default mg

fix
2025-09-25 16:30:23 -04:00 · 2025-09-25 16:08:35 -04:00 · 2025-09-25 15:55:08 -04:00 · 2025-09-25 15:47:23 -04:00 · 2025-09-25 14:54:03 -04:00 · 2025-09-25 14:27:34 -04:00
97 changed files with 4222 additions and 9140 deletions
--- a/.coveragerc
+++ b/.coveragerc
@@ -2,6 +2,7 @@
 source = axolotl
 omit =
    */tests/*
+    setup.py

 [report]
 exclude_lines =
--- a/.github/CONTRIBUTING.md
+++ b/.github/CONTRIBUTING.md
@@ -29,18 +29,13 @@ PRs are **greatly welcome**!
 2. Set up the development environment by following the instructions in the [README.md](https://github.com/axolotl-ai-cloud/axolotl/tree/main/README.md) file.
 3. Explore the codebase, run tests, and verify that everything works as expected.

-Please run the below to setup:
-
+Please run below to setup env
 ```bash
-git clone https://github.com/axolotl-ai-cloud/axolotl.git
-cd axolotl
+pip3 install -r requirements-dev.txt -r requirements-tests.txt
+pre-commit install

-uv sync --dev && uv pip install flash-attn --no-build-isolation
-source .venv/bin/activate
-
-pre-commit install  # install pre-commit hooks
-
-pytest tests/  # optional; run test suite
+# test
+pytest tests/
 ```

 ## How to Contribute
--- a/.github/workflows/base.yml
+++ b/.github/workflows/base.yml
@@ -39,6 +39,13 @@ jobs:
            pytorch: 2.6.0
            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
            dockerfile: "Dockerfile-base"
+          - cuda: "126"
+            cuda_version: 12.6.3
+            cudnn_version: ""
+            python_version: "3.11"
+            pytorch: 2.7.0
+            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
+            dockerfile: "Dockerfile-base"
          - cuda: "126"
            cuda_version: 12.6.3
            cudnn_version: ""
@@ -98,9 +105,7 @@ jobs:
          context: .
          file: ./docker/${{ matrix.dockerfile }}
          push: ${{ github.event_name != 'pull_request' }}
-          tags: |
-            ${{ steps.metadata.outputs.tags }}-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
-            ${{ steps.metadata.outputs.tags }}-base-uv-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
+          tags: ${{ steps.metadata.outputs.tags }}-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
          labels: ${{ steps.metadata.outputs.labels }}
          build-args: |
            CUDA_VERSION=${{ matrix.cuda_version }}
--- a/.github/workflows/docs.yml
+++ b/.github/workflows/docs.yml
@@ -20,14 +20,10 @@ jobs:
          uses: actions/setup-python@v5
          with:
            python-version: '3.11'
-        - name: Install uv
-          uses: astral-sh/setup-uv@v4
-          with:
-            version: "latest"
        - name: Install dependencies
          run: |
-            uv pip install --system jupyter quartodoc
-            uv pip install --system -e .
+            python3 -m pip install jupyter quartodoc
+            python3 -m pip install -e .
        - name: Build autodoc
          run: quartodoc build
        - name: Publish to GitHub Pages (and render)
--- a/.github/workflows/lint.yml
+++ b/.github/workflows/lint.yml
@@ -6,7 +6,7 @@ on:
      types: [opened, synchronize, reopened, ready_for_review]
      paths:
       - '**.py'
-       - 'pyproject.toml'
+       - 'requirements.txt'
       - '.github/workflows/*.yml'
       - "*.[q]md"
       - "examples/**/*.y[a]?ml"
@@ -23,4 +23,5 @@ jobs:
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
+          cache: 'pip' # caching pip dependencies
      - uses: pre-commit/action@v3.0.1
--- a/.github/workflows/main.yml
+++ b/.github/workflows/main.yml
@@ -20,6 +20,11 @@ jobs:
            python_version: "3.11"
            pytorch: 2.6.0
            axolotl_extras:
+          - cuda: 126
+            cuda_version: 12.6.3
+            python_version: "3.11"
+            pytorch: 2.7.0
+            axolotl_extras:
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
@@ -68,8 +73,6 @@ jobs:
            PYTORCH_VERSION=${{ matrix.pytorch }}
            AXOLOTL_ARGS=${{ matrix.axolotl_args }}
            AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
@@ -90,6 +93,11 @@ jobs:
            python_version: "3.11"
            pytorch: 2.6.0
            axolotl_extras:
+          - cuda: 126
+            cuda_version: 12.6.3
+            python_version: "3.11"
+            pytorch: 2.7.0
+            axolotl_extras:
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
@@ -140,8 +148,6 @@ jobs:
          build-args: |
            BASE_TAG=${{ github.ref_type == 'tag' && 'main' || github.ref_name }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
            CUDA=${{ matrix.cuda }}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile-cloud
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
@@ -207,8 +213,6 @@ jobs:
          build-args: |
            BASE_TAG=${{ github.ref_type == 'tag' && 'main' || github.ref_name }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
            CUDA=${{ matrix.cuda }}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile-cloud-no-tmux
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
--- a/.github/workflows/multi-gpu-e2e.yml
+++ b/.github/workflows/multi-gpu-e2e.yml
@@ -4,6 +4,8 @@ on:
  pull_request:
    paths:
      - 'tests/e2e/multigpu/**.py'
+      - 'requirements.txt'
+      - 'setup.py'
      - 'pyproject.toml'
      - '.github/workflows/multi-gpu-e2e.yml'
      - 'src/axolotl/core/trainers/mixins/sequence_parallel.py'
@@ -54,17 +56,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
          python -m pip install --upgrade pip
-          pip install modal==1.0.2 jinja2 protobuf
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=${{ github.ref_name }}-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
+          echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
          echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
          echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
          echo "AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}" >> $GITHUB_ENV
@@ -74,4 +72,4 @@ jobs:
          echo "CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }}" >> $GITHUB_ENV
      - name: Run tests job on Modal
        run: |
-          modal run -m cicd.multigpu
+          modal run cicd.multigpu
--- a/.github/workflows/nightlies.yml
+++ b/.github/workflows/nightlies.yml
@@ -52,8 +52,6 @@ jobs:
            CUDA=${{ matrix.cuda }}
            PYTORCH_VERSION=${{ matrix.pytorch }}
            AXOLOTL_ARGS=${{ matrix.axolotl_args }}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
@@ -104,8 +102,6 @@ jobs:
          build-args: |
            BASE_TAG=${{ github.ref_name }}-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}${{ matrix.axolotl_extras != '' && '-' || '' }}${{ matrix.axolotl_extras }}
            CUDA=${{ matrix.cuda }}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile-cloud
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
--- a/.github/workflows/precommit-autoupdate.yml
+++ b/.github/workflows/precommit-autoupdate.yml
@@ -18,15 +18,10 @@ jobs:
        with:
          python-version: '3.11'

-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
-
      - name: Update pre-commit hooks
        id: update
        run: |
-          uv pip install --system pre-commit
+          pip install pre-commit
          pre-commit autoupdate
          if [[ -n $(git status --porcelain) ]]; then
            echo "changes=true" >> $GITHUB_OUTPUT
--- a/.github/workflows/preview-docs.yml
+++ b/.github/workflows/preview-docs.yml
@@ -40,15 +40,10 @@ jobs:
        with:
          python-version: '3.11'

-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
-
      - name: Install dependencies
        run: |
-          uv pip install --system jupyter quartodoc
-          uv pip install --system -e .
+          python3 -m pip install jupyter quartodoc
+          python3 -m pip install -e .

