From c92e815f4b891b21f43929ad337fe8db2c3cce0d Mon Sep 17 00:00:00 2001 From: Michael Wyatt Date: Thu, 20 Jul 2023 14:07:31 -0700 Subject: [PATCH 1/4] specify triton 2.0.0 requirement --- .github/workflows/nv-inference.yml | 2 +- requirements/requirements-dev.txt | 1 - requirements/requirements-inf.txt | 1 - requirements/requirements-triton.txt | 2 +- 4 files changed, 2 insertions(+), 4 deletions(-) diff --git a/.github/workflows/nv-inference.yml b/.github/workflows/nv-inference.yml index 61f85d8d0598..065f8b93f1e0 100644 --- a/.github/workflows/nv-inference.yml +++ b/.github/workflows/nv-inference.yml @@ -39,7 +39,7 @@ jobs: - name: Install deepspeed run: | - pip install .[dev,1bit,autotuning,inf] + pip install .[dev,1bit,autotuning,inf,triton] ds_report - name: Python environment diff --git a/requirements/requirements-dev.txt b/requirements/requirements-dev.txt index 0106e3785459..0f0cb337fc51 100644 --- a/requirements/requirements-dev.txt +++ b/requirements/requirements-dev.txt @@ -14,5 +14,4 @@ sphinx-rtd-theme tensorboard torchvision transformers -triton wandb diff --git a/requirements/requirements-inf.txt b/requirements/requirements-inf.txt index ef8bfff774d6..848a7f7a485d 100644 --- a/requirements/requirements-inf.txt +++ b/requirements/requirements-inf.txt @@ -3,4 +3,3 @@ lm-eval==0.3.0 protobuf transformers transformers[sentencepiece] -triton diff --git a/requirements/requirements-triton.txt b/requirements/requirements-triton.txt index a59a965090a6..7255b2d76738 100644 --- a/requirements/requirements-triton.txt +++ b/requirements/requirements-triton.txt @@ -1 +1 @@ -triton +triton>=2.0.0 From d98f1f6c49e609a95a24c9ddc1a981653e4e0306 Mon Sep 17 00:00:00 2001 From: Michael Wyatt Date: Thu, 20 Jul 2023 16:19:11 -0700 Subject: [PATCH 2/4] fix for setup-venv action --- .github/workflows/setup-venv/action.yml | 1 + 1 file changed, 1 insertion(+) diff --git a/.github/workflows/setup-venv/action.yml b/.github/workflows/setup-venv/action.yml index cfffd0abeafe..375fdf70f1ce 100644 --- a/.github/workflows/setup-venv/action.yml +++ b/.github/workflows/setup-venv/action.yml @@ -12,6 +12,7 @@ runs: shell: bash - id: create-venv run: | + rm -rf ./unit-test-venv python -m venv unit-test-venv source ./unit-test-venv/bin/activate python -m pip install --upgrade pip From a0958a6f937e8889c233f4ee7401ce9270f38361 Mon Sep 17 00:00:00 2001 From: Michael Wyatt Date: Thu, 20 Jul 2023 16:35:11 -0700 Subject: [PATCH 3/4] fix for install error --- .github/workflows/nv-transformers-v100.yml | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/.github/workflows/nv-transformers-v100.yml b/.github/workflows/nv-transformers-v100.yml index 22a341e8b687..663b9bb98f47 100644 --- a/.github/workflows/nv-transformers-v100.yml +++ b/.github/workflows/nv-transformers-v100.yml @@ -31,6 +31,15 @@ jobs: python -c "import torch; print('torch:', torch.__version__, torch)" python -c "import torch; print('CUDA available:', torch.cuda.is_available())" + - name: Install transformers + run: | + git clone https://github.com/huggingface/transformers + cd transformers + # if needed switch to the last known good SHA until transformers@master is fixed + # git checkout 1cc453d33 + git rev-parse --short HEAD + pip install . + - name: Install deepspeed run: | pip install .[dev,autotuning] @@ -43,12 +52,7 @@ jobs: - name: HF transformers tests run: | unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch - git clone https://github.com/huggingface/transformers cd transformers - # if needed switch to the last known good SHA until transformers@master is fixed - #git checkout 6268694e2 - git rev-parse --short HEAD - # scipy/sklearn required for tests, using the 'dev' extra forces torch re-install pip install .[testing] # find reqs used in ds integration tests find examples/pytorch -regextype posix-egrep -regex '.*(language-modeling|question-answering|summarization|image-classification|text-classification|translation).*/requirements.txt' -exec grep -v 'torch' {} \; | xargs -I {} pip install --upgrade {} From 07eb4acbf28107009ef8e99002fcdacd139bf045 Mon Sep 17 00:00:00 2001 From: Michael Wyatt Date: Fri, 21 Jul 2023 10:04:45 -0700 Subject: [PATCH 4/4] fix torch install error --- .github/workflows/nv-transformers-v100.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/nv-transformers-v100.yml b/.github/workflows/nv-transformers-v100.yml index 663b9bb98f47..a8f3e216ec48 100644 --- a/.github/workflows/nv-transformers-v100.yml +++ b/.github/workflows/nv-transformers-v100.yml @@ -27,7 +27,7 @@ jobs: - name: Install pytorch run: | # use the same pytorch version as transformers CI - pip install -U --cache-dir $TORCH_CACHE torch torchvision torchaudio -f https://download.pytorch.org/whl/torch_stable.html + pip install -U --cache-dir $TORCH_CACHE torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu116 python -c "import torch; print('torch:', torch.__version__, torch)" python -c "import torch; print('CUDA available:', torch.cuda.is_available())"