf llvm15 clinfo clover opencl clblast openblas | |||
| gmake CC=/usr/local/bin/clang15 CXX=/usr/local/bin/clang++15 | |||
| gmake CC=/usr/local/bin/clang15 CXX=/usr/local/bin/clang++15 -j `sysctl -n hw.ncpu` | |||
There was a problem hiding this comment.
any reason you switched syntax for this one?
There was a problem hiding this comment.
The FreeBSD shell doesn't support the $() syntax.
There was a problem hiding this comment.
I see. Should we switch all the others then?
This is probably the difference between sh and bash
There was a problem hiding this comment.
This is probably the difference between
shandbash
Nope. Modern POSIX sh supports $(), you can try it out for yourself with the dash shell. FreeBSD defaults to tcsh which doesn't support $().
…example * 'master' of github.com:ggerganov/llama.cpp: ggml-cuda : perform cublas mat mul of quantized types as f16 (ggml-org#3412) llama.cpp : add documentation about rope_freq_base and scale values (ggml-org#3401) train : fix KQ_pos allocation (ggml-org#3392) llama : quantize up to 31% faster on Linux and Windows with mmap (ggml-org#3206) readme : update hot topics + model links (ggml-org#3399) readme : add link to grammars app (ggml-org#3388) swift : fix build on xcode 15 (ggml-org#3387) build : enable more non-default compiler warnings (ggml-org#3200) ggml_tensor: update the structure comments. (ggml-org#3283) ggml : release the requested thread pool resource (ggml-org#3292) llama.cpp : split llama_context_params into model and context params (ggml-org#3301) ci : multithreaded builds (ggml-org#3311) train : finetune LORA (ggml-org#2632) gguf : basic type checking in gguf_get_* (ggml-org#3346) gguf : make token scores and types optional (ggml-org#3347) ci : disable freeBSD builds due to lack of VMs (ggml-org#3381) llama : custom attention mask + parallel decoding + no context swaps (ggml-org#3228) docs : mark code as Bash (ggml-org#3375) readme : add Mistral AI release 0.1 (ggml-org#3362) ggml-cuda : perform cublas fp16 matrix multiplication as fp16 (ggml-org#3370)
* mac and linux threads * windows * Update build.yml * Update build.yml * Update build.yml * automatically get thread count * windows syntax * try to fix freebsd * Update build.yml * Update build.yml * Update build.yml
The CUDA install is still super slow, but if you have the threads you might as well use them! This halves build time for some targets.