      - name: Build autodoc
        run: quartodoc build
--- a/.github/workflows/pypi.yml
+++ b/.github/workflows/pypi.yml
@@ -38,24 +38,23 @@ jobs:
        with:
          python-version: "3.11"

-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
-
      - name: Install dependencies
        run: |
-          uv pip install --system wheel packaging==23.2
-          uv pip install --system --no-build-isolation -e ".[dev]"
+          pip3 install wheel packaging==23.2
+          pip3 install --no-build-isolation -e .
+          pip3 install -r requirements-dev.txt -r requirements-tests.txt

      - name: Extract tag name
        id: tag
-        run: echo "TAG_NAME=$(echo "$GITHUB_REF" | cut -d / -f 3)" >> "$GITHUB_OUTPUT"
+        run: echo ::set-output name=TAG_NAME::$(echo $GITHUB_REF | cut -d / -f 3)

-      - name: Build package
+      - name: Update version in setup.py
        run: |
-          uv pip install --system build
-          python -m build
+          sed -i -E 's/version="([0-9.]+)",/version="${{ steps.tag.outputs.TAG_NAME }}",/g' setup.py
+
+      - name: Build a source dist
+        run: |
+          python setup.py sdist

      - name: Publish package distributions to PyPI
        uses: pypa/gh-action-pypi-publish@release/v1
--- a/.github/workflows/tests-nightly.yml
+++ b/.github/workflows/tests-nightly.yml
@@ -13,6 +13,7 @@ jobs:
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
+          cache: 'pip' # caching pip dependencies
      - uses: pre-commit/action@v3.0.1
        env:
          SKIP: no-commit-to-branch
@@ -42,30 +43,32 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: ${{ matrix.python_version }}
+          cache: 'pip' # caching pip dependencies

-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
+      - name: upgrade pip
+        run: |
+          pip3 install --upgrade pip
+          pip3 install --upgrade packaging==23.2 setuptools==75.8.0 wheel

      - name: Install PyTorch
        run: |
-          uv pip install --system torch==${{ matrix.pytorch_version }} torchvision
+          pip3 install torch==${{ matrix.pytorch_version }} torchvision

-      - name: Update pyproject.toml for nightly builds
+      - name: Update requirements.txt
        run: |
-          sed -i 's#"transformers==.*"#"transformers @ git+https://github.com/huggingface/transformers.git@main"#' pyproject.toml
-          sed -i 's#"peft==.*"#"peft @ git+https://github.com/huggingface/peft.git@main"#' pyproject.toml
-          sed -i 's#"accelerate==.*"#"accelerate @ git+https://github.com/huggingface/accelerate.git@main"#' pyproject.toml
-          sed -i 's#"trl==.*"#"trl @ git+https://github.com/huggingface/trl.git@main"#' pyproject.toml
-          sed -i 's#"datasets==.*"#"datasets @ git+https://github.com/huggingface/datasets.git@main"#' pyproject.toml
+          sed -i 's#^transformers.*#transformers @ git+https://github.com/huggingface/transformers.git@main#' requirements.txt
+          sed -i 's#^peft.*#peft @ git+https://github.com/huggingface/peft.git@main#' requirements.txt
+          sed -i 's#^accelerate.*#accelerate @ git+https://github.com/huggingface/accelerate.git@main#' requirements.txt
+          sed -i 's#^trl.*#trl @ git+https://github.com/huggingface/trl.git@main#' requirements.txt
+          sed -i 's#^datasets.*#datasets @ git+https://github.com/huggingface/datasets.git@main#' requirements.txt

      - name: Install dependencies
        run: |
-          uv pip show --system torch
-          uv pip install --system --no-build-isolation -e ".[dev]"
+          pip3 show torch
+          pip3 install --no-build-isolation -U -e .
          python scripts/unsloth_install.py | sh
          python scripts/cutcrossentropy_install.py | sh
+          pip3 install -r requirements-dev.txt -r requirements-tests.txt

      - name: Make sure PyTorch version wasn't clobbered
        run: |
@@ -81,6 +84,9 @@ jobs:
          pytest -v --durations=10 tests/patched/
          pytest -v --durations=10 tests/cli/

+      - name: cleanup pip cache
+        run: |
+          find "$(pip cache dir)/http-v2" -type f -mtime +14 -exec rm {} \;

  docker-e2e-tests:
    if: github.repository_owner == 'axolotl-ai-cloud'
@@ -114,16 +120,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
-          uv pip install --system modal==1.0.2 jinja2
+          python -m pip install --upgrade pip
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=main-base-uv-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
+          echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
          echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
          echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
          echo "AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}" >> $GITHUB_ENV
@@ -133,7 +136,7 @@ jobs:
          echo "CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }}" >> $GITHUB_ENV
      - name: Run tests job on Modal
        run: |
-          modal run -m cicd.e2e_tests
+          modal run cicd.e2e_tests
  docker-e2e-multigpu-tests:
    if: github.repository_owner == 'axolotl-ai-cloud'
    # this job needs to be run on self-hosted GPU runners...
@@ -159,16 +162,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
-          uv pip install --system modal==1.0.2 jinja2
+          python -m pip install --upgrade pip
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=main-base-uv-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
+          echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
          echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
          echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
          echo "AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}" >> $GITHUB_ENV
--- a/.github/workflows/tests.yml
+++ b/.github/workflows/tests.yml
@@ -7,16 +7,18 @@ on:
      - "main"
    paths:
      - '**.py'
-      - 'pyproject.toml'
+      - 'requirements.txt'
      - '.github/workflows/*.yml'
+      - 'requirements-tests.txt'
      - 'cicd/cicd.sh'
      - 'cicd/Dockerfile.jinja'
  pull_request:
      types: [opened, synchronize, reopened, ready_for_review]
      paths:
       - '**.py'
-       - 'pyproject.toml'
+       - 'requirements.txt'
       - '.github/workflows/*.yml'
+       - 'requirements-tests.txt'
       - 'cicd/cicd.sh'
       - 'cicd/Dockerfile.jinja'
  workflow_dispatch:
@@ -39,6 +41,7 @@ jobs:
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
+          cache: 'pip' # caching pip dependencies
      - uses: pre-commit/action@v3.0.1
        env:
          SKIP: no-commit-to-branch
@@ -69,25 +72,24 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: ${{ matrix.python_version }}
+          cache: 'pip' # caching pip dependencies

-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
+      - name: upgrade pip
+        run: |
+          pip3 install --upgrade pip
+          pip3 install --upgrade packaging==23.2 setuptools==75.8.0 wheel

      - name: Install PyTorch
        run: |
-          uv pip install --system torch==${{ matrix.pytorch_version }} torchvision
+          pip3 install torch==${{ matrix.pytorch_version }} torchvision

      - name: Install dependencies
        run: |
-          uv pip show --system torch
-          uv pip install --system wheel
-          printf "torch==${{ matrix.pytorch_version }}\n" > torch-constraints.txt
-          uv pip install --system --no-cache-dir --no-build-isolation -e ".[dev]" --constraints torch-constraints.txt
-          set -o pipefail
-          python scripts/unsloth_install.py | bash
-          python scripts/cutcrossentropy_install.py | bash
+          pip3 show torch
+          pip3 install --no-build-isolation -U -e .
+          python scripts/unsloth_install.py | sh
+          python scripts/cutcrossentropy_install.py | sh
+          pip3 install -r requirements-dev.txt -r requirements-tests.txt

      - name: Make sure PyTorch version wasn't clobbered
        run: |
@@ -103,10 +105,10 @@ jobs:

      - name: Run tests
        run: |
-          python -m pytest -v --durations=10 -n 8 --dist loadfile --cov=axolotl --cov-report=xml --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ --ignore=tests/monkeypatch/ tests/
-          python -m pytest -v --durations=10 -n 8 --cov=axolotl --cov-append --cov-report=xml tests/monkeypatch/
-          python -m pytest -v --durations=10 -n 8 --cov=axolotl --cov-append --cov-report=xml tests/patched/
-          python -m pytest -v --durations=10 -n 8 --cov=axolotl --cov-append --cov-report=xml tests/cli/
+          pytest -v --durations=10 -n8 --dist loadfile --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ --ignore=tests/monkeypatch/ tests/ --cov=axolotl --cov-report=xml
+          pytest -v --durations=10 tests/monkeypatch/ --cov=axolotl --cov-append --cov-report=xml
+          pytest -v --durations=10 tests/patched/ --cov=axolotl --cov-append --cov-report=xml
+          pytest -v --durations=10 tests/cli/ --cov=axolotl --cov-append --cov-report=xml

      - name: Upload coverage to Codecov
        uses: codecov/codecov-action@v5
@@ -116,6 +118,9 @@ jobs:
          flags: unittests,pytorch-${{ matrix.pytorch_version }}
          fail_ci_if_error: false

+      - name: cleanup pip cache
+        run: |
+          find "$(pip cache dir)/http-v2" -type f -mtime +14 -exec rm {} \;

  pytest-sdist:
    name: PyTest from Source Dist
@@ -142,26 +147,25 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: ${{ matrix.python_version }}
+          cache: 'pip' # caching pip dependencies

