Merge branch 'master' into tohtana/support_autocast

.github/workflows/cpu-torch-latest.yml

@@ -42,7 +42,7 @@ jobs:
           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 6c3f168b3
+          git checkout 981c276
           git rev-parse --short HEAD
           pip install .
 

.github/workflows/hpu-gaudi2.yml

@@ -112,7 +112,7 @@ jobs:
           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 6c3f168b3
+          git checkout 981c276
           git rev-parse --short HEAD
           pip install .
 

.github/workflows/no-torch.yml

@@ -32,11 +32,12 @@ jobs:
         run: |
           pip uninstall torch --yes
           pip install setuptools
+          pip install build
           pip list
 
       - name: Build deepspeed
         run: |
-          DS_BUILD_STRING=" " python setup.py sdist
+          DS_BUILD_STRING=" " python -m build --sdist
 
       - name: Open GitHub issue if nightly CI fails
         if: ${{ failure() && (github.event_name == 'schedule') }}

.github/workflows/nv-a6000.yml

@@ -42,7 +42,8 @@ jobs:
         run: |
           git clone https://github.com/huggingface/transformers
           cd transformers
-          git checkout v4.47.1
+          # if you need to use an older transformers version temporarily in case of breakage
+          git checkout 981c276
           git rev-parse --short HEAD
           python -m pip install .
       - name: Install deepspeed

.github/workflows/nv-accelerate-v100.yml

@@ -19,7 +19,7 @@ concurrency:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -29,7 +29,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu121
+          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 

.github/workflows/nv-ds-chat.yml

@@ -27,7 +27,7 @@ permissions:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -37,13 +37,13 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu121
+          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 
       - name: Install deepspeed
         run: |
-          pip install transformers
+          pip install transformers==4.48.3
           pip install .[dev]
           ds_report
 

.github/workflows/nv-inference.yml

@@ -22,7 +22,7 @@ concurrency:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -32,7 +32,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install -U --cache-dir $TORCH_CACHE torch==2.1.2 torchvision==0.16.2 --index-url https://download.pytorch.org/whl/cu121
+          pip install -U --cache-dir $TORCH_CACHE torch==2.1.2 torchvision==0.16.2 --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 
@@ -58,8 +58,8 @@ jobs:
         run: |
           unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch
           cd tests
-          #pytest $PYTEST_OPTS -m 'seq_inference' unit/ --torch_ver="2.1" --cuda_ver="12.1"
-          pytest $PYTEST_OPTS -m 'inference_ops' unit/ --torch_ver="2.1" --cuda_ver="12.1"
-          pytest $PYTEST_OPTS --forked -n 4 -m 'inference' unit/ --torch_ver="2.1" --cuda_ver="12.1"
+          #pytest $PYTEST_OPTS -m 'seq_inference' unit/ --torch_ver="2.1" --cuda_ver="12.4"
+          pytest $PYTEST_OPTS -m 'inference_ops' unit/ --torch_ver="2.1" --cuda_ver="12.4"
+          pytest $PYTEST_OPTS --forked -n 4 -m 'inference' unit/ --torch_ver="2.1" --cuda_ver="12.4"
           # run ds_report again to check updated op list
           ds_report

.github/workflows/nv-lightning-v100.yml

@@ -19,7 +19,7 @@ concurrency:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -29,7 +29,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu121
+          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 

.github/workflows/nv-mii.yml

@@ -27,7 +27,7 @@ concurrency:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -37,7 +37,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip3 install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu121
+          pip3 install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 

.github/workflows/nv-nightly.yml

@@ -18,7 +18,7 @@ permissions:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -28,7 +28,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu121
+          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 
@@ -58,7 +58,7 @@ jobs:
         run: |
           unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch
           cd tests
-          pytest $PYTEST_OPTS --forked -m 'nightly' unit/ --torch_ver="2.5" --cuda_ver="12.1"
+          pytest $PYTEST_OPTS --forked -m 'nightly' unit/ --torch_ver="2.6" --cuda_ver="12.4"
 
       - name: Open GitHub issue if nightly CI fails
         if: ${{ failure() && (github.event_name == 'schedule') }}

.github/workflows/nv-torch-latest-v100.yml

@@ -19,7 +19,7 @@ concurrency:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -29,7 +29,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu121
+          pip install -U --cache-dir $TORCH_CACHE torch torchvision --index-url https://download.pytorch.org/whl/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 
@@ -38,7 +38,7 @@ jobs:
           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 6c3f168b3
+          git checkout 981c276
           git rev-parse --short HEAD
           pip install .
 
@@ -55,5 +55,5 @@ jobs:
         run: |
           unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch
           cd tests
-          pytest $PYTEST_OPTS --forked -n 4 unit/ --torch_ver="2.5" --cuda_ver="12.1"
-          pytest $PYTEST_OPTS --forked -m 'sequential' unit/ --torch_ver="2.5" --cuda_ver="12.1"
+          pytest $PYTEST_OPTS --forked -n 8 unit/ --torch_ver="2.6" --cuda_ver="12.4"
+          pytest $PYTEST_OPTS --forked -m 'sequential' unit/ --torch_ver="2.6" --cuda_ver="12.4"

.github/workflows/nv-torch-nightly-v100.yml

@@ -2,11 +2,11 @@ name: nv-torch-nightly-v100
 
 on:
   workflow_dispatch:
-    pull_request:
-    paths:
-      - '.github/workflows/nv-torch-nightly-v100.yml'
   schedule:
     - cron: "0 0 * * *"
+  pull_request:
+    paths:
+      - '.github/workflows/nv-torch-nightly-v100.yml'
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
@@ -18,7 +18,7 @@ permissions:
 
 jobs:
   unit-tests:
-    runs-on: [self-hosted, nvidia, cu121, v100]
+    runs-on: [self-hosted, nvidia, cu124, v100]
 
     steps:
       - uses: actions/checkout@v4
@@ -28,7 +28,7 @@ jobs:
 
       - name: Install pytorch
         run: |
-          pip install --pre torch torchvision --index-url https://download.pytorch.org/whl/nightly/cu121
+          pip install --pre torch torchvision --index-url https://download.pytorch.org/whl/nightly/cu124
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 
@@ -37,7 +37,7 @@ jobs:
           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 6c3f168b3
+          git checkout 981c276
           git rev-parse --short HEAD
           pip install .
 
@@ -54,7 +54,7 @@ jobs:
         run: |
           unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch
           cd tests
-          pytest $PYTEST_OPTS --forked -n 4 unit/
+          pytest $PYTEST_OPTS --forked -n 8 unit/
           pytest $PYTEST_OPTS --forked -m 'sequential' unit/
 
       - name: Open GitHub issue if nightly CI fails