-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
+      - name: upgrade pip
+        run: |
+          pip3 install --upgrade pip
+          pip3 install --upgrade packaging==23.2 setuptools==75.8.0 setuptools_scm build wheel

      - name: Install PyTorch
        run: |
-          uv pip install --system torch==${{ matrix.pytorch_version }} torchvision
+          pip3 install torch==${{ matrix.pytorch_version }} torchvision

      - name: Install dependencies
        run: |
-          uv pip show --system torch
-          uv pip install --system wheel build setuptools_scm
-          python -m build --sdist
-          printf "torch==${{ matrix.pytorch_version }}\n" > torch-constraints.txt
-          tarball_path=$(echo dist/axolotl*.tar.gz)
-          uv pip install --no-cache-dir --no-build-isolation --system "${tarball_path}[dev]" --constraints torch-constraints.txt
+          pip3 show torch
+          python -m build --no-isolation --sdist
+          pip3 install --no-build-isolation dist/axolotl*.tar.gz
          python scripts/unsloth_install.py | sh
          python scripts/cutcrossentropy_install.py | sh
+          pip3 install -r requirements-dev.txt -r requirements-tests.txt

      - name: Make sure PyTorch version wasn't clobbered
        run: |
@@ -176,9 +180,13 @@ jobs:

      - name: Run tests
        run: |
-          python -m pytest -v --durations=10 -n 8 --dist loadfile --cov=axolotl --cov-report=xml --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ --ignore=tests/monkeypatch/ tests/
-          python -m pytest -v --durations=10 -n 8 --cov=axolotl --cov-append --cov-report=xml tests/monkeypatch/
-          python -m pytest -v --durations=10 -n 8 tests/cli/
+          pytest -v --durations=10 -n8 --dist loadfile --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ --ignore=tests/monkeypatch/ tests/ --cov=axolotl --cov-report=xml
+          pytest -v --durations=10 tests/monkeypatch/ --cov=axolotl --cov-append --cov-report=xml
+          pytest -v --durations=10 tests/cli/
+
+      - name: cleanup pip cache
+        run: |
+          find "$(pip cache dir)/http-v2" -type f -mtime +14 -exec rm {} \;

  gate-skip-e2e:
    needs: [pre-commit, pytest, pytest-sdist]
@@ -235,7 +243,7 @@ jobs:
            pytorch: 2.7.1
            num_gpus: 1
            axolotl_extras:
-            dockerfile: "Dockerfile.jinja"
+            dockerfile: "Dockerfile-uv.jinja"
    steps:
      - name: Checkout
        uses: actions/checkout@v4
@@ -243,17 +251,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
          python -m pip install --upgrade pip
-          pip install modal==1.0.2 jinja2 protobuf
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=${{ github.ref_name }}-base-uv-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
+          echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
          echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
          echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
          echo "AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}" >> $GITHUB_ENV
@@ -308,17 +312,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
          python -m pip install --upgrade pip
-          pip install modal==1.0.2 jinja2 protobuf
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=${{ github.ref_name }}-base-uv-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
+          echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
          echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
          echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
          echo "AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}" >> $GITHUB_ENV
@@ -355,17 +355,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
          python -m pip install --upgrade pip
-          pip install modal==1.0.2 jinja2 protobuf
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=${{ github.ref_name }}-base-uv-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
+          echo "BASE_TAG=main-base-py${{ matrix.python_version }}-cu${{ matrix.cuda }}-${{ matrix.pytorch }}" >> $GITHUB_ENV
          echo "PYTORCH_VERSION=${{ matrix.pytorch}}" >> $GITHUB_ENV
          echo "AXOLOTL_ARGS=${{ matrix.axolotl_args}}" >> $GITHUB_ENV
          echo "AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}" >> $GITHUB_ENV
--- a/.gitignore
+++ b/.gitignore
@@ -191,5 +191,5 @@ out/
 # vim
 *.swp

-# setuptools-scm generated version file
+# scm auto-versioning
 src/axolotl/_version.py
--- a/.runpod/Dockerfile
+++ b/.runpod/Dockerfile
@@ -1,8 +1,9 @@
 FROM axolotlai/axolotl-cloud:main-py3.11-cu124-2.6.0

 COPY .runpod/requirements.txt /requirements.txt
-RUN curl -LsSf https://astral.sh/uv/install.sh | sh && \
-    /root/.local/bin/uv pip install --system -r /requirements.txt
+RUN --mount=type=cache,target=/root/.cache/pip \
+    python3 -m pip install --upgrade pip && \
+    python3 -m pip install --upgrade -r /requirements.txt

 # Environment settings
 ARG BASE_VOLUME="/runpod-volume"
--- a/MANIFEST.in
+++ b/MANIFEST.in
@@ -1,5 +1,6 @@
-include pyproject.toml
+include requirements.txt
 include README.md
 include LICENSE
+include src/setuptools_axolotl_dynamic_dependencies.py
 include src/axolotl/utils/chat_templates/templates/*.jinja
-recursive-include src/axolotl *.py
+recursive-include axolotl *.py
--- a/README.md
+++ b/README.md
@@ -65,9 +65,15 @@ Features:
 - **Flexible Dataset Handling**: Load from local, HuggingFace, and cloud (S3, Azure, GCP, OCI) datasets.
 - **Cloud Ready**: We ship [Docker images](https://hub.docker.com/u/axolotlai) and also [PyPI packages](https://pypi.org/project/axolotl/) for use on cloud platforms and local hardware.

+
+
 ## 🚀 Quick Start - LLM Fine-tuning in Minutes

-**Requirements**: NVIDIA GPU (Ampere+) or AMD GPU, Python 3.11+
+**Requirements**:
+
+- NVIDIA GPU (Ampere or newer for `bf16` and Flash Attention) or AMD GPU
+- Python 3.11
+- PyTorch ≥2.6.0

 ### Google Colab

@@ -75,35 +81,15 @@ Features:

 ### Installation

-#### Project setup (uv add)
+#### Using pip

 ```bash
-# Install uv
-curl -LsSf https://astral.sh/uv/install.sh | sh
-
-# Initialize or enter your project
-uv init my-project && cd my-project
-uv add axolotl
-uv pip install flash-attn --no-build-isolation
-source .venv/bin/activate
+pip3 install -U packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation axolotl[flash-attn,deepspeed]

 # Download example axolotl configs, deepspeed configs
 axolotl fetch examples
-axolotl fetch deepspeed_configs  # optional
-```
-
-#### Quick try (uv pip)
-
-```bash
-# Install uv if needed
-curl -LsSf https://astral.sh/uv/install.sh | sh
-
-uv pip install axolotl
-uv pip install flash-attn --no-build-isolation
-
-# Download example axolotl configs, deepspeed configs
-axolotl fetch examples
-axolotl fetch deepspeed_configs  # optional
+axolotl fetch deepspeed_configs  # OPTIONAL
 ```

 #### Using Docker
--- a/cicd/Dockerfile-uv.jinja
+++ b/cicd/Dockerfile-uv.jinja
@@ -0,0 +1,52 @@
+FROM axolotlai/axolotl-base-uv:{{ BASE_TAG }}
+
+ENV TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6 9.0+PTX"
+ENV AXOLOTL_EXTRAS="{{ AXOLOTL_EXTRAS }}"
+ENV AXOLOTL_ARGS="{{ AXOLOTL_ARGS }}"
+ENV CUDA="{{ CUDA }}"
+ENV PYTORCH_VERSION="{{ PYTORCH_VERSION }}"
+ENV GITHUB_REF="{{ GITHUB_REF }}"
+ENV GITHUB_SHA="{{ GITHUB_SHA }}"
+ENV NIGHTLY_BUILD="{{ NIGHTLY_BUILD }}"
+ENV HF_HOME="{{ HF_HOME }}"
+
+RUN apt-get update && \
+    apt-get install -y --allow-change-held-packages vim curl nano libnccl2 libnccl-dev ibverbs-providers ibverbs-utils infiniband-diags librdmacm-dev librdmacm1 rdmacm-utils slurm-wlm
+
+WORKDIR /workspace
+
+RUN git clone --depth=1 https://github.com/axolotl-ai-cloud/axolotl.git
+
+WORKDIR /workspace/axolotl
+
+RUN git fetch origin +$GITHUB_REF && \
+    git checkout FETCH_HEAD
+
+# If AXOLOTL_EXTRAS is set, append it in brackets
+RUN if [ "$NIGHTLY_BUILD" = "true" ] ; then \
+        sed -i 's#^transformers.*#transformers @ git+https://github.com/huggingface/transformers.git@main#' requirements.txt; \
+        sed -i 's#^peft.*#peft @ git+https://github.com/huggingface/peft.git@main#' requirements.txt; \
+        sed -i 's#^accelerate.*#accelerate @ git+https://github.com/huggingface/accelerate.git@main#' requirements.txt; \
+        sed -i 's#^trl.*#trl @ git+https://github.com/huggingface/trl.git@main#' requirements.txt; \
+        sed -i 's#^datasets.*#datasets @ git+https://github.com/huggingface/datasets.git@main#' requirements.txt; \
+    fi
+
+RUN uv pip install packaging==23.2 setuptools==75.8.0
+RUN if [ "$AXOLOTL_EXTRAS" != "" ] ; then \
+        uv pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray,$AXOLOTL_EXTRAS] $AXOLOTL_ARGS; \
+    else \
+        uv pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray] $AXOLOTL_ARGS; \
+    fi
+
+RUN python scripts/unsloth_install.py --uv | sh
+RUN python scripts/cutcrossentropy_install.py --uv | sh
+
+# So we can test the Docker image
+RUN uv pip install -r requirements-dev.txt -r requirements-tests.txt
+
+# fix so that git fetch/pull from remote works
+RUN git config remote.origin.fetch "+refs/heads/*:refs/remotes/origin/*" && \
+    git config --get remote.origin.fetch
+
+# helper for huggingface-login cli
+RUN git config --global credential.helper store
--- a/cicd/Dockerfile.jinja
+++ b/cicd/Dockerfile.jinja
@@ -1,10 +1,6 @@
-FROM axolotlai/axolotl-base-uv:{{ BASE_TAG }}
+FROM axolotlai/axolotl-base:{{ BASE_TAG }}

-SHELL ["/bin/bash", "-euxo", "pipefail", "-c"]
-
-ARG VENV_PYTHON="/workspace/axolotl-venv/bin/python"
-
-ENV TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6 9.0+PTX"
+ENV TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6+PTX"
 ENV AXOLOTL_EXTRAS="{{ AXOLOTL_EXTRAS }}"
 ENV AXOLOTL_ARGS="{{ AXOLOTL_ARGS }}"
 ENV CUDA="{{ CUDA }}"
@@ -13,7 +9,7 @@ ENV GITHUB_REF="{{ GITHUB_REF }}"
 ENV GITHUB_SHA="{{ GITHUB_SHA }}"
 ENV NIGHTLY_BUILD="{{ NIGHTLY_BUILD }}"
 ENV HF_HOME="{{ HF_HOME }}"
-ENV VENV_PYTHON=$VENV_PYTHON
+ENV AXOLOTL_DATASET_PROCESSES="8"

 RUN apt-get update && \
    apt-get install -y --allow-change-held-packages vim curl nano libnccl2 libnccl-dev ibverbs-providers ibverbs-utils infiniband-diags librdmacm-dev librdmacm1 rdmacm-utils slurm-wlm
@@ -29,27 +25,25 @@ RUN git fetch origin +$GITHUB_REF && \

 # If AXOLOTL_EXTRAS is set, append it in brackets
 RUN if [ "$NIGHTLY_BUILD" = "true" ] ; then \
-        sed -i 's#"transformers[^"]*"#"transformers @ git+https://github.com/huggingface/transformers.git@main"#' pyproject.toml; \
-        sed -i 's#"peft[^"]*"#"peft @ git+https://github.com/huggingface/peft.git@main"#' pyproject.toml; \
-        sed -i 's#"accelerate[^"]*"#"accelerate @ git+https://github.com/huggingface/accelerate.git@main"#' pyproject.toml; \
-        sed -i 's#"trl[^"]*"#"trl @ git+https://github.com/huggingface/trl.git@main"#' pyproject.toml; \
-        sed -i 's#"datasets[^"]*"#"datasets @ git+https://github.com/huggingface/datasets.git@main"#' pyproject.toml; \
+        sed -i 's#^transformers.*#transformers @ git+https://github.com/huggingface/transformers.git@main#' requirements.txt; \
+        sed -i 's#^peft.*#peft @ git+https://github.com/huggingface/peft.git@main#' requirements.txt; \
+        sed -i 's#^accelerate.*#accelerate @ git+https://github.com/huggingface/accelerate.git@main#' requirements.txt; \
+        sed -i 's#^trl.*#trl @ git+https://github.com/huggingface/trl.git@main#' requirements.txt; \
+        sed -i 's#^datasets.*#datasets @ git+https://github.com/huggingface/datasets.git@main#' requirements.txt; \
    fi

-RUN uv pip install --python "$VENV_PYTHON" packaging==23.2 setuptools==75.8.0 pip
+RUN pip install packaging==23.2 setuptools==75.8.0
 RUN if [ "$AXOLOTL_EXTRAS" != "" ] ; then \
-        uv pip install --python "$VENV_PYTHON" --no-build-isolation -e .[ring-flash-attn,optimizers,ray,${AXOLOTL_EXTRAS}] $AXOLOTL_ARGS; \
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray,$AXOLOTL_EXTRAS] $AXOLOTL_ARGS; \
    else \
-        uv pip install --python "$VENV_PYTHON" --no-build-isolation -e .[ring-flash-attn,optimizers,ray] $AXOLOTL_ARGS; \
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray] $AXOLOTL_ARGS; \
    fi

-RUN uv pip install --python "$VENV_PYTHON" --no-build-isolation flash-attn $AXOLOTL_ARGS
-
-RUN "$VENV_PYTHON" scripts/unsloth_install.py | sh
-RUN "$VENV_PYTHON" scripts/cutcrossentropy_install.py | sh
+RUN python scripts/unsloth_install.py | sh
+RUN python scripts/cutcrossentropy_install.py | sh

 # So we can test the Docker image
-RUN uv pip install --python "$VENV_PYTHON" -e ".[dev]"
+RUN pip install -r requirements-dev.txt -r requirements-tests.txt

 # fix so that git fetch/pull from remote works
 RUN git config remote.origin.fetch "+refs/heads/*:refs/remotes/origin/*" && \
--- a/cicd/cicd.sh
+++ b/cicd/cicd.sh
@@ -4,7 +4,7 @@ set -e
 python -c "import torch; assert '$PYTORCH_VERSION' in torch.__version__"

 # Run unit tests with initial coverage report
-uv run pytest -v --durations=10 -n8 \
+pytest -v --durations=10 -n8 \
  --ignore=tests/e2e/ \
  --ignore=tests/patched/ \
  --ignore=tests/cli \
@@ -12,36 +12,36 @@ uv run pytest -v --durations=10 -n8 \
  --cov=axolotl

 # Run lora kernels tests with coverage append
-uv run pytest -v --durations=10 \
+pytest -v --durations=10 \
  /workspace/axolotl/tests/e2e/patched/lora_kernels \
  --cov=axolotl \
  --cov-append

 # Run patched tests excluding lora kernels with coverage append
-uv run pytest --full-trace -vvv --durations=10 \
+pytest --full-trace -vvv --durations=10 \
  --ignore=tests/e2e/patched/lora_kernels \
  /workspace/axolotl/tests/e2e/patched \
  --cov=axolotl \
  --cov-append

 # Run solo tests with coverage append
-uv run pytest -v --durations=10 -n1 \
+pytest -v --durations=10 -n1 \
  /workspace/axolotl/tests/e2e/solo/ \
  --cov=axolotl \
  --cov-append

 # Run integration tests with coverage append
-uv run pytest -v --durations=10 \
+pytest -v --durations=10 \
  /workspace/axolotl/tests/e2e/integrations/ \
  --cov=axolotl \
  --cov-append

-uv run pytest -v --durations=10 /workspace/axolotl/tests/cli \
+pytest -v --durations=10 /workspace/axolotl/tests/cli \
  --cov=axolotl \
  --cov-append

 # Run remaining e2e tests with coverage append and final report
-uv run pytest -v --durations=10 \
+pytest -v --durations=10 \
  --ignore=tests/e2e/solo/ \
  --ignore=tests/e2e/patched/ \
  --ignore=tests/e2e/multigpu/ \
@@ -52,4 +52,4 @@ uv run pytest -v --durations=10 \
  --cov-append \
  --cov-report=xml:e2e-coverage.xml

-uv run codecov upload-process -t $CODECOV_TOKEN -f e2e-coverage.xml -F e2e,pytorch-${PYTORCH_VERSION} || true
+codecov upload-process -t $CODECOV_TOKEN -f e2e-coverage.xml -F e2e,pytorch-${PYTORCH_VERSION} || true
--- a/cicd/multigpu.py
+++ b/cicd/multigpu.py
@@ -23,7 +23,7 @@ df_args = {
    "AXOLOTL_EXTRAS": os.environ.get("AXOLOTL_EXTRAS", ""),
    "AXOLOTL_ARGS": os.environ.get("AXOLOTL_ARGS", ""),
    "PYTORCH_VERSION": os.environ.get("PYTORCH_VERSION", "2.6.0"),
-    "BASE_TAG": os.environ.get("BASE_TAG", "main-base-uv-py3.11-cu126-2.6.0"),
+    "BASE_TAG": os.environ.get("BASE_TAG", "main-base-py3.11-cu126-2.6.0"),
    "CUDA": os.environ.get("CUDA", "126"),
    "GITHUB_REF": os.environ.get("GITHUB_REF", "refs/heads/main"),
    "GITHUB_SHA": os.environ.get("GITHUB_SHA", ""),
--- a/cicd/single_gpu.py
+++ b/cicd/single_gpu.py
@@ -23,7 +23,7 @@ df_args = {
    "AXOLOTL_EXTRAS": os.environ.get("AXOLOTL_EXTRAS", ""),
    "AXOLOTL_ARGS": os.environ.get("AXOLOTL_ARGS", ""),
    "PYTORCH_VERSION": os.environ.get("PYTORCH_VERSION", "2.6.0"),
-    "BASE_TAG": os.environ.get("BASE_TAG", "main-base-uv-py3.11-cu126-2.6.0"),
+    "BASE_TAG": os.environ.get("BASE_TAG", "main-base-py3.11-cu126-2.6.0"),
    "CUDA": os.environ.get("CUDA", "126"),
    "GITHUB_REF": os.environ.get("GITHUB_REF", "refs/heads/main"),
    "GITHUB_SHA": os.environ.get("GITHUB_SHA", ""),
--- a/docker/Dockerfile
+++ b/docker/Dockerfile
@@ -1,19 +1,13 @@
-ARG BASE_TAG=main-base-uv
-FROM axolotlai/axolotl-base-uv:$BASE_TAG
+ARG BASE_TAG=main-base
+FROM axolotlai/axolotl-base:$BASE_TAG

 ARG TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6+PTX"
 ARG AXOLOTL_EXTRAS=""
 ARG AXOLOTL_ARGS=""
 ARG CUDA="118"
 ARG PYTORCH_VERSION="2.1.2"
-ARG GIT_REF="refs/heads/main"
-ARG GIT_SHA="HEAD"
-ARG VENV_PYTHON="/workspace/axolotl-venv/bin/python"

 ENV PYTORCH_VERSION=$PYTORCH_VERSION
-ENV GIT_REF=$GIT_REF
-ENV GIT_SHA=$GIT_SHA
-ENV VENV_PYTHON=$VENV_PYTHON

 RUN apt-get update && \
    apt-get install -y --allow-change-held-packages vim curl nano libnccl2 libnccl-dev rsync s3fs && \
@@ -26,19 +20,16 @@ RUN git clone --depth=1 https://github.com/axolotl-ai-cloud/axolotl.git

 WORKDIR /workspace/axolotl

-# Ensure we are on the expected commit and break Docker cache between revisions
-RUN git fetch origin "$GIT_REF" && git checkout "$GIT_SHA"
-
 # If AXOLOTL_EXTRAS is set, append it in brackets
 RUN if [ "$AXOLOTL_EXTRAS" != "" ] ; then \
-        uv pip install --python "$VENV_PYTHON" --no-build-isolation -e .[ring-flash-attn,optimizers,ray,$AXOLOTL_EXTRAS] $AXOLOTL_ARGS; \
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray,$AXOLOTL_EXTRAS] $AXOLOTL_ARGS; \
    else \
-        uv pip install --python "$VENV_PYTHON" --no-build-isolation -e .[ring-flash-attn,optimizers,ray] $AXOLOTL_ARGS; \
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray] $AXOLOTL_ARGS; \
    fi && \
-    uv pip install --python "$VENV_PYTHON" --no-build-isolation flash-attn $AXOLOTL_ARGS && \
-    "$VENV_PYTHON" scripts/unsloth_install.py | sh && \
-    "$VENV_PYTHON" scripts/cutcrossentropy_install.py | sh && \
-    uv pip install --python "$VENV_PYTHON" pytest
+    python scripts/unsloth_install.py | sh && \
+    python scripts/cutcrossentropy_install.py | sh && \
+    pip install pytest && \
+    pip cache purge

 # fix so that git fetch/pull from remote works with shallow clone
 RUN git config remote.origin.fetch "+refs/heads/*:refs/remotes/origin/*" && \
--- a/docker/Dockerfile-base
+++ b/docker/Dockerfile-base
@@ -48,5 +48,5 @@ RUN git lfs install --skip-repo && \
    pip3 cache purge

 RUN if [ "$PYTORCH_VERSION" = "2.6.0" ] && [ "$CUDA" = "124" ] ; then \
-        FLASH_ATTENTION_FORCE_BUILD="TRUE" uv pip install --no-build-isolation flash-attn==2.8.0.post2; \
+        FLASH_ATTENTION_FORCE_BUILD="TRUE" pip3 install --no-build-isolation flash-attn==2.8.0.post2; \
    fi
--- a/docker/Dockerfile-cloud
+++ b/docker/Dockerfile-cloud
@@ -12,8 +12,8 @@ EXPOSE 22
 COPY scripts/cloud-entrypoint.sh /root/cloud-entrypoint.sh
 COPY scripts/motd /etc/motd

-RUN uv pip install --python "$VENV_PYTHON" jupyterlab notebook ipywidgets && \
-    "$VENV_PYTHON" -m jupyter lab clean
+RUN pip install jupyterlab notebook ipywidgets && \
+    jupyter lab clean
 RUN apt update && \
    apt install --yes --no-install-recommends openssh-server tmux iproute2 nvtop && \
    rm -rf /var/cache/apt/archives && \
--- a/docker/Dockerfile-cloud-no-tmux
+++ b/docker/Dockerfile-cloud-no-tmux
@@ -12,8 +12,8 @@ EXPOSE 22
 COPY scripts/cloud-entrypoint.sh /root/cloud-entrypoint.sh
 COPY scripts/motd /etc/motd

-RUN uv pip install --python "$VENV_PYTHON" jupyterlab notebook ipywidgets && \
-    "$VENV_PYTHON" -m jupyter lab clean
+RUN pip install jupyterlab notebook ipywidgets && \
+    jupyter lab clean
 RUN apt update && \
    apt install --yes --no-install-recommends openssh-server tmux iproute2 nvtop ibverbs-providers ibverbs-utils infiniband-diags librdmacm-dev librdmacm1 rdmacm-utils slurm-wlm && \
    rm -rf /var/cache/apt/archives && \
--- a/docker/Dockerfile-tests
+++ b/docker/Dockerfile-tests
@@ -24,14 +24,13 @@ RUN git fetch origin +$GITHUB_REF && \

 # If AXOLOTL_EXTRAS is set, append it in brackets
 RUN if [ "$AXOLOTL_EXTRAS" != "" ] ; then \
-        uv pip install --no-build-isolation -e .[deepspeed,mamba-ssm,$AXOLOTL_EXTRAS] $AXOLOTL_ARGS; \
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,mamba-ssm,$AXOLOTL_EXTRAS] $AXOLOTL_ARGS; \
    else \
-        uv pip install --no-build-isolation -e .[deepspeed,mamba-ssm] $AXOLOTL_ARGS; \
-    fi && \
-    uv pip install --no-build-isolation flash-attn $AXOLOTL_ARGS
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,mamba-ssm] $AXOLOTL_ARGS; \
+    fi

 # So we can test the Docker image
-RUN uv pip install pytest
+RUN pip install pytest

 # fix so that git fetch/pull from remote works
 RUN git config remote.origin.fetch "+refs/heads/*:refs/remotes/origin/*" && \
--- a/docker/Dockerfile-uv-base
+++ b/docker/Dockerfile-uv-base
@@ -13,7 +13,6 @@ ARG TORCH_CUDA_ARCH_LIST="7.0 7.5 8.0 8.6 9.0+PTX"
 ENV PYTHON_VERSION=$PYTHON_VERSION
 ENV TORCH_CUDA_ARCH_LIST=$TORCH_CUDA_ARCH_LIST
 ENV UV_TORCH_BACKEND="cu${CUDA}"
-ENV VENV_PYTHON=/workspace/axolotl-venv/bin/python

 RUN apt-get update \
    && apt-get install -y wget git build-essential ninja-build git-lfs libaio-dev pkg-config curl && rm -rf /var/lib/apt/lists/* \
@@ -30,8 +29,8 @@ RUN uv venv --no-project --relocatable axolotl-venv

 ENV PATH="/workspace/axolotl-venv/bin:${PATH}"

-RUN uv pip install --python "$VENV_PYTHON" packaging setuptools wheel psutil protobuf grpclib \
-    && uv pip install --python "$VENV_PYTHON" torch==${PYTORCH_VERSION} \
-    && uv pip install --python "$VENV_PYTHON" --no-build-isolation "causal_conv1d @ git+https://github.com/Dao-AILab/causal-conv1d.git@main" \
-    && uv pip install --python "$VENV_PYTHON" "mamba_ssm @ git+https://github.com/state-spaces/mamba.git@main" \
-    && uv pip install --python "$VENV_PYTHON" awscli pydantic
+RUN uv pip install packaging setuptools wheel psutil \
+    && uv pip install torch==${PYTORCH_VERSION} \
+    && uv pip install --no-build-isolation "causal_conv1d @ git+https://github.com/Dao-AILab/causal-conv1d.git@main" \
+    && uv pip install "mamba_ssm @ git+https://github.com/state-spaces/mamba.git@main" \
+    && uv pip install awscli pydantic
--- a/docs/debugging.qmd
+++ b/docs/debugging.qmd
@@ -72,8 +72,8 @@ datasets:
 Make sure you have an [editable install](https://setuptools.pypa.io/en/latest/userguide/development_mode.html) of Axolotl, which ensures that changes you make to the code are reflected at runtime.  Run the following commands from the root of this project:

 ```bash
-uv sync --extra deepspeed
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging
+pip3 install --no-build-isolation -e '.[flash-attn,deepspeed]'
 ```

 #### Remote Hosts
@@ -213,8 +213,8 @@ docker run --privileged --gpus '"all"' --shm-size 10g --rm -it --name axolotl --
 You will now be in the container.  Next, perform an editable install of Axolotl:

 ```bash
-uv sync --extra deepspeed
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging
+pip3 install --no-build-isolation -e '.[flash-attn,deepspeed]'
 ```

 ### Attach To Container
--- a/docs/fsdp_qlora.qmd
+++ b/docs/fsdp_qlora.qmd
@@ -1,5 +1,5 @@
 ---
-title: "FSDP + QLoRA"
+title: "FDSP + QLoRA"
 description: Use FSDP with QLoRA to fine-tune large LLMs on consumer GPUs.
 format:
  html:
@@ -23,12 +23,6 @@ To enable `QLoRA` with `FSDP`, you need to perform the following steps:
 2. Enable FSDP in your axolotl config, as [described here](multi-gpu.qmd#sec-fsdp).
 3. Use one of the supported model types: `llama`, `mistral` or `mixtral`.

-## Enabling Swap for FSDP2
-
-If available memory is insufficient even after FSDP's CPU offloading, you can enable swap memory usage by setting `cpu_offload_pin_memory: false` alongside `offload_params: true` in FSDP config.
-
-This disables memory pinning, allowing FSDP to use disk swap space as fallback. Disabling memory pinning itself incurs performance overhead, and actually having to use swap adds more, but it may enable training larger models that would otherwise cause OOM errors on resource constrained systems.
-
 ## Example Config

 [examples/llama-2/qlora-fsdp.yml](../examples/llama-2/qlora-fsdp.yml) contains an example of how to enable QLoRA + FSDP in axolotl.
--- a/docs/installation.qmd
+++ b/docs/installation.qmd
@@ -29,40 +29,19 @@ Follow the instructions at: [https://pytorch.org/get-started/locally/](https://p
 For Blackwell GPUs, please use Pytorch 2.7.0 and CUDA 12.8.
 :::

-### uv Installation (Recommended) {#sec-uv-quick}
+### PyPI Installation (Recommended) {#sec-pypi}

 ```{.bash}
-# Install uv if not already installed
-curl -LsSf https://astral.sh/uv/install.sh | sh
-
-# Add Axolotl to a project (recommended)
-uv init my-project && cd my-project
-uv add axolotl
-uv pip install flash-attn --no-build-isolation
-source .venv/bin/activate
-```
-
-For a quick one-off install without creating a project:
-
-```{.bash}
-uv pip install axolotl
-uv pip install flash-attn --no-build-isolation
-```
-
-### pip Installation {#sec-pypi}
-
-```{.bash}
-pip install --no-build-isolation axolotl[deepspeed]
-pip install --no-build-isolation flash-attn
+pip3 install -U packaging setuptools wheel ninja
+pip3 install --no-build-isolation axolotl[flash-attn,deepspeed]
 ```

 We use `--no-build-isolation` in order to detect the installed PyTorch version (if
 installed) in order not to clobber it, and so that we set the correct version of
 dependencies that are specific to the PyTorch version or other installed
-co-dependencies. Flash Attention is resolved separately so it can be built against
-the environment configured by the previous step.
+co-dependencies.

-### Advanced uv Installation {#sec-uv}
+### uv Installation {#sec-uv}

 uv is a fast, reliable Python package installer and resolver built in Rust. It offers significant performance improvements over pip and provides better dependency resolution, making it an excellent choice for complex environments.

@@ -83,38 +62,28 @@ source .venv/bin/activate
 Install PyTorch
 - PyTorch 2.6.0 recommended
 ```{.bash}
+uv pip install packaging setuptools wheel
 uv pip install torch==2.6.0
 uv pip install awscli pydantic
 ```

 Install axolotl from PyPi
 ```{.bash}
-uv pip install --no-build-isolation axolotl[deepspeed]
-# optionally install with vLLM if you're using torch==2.6.0 and want to train w/ GRPO
-# uv pip install --no-build-isolation axolotl[deepspeed,vllm]
+uv pip install --no-build-isolation axolotl[deepspeed,flash-attn]

-uv pip install flash-attn --no-build-isolation
+# optionally install with vLLM if you're using torch==2.6.0 and want to train w/ GRPO
+uv pip install --no-build-isolation axolotl[deepspeed,flash-attn,vllm]
 ```

 ### Edge/Development Build {#sec-edge-build}

 For the latest features between releases:

-#### Using uv (recommended)
 ```{.bash}
 git clone https://github.com/axolotl-ai-cloud/axolotl.git
 cd axolotl
-curl -LsSf https://astral.sh/uv/install.sh | sh  # If not already installed
-uv sync
-uv pip install flash-attn --no-build-isolation
-```
-
-#### Using pip
-```{.bash}
-git clone https://github.com/axolotl-ai-cloud/axolotl.git
-cd axolotl
-pip install --no-build-isolation -e '.[deepspeed]'
-pip install --no-build-isolation flash-attn
+pip3 install -U packaging setuptools wheel ninja
+pip3 install --no-build-isolation -e '.[flash-attn,deepspeed]'
 ```

 ### Docker {#sec-docker}
@@ -172,7 +141,7 @@ For providers supporting Docker:
 ### macOS {#sec-macos}

 ```{.bash}
-uv pip install --no-build-isolation -e '.'
+pip3 install --no-build-isolation -e '.'
 ```

 See @sec-troubleshooting for Mac-specific issues.
@@ -190,15 +159,10 @@ We recommend using WSL2 (Windows Subsystem for Linux) or Docker.
 1. Install Python ≥3.11
 2. Install PyTorch: https://pytorch.org/get-started/locally/
 3. Install Axolotl:
-```{.bash}
-# Option A: add Axolotl to the environment
-uv add axolotl
-uv pip install flash-attn --no-build-isolation
-
-# Option B: quick install
-uv pip install axolotl
-uv pip install flash-attn --no-build-isolation
-```
+   ```{.bash}
+   pip3 install -U packaging setuptools wheel ninja
+   pip3 install --no-build-isolation -e '.[flash-attn,deepspeed]'
+   ```
 4. (Optional) Login to Hugging Face:
   ```{.bash}
   huggingface-cli login
--- a/docs/lora_optims.qmd
+++ b/docs/lora_optims.qmd
@@ -5,11 +5,10 @@ description: "Custom autograd functions and Triton kernels in Axolotl for optimi

 Inspired by [Unsloth](https://github.com/unslothai/unsloth), we've implemented two
 optimizations for LoRA and QLoRA fine-tuning, supporting both single GPU and multi-GPU
-(including the DDP, DeepSpeed, and FSDP2 settings) training. These include (1) SwiGLU
-and GEGLU activation function Triton kernels, and (2) LoRA MLP and attention custom
-autograd functions. Our goal was to leverage operator fusion and tensor re-use in order
-to improve speed and reduce memory usage during the forward and backward passes of
-these calculations.
+(in the DDP and DeepSpeed settings) training. These include (1) SwiGLU and GEGLU activation function
+Triton kernels, and (2) LoRA MLP and attention custom autograd functions. Our goal was
+to leverage operator fusion and tensor re-use in order to improve speed and reduce
+memory usage during the forward and backward passes of these calculations.

 We currently support several common model architectures, including (but not limited to):

@@ -132,5 +131,6 @@ computation path.
 ## Future Work

 - Support for additional model architectures
+- Support for the FSDP setting
 - Support for dropout and bias
 - Additional operator fusions
--- a/docs/multimodal.qmd
+++ b/docs/multimodal.qmd
@@ -95,7 +95,7 @@ chat_template: llava
 ### Mistral-Small-3.1 {#sec-mistral-small-31}

 ::: {.callout-tip}
-Please make sure to install vision lib via `uv pip install 'mistral-common[opencv]==1.8.5'`
+Please make sure to install vision lib via `pip install 'mistral-common[opencv]==1.8.5'`
 :::

 ```yaml
@@ -105,7 +105,7 @@ base_model: mistralai/Mistral-Small-3.1-24B-Instruct-2503
 ### Magistral-Small-2509 {#sec-magistral-small-2509}

 ::: {.callout-tip}
-Please make sure to install vision lib via `uv pip install 'mistral-common[opencv]==1.8.5'`
+Please make sure to install vision lib via `pip install 'mistral-common[opencv]==1.8.5'`
 :::

 ```yaml
@@ -115,7 +115,7 @@ base_model: mistralai/Magistral-Small-2509
 ### Voxtral {#sec-voxtral}

 ::: {.callout-tip}
-Please make sure to install audio lib via `uv pip install librosa==0.11.0 'mistral_common[audio]==1.8.3'`
+Please make sure to install audio lib via `pip3 install librosa==0.11.0 'mistral_common[audio]==1.8.3'`
 :::

 ```yaml
@@ -143,7 +143,7 @@ The model's initial loss and grad norm will be very high. We suspect this to be
 :::

 ::: {.callout-tip}
-Please make sure to install `timm` via `uv pip install timm==1.0.17`
+Please make sure to install `timm` via `pip3 install timm==1.0.17`
 :::

 ```yaml
@@ -171,7 +171,7 @@ chat_template: qwen2_vl  # same as qwen2-vl
 ### SmolVLM2 {#sec-smolvlm2}

 ::: {.callout-tip}
-Please make sure to install `num2words` via `uv pip install num2words==0.5.14`
+Please make sure to install `num2words` via `pip3 install num2words==0.5.14`
 :::

 ```yaml
@@ -181,7 +181,7 @@ base_model: HuggingFaceTB/SmolVLM2-500M-Video-Instruct
 ### LFM2-VL {#sec-lfm2-vl}

 ::: {.callout-warning}
-Please uninstall `causal-conv1d` via `uv pip uninstall -y causal-conv1d`
+Please uninstall `causal-conv1d` via `pip3 uninstall -y causal-conv1d`
 :::

 ```yaml
@@ -222,7 +222,7 @@ For audio loading, you can use the following keys within `content` alongside `"t

 ::: {.callout-tip}

-You may need to install `librosa` via `uv pip install librosa==0.11.0`.
+You may need to install `librosa` via `pip3 install librosa==0.11.0`.

 :::

--- a/docs/sequence_parallelism.qmd
+++ b/docs/sequence_parallelism.qmd
@@ -49,9 +49,9 @@ When sequence parallelism is enabled:
 To use sequence parallelism, you need:

 - Multiple GPUs (at least 2)
- The `ring-flash-attn` package. Install with either `uv sync --extra ring-flash-attn`
-  (from a cloned repository) or `uv pip install ring-flash-attn>=0.1.4`.
- Flash Attention installed separately with `uv pip install flash-attn --no-build-isolation`.
+- The `ring-flash-attn` package. Install with:
+  - `pip install axolotl[ring-flash-attn]` (preferred)
+  - `pip install ring-flash-attn>=0.1.4`

 ## Limitations

--- a/examples/LiquidAI/README.md
+++ b/examples/LiquidAI/README.md
@@ -12,14 +12,9 @@ This guide shows how to fine-tune both the LFM2 and LFM2-VL models with Axolotl.

    Here is an example of how to install from pip:
    ```bash
-    # Ensure you have a compatible version of PyTorch installed
-    # Option A: manage dependencies in your project
-    uv add 'axolotl>=0.12.0'
-    uv pip install flash-attn --no-build-isolation
-
-    # Option B: quick install
-    uv pip install 'axolotl>=0.12.0'
-    uv pip install flash-attn --no-build-isolation
+    # Ensure you have a compatible version of Pytorch installed
+    pip3 install packaging setuptools wheel ninja
+    pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
    ```

 2.  Run one of the finetuning examples below.
@@ -40,7 +35,7 @@ This guide shows how to fine-tune both the LFM2 and LFM2-VL models with Axolotl.

 - **Installation Error**: If you encounter `ImportError: ... undefined symbol ...` or `ModuleNotFoundError: No module named 'causal_conv1d_cuda'`, the `causal-conv1d` package may have been installed incorrectly. Try uninstalling it:
  ```bash
-  uv pip uninstall -y causal-conv1d
+  pip uninstall -y causal-conv1d
  ```

 - **Dataset Loading**: Read more on how to load your own dataset in our [documentation](https://docs.axolotl.ai/docs/dataset_loading.html).
--- a/examples/apertus/README.md
+++ b/examples/apertus/README.md
@@ -15,8 +15,8 @@ This guide shows how to fine-tune it with Axolotl with multi-turn conversations
 git clone https://github.com/axolotl-ai-cloud/axolotl.git
 cd axolotl

-uv sync
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation -e '.[flash-attn]'

 # Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
 python scripts/cutcrossentropy_install.py | sh
@@ -31,7 +31,7 @@ python scripts/cutcrossentropy_install.py | sh
 # For those using our Docker image, use the below path.
 export CUDA_HOME=/usr/local/cuda

-uv pip install git+https://github.com/nickjbrowning/XIELU@59d6031 --no-build-isolation --no-deps
+pip3 install git+https://github.com/nickjbrowning/XIELU@59d6031 --no-build-isolation --no-deps
 ```

 For any installation errors, see [XIELU Installation Issues](#xielu-installation-issues)
@@ -67,7 +67,7 @@ If those didn't help, please try the below solutions:
 1. Pass env for CMAKE and try install again:

    ```bash
-    Python_EXECUTABLE=$(which python) uv pip install git+https://github.com/nickjbrowning/XIELU@59d6031 --no-build-isolation --no-deps
+    Python_EXECUTABLE=$(which python) pip3 install git+https://github.com/nickjbrowning/XIELU@59d6031 --no-build-isolation --no-deps
    ```

 2. Git clone the repo and manually hardcode python path:
@@ -92,7 +92,7 @@ If those didn't help, please try the below solutions:
    ```

    ```bash
-    uv pip install . --no-build-isolation --no-deps
+    pip3 install . --no-build-isolation --no-deps
    ```

 ## Optimization Guides
--- a/examples/arcee/README.md
+++ b/examples/arcee/README.md
@@ -17,8 +17,8 @@ Thanks to the team at Arcee.ai for using Axolotl in supervised fine-tuning the A
 git clone https://github.com/axolotl-ai-cloud/axolotl.git
 cd axolotl

-uv sync
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation -e '.[flash-attn]'

 # Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
 python scripts/cutcrossentropy_install.py | sh
--- a/examples/colab-notebooks/colab-axolotl-example.ipynb
+++ b/examples/colab-notebooks/colab-axolotl-example.ipynb
@@ -12,10 +12,10 @@
    "\n",
    "Axolotl is the most performant LLM post-training framework available, delivering faster training with efficient, consistent and stable performance. Train your workload and ship your product 30% faster; saving you both time and money.\n",
    "\n",
-    "- \u2b50 us on [GitHub](https://github.com/axolotl-ai-cloud/axolotl)\n",
-    "- \ud83d\udcdc Read the [Docs](http://docs.axolotl.ai/)\n",
-    "- \ud83d\udcac Chat with us on [Discord](https://discord.gg/mnpEYgRUmD)\n",
-    "- \ud83d\udcf0 Get updates on [X/Twitter](https://x.com/axolotl_ai)\n"
+    "- ⭐ us on [GitHub](https://github.com/axolotl-ai-cloud/axolotl)\n",
+    "- 📜 Read the [Docs](http://docs.axolotl.ai/)\n",
+    "- 💬 Chat with us on [Discord](https://discord.gg/mnpEYgRUmD)\n",
+    "- 📰 Get updates on [X/Twitter](https://x.com/axolotl_ai)\n"
   ]
  },
  {
@@ -39,8 +39,8 @@
   "source": [
    "%%capture\n",
    "# This step can take ~5-10 minutes to install dependencies\n",
-    "!uv pip install --no-build-isolation axolotl>=0.9.1\n!uv pip install flash-attn --no-build-isolation\n",
-    "!uv pip install \"cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@147ea28\""
+    "!pip install --no-build-isolation axolotl[flash-attn]>=0.9.1\n",
+    "!pip install \"cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@c5aa3ef\""
   ]
  },
  {
@@ -1371,7 +1371,7 @@
       "version_minor": 0
      },
      "text/plain": [
-       "VBox(children=(HTML(value='<center> <img\\nsrc=https://huggingface.co/front/assets/huggingface_logo-noborder.sv\u2026"
+       "VBox(children=(HTML(value='<center> <img\\nsrc=https://huggingface.co/front/assets/huggingface_logo-noborder.sv…"
      ]
     },
     "metadata": {},
@@ -1729,9 +1729,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_12815f401eba44658caa7b2e490137a8",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_30e02aa2d0d241979369e598287f2639",
-      "value": "Drop\u2007Samples\u2007with\u2007Zero\u2007Trainable\u2007Tokens\u2007(num_proc=2):\u2007100%"
+      "value": "Drop Samples with Zero Trainable Tokens (num_proc=2): 100%"
     }
    },
    "083f9cda8d754c168beee10d2f8955a2": {
@@ -1774,9 +1774,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_b195f160ca20442fadd8b5aed0ee41af",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_ca65e32eb52f48c09a84b33cb18f22cd",
-      "value": "\u200711.4M/11.4M\u2007[00:00&lt;00:00,\u200721.8MB/s]"
+      "value": " 11.4M/11.4M [00:00&lt;00:00, 21.8MB/s]"
     }
    },
    "0a46ad75c198463d843fb35e813642cb": {
@@ -1917,7 +1917,7 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_b1bea589efa14258a9982071b87938bf",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_590eef89881545aa8bbef9a8bbe7fb00",
      "value": "\n<b>Pro Tip:</b> If you don't already have one, you can create a dedicated\n'notebooks' token with 'write' access, that you can then easily reuse for all\nnotebooks. </center>"
     }
@@ -1938,9 +1938,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_bfcdbba993b74972a9e3e575f86908ff",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_6ebb2ec171414e47a14765505f64bb3c",
-      "value": "\u20073.84G/3.84G\u2007[00:09&lt;00:00,\u2007664MB/s]"
+      "value": " 3.84G/3.84G [00:09&lt;00:00, 664MB/s]"
     }
    },
    "0e936d9dbf9c4fdd86bbfe9730dedc47": {
@@ -2296,9 +2296,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_349eee9f56d64f0cba6fc24ff2c50c9b",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_7e5d3774060e4589aa65982da5ea4ef4",
-      "value": "\u20079985/9985\u2007[00:04&lt;00:00,\u20072604.11\u2007examples/s]"
+      "value": " 9985/9985 [00:04&lt;00:00, 2604.11 examples/s]"
     }
    },
    "16d1283741404b7bb319094c992fce01": {
@@ -2317,9 +2317,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_a4e5789584564049b83df7c6c54a3e08",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_ff3a94b146a948b6907f5d80c7157f99",
-      "value": "\u20079985/0\u2007[00:00&lt;00:00,\u200750763.46\u2007examples/s]"
+      "value": " 9985/0 [00:00&lt;00:00, 50763.46 examples/s]"
     }
    },
    "1811cda0644e4190a9469d1774435d82": {
@@ -2390,9 +2390,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_e366ae3fceec4566b9ed303d6c5f90af",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_5dd7d150dbe04f08b165ce7f2c27cd11",
-      "value": "model-00008-of-00008.safetensors:\u2007100%"
+      "value": "model-00008-of-00008.safetensors: 100%"
     }
    },
    "19127c7bb1554ccbac877059f9a82db0": {
@@ -2561,9 +2561,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_0dea5caa27384f5689e3cab51f558727",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_a6f48410b9964fefba0c3009a77dc838",
-      "value": "\u20079.68k/9.68k\u2007[00:00&lt;00:00,\u2007812kB/s]"
+      "value": " 9.68k/9.68k [00:00&lt;00:00, 812kB/s]"
     }
    },
    "1f7d30f71bbd4547a9150d21da071055": {
@@ -2634,9 +2634,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_f4a1795dc7514a718f478245f521f0ba",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_5e746eb25bbe416fb585fa24e79f5177",
-      "value": "model-00002-of-00008.safetensors:\u2007100%"
+      "value": "model-00002-of-00008.safetensors: 100%"
     }
    },
    "20352e5f58d24bb8b1f3940efd14fe4a": {
@@ -2707,9 +2707,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_1c6f1f10667545aaab958016ba7e2c94",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_e6e969610738449887259063967f82b0",
-      "value": "\u20072.78M/2.78M\u2007[00:00&lt;00:00,\u200717.8MB/s]"
+      "value": " 2.78M/2.78M [00:00&lt;00:00, 17.8MB/s]"
     }
    },
    "258b7c635c1045329d4669e48c46ccd5": {
@@ -3056,9 +3056,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_be724f04b03942b2a033a7e8898bb4fd",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_fcbab4d8dced41a18dfccce81e3a45a0",
-      "value": "model-00005-of-00008.safetensors:\u2007100%"
+      "value": "model-00005-of-00008.safetensors: 100%"
     }
    },
    "3036608c71904ce9ae4bb2a9fa8802d9": {
@@ -3077,9 +3077,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_5ca6be24acb548cea130bd58e9954c7c",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_5cfb02ee044b4011a378efa8b54a370f",
-      "value": "\u20073.96G/3.96G\u2007[00:10&lt;00:00,\u2007531MB/s]"
+      "value": " 3.96G/3.96G [00:10&lt;00:00, 531MB/s]"
     }
    },
    "30a81da86f8043eca301e86a8651201a": {
@@ -3629,9 +3629,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_8f5bd719974e41c3a8dd9a5b0d3d71e6",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_b87c84de30e84b3abf4871461fb9cbd3",
-      "value": "Loading\u2007checkpoint\u2007shards:\u2007100%"
+      "value": "Loading checkpoint shards: 100%"
     }
    },
    "41f3b32c2f6b4034ae7a3b9124e28bc7": {
@@ -3791,7 +3791,7 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_39789237703c4a418134243055c9cbf5",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_a3a945817f684328b34651fe052393ec",
      "value": "Connecting..."
     }
@@ -4077,9 +4077,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_4d468f96ec924681ad65eb671674b93e",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_ad7599de524549c48bf2d3124ad4b299",
-      "value": "Dropping\u2007Long\u2007Sequences\u2007(num_proc=2):\u2007100%"
+      "value": "Dropping Long Sequences (num_proc=2): 100%"
     }
    },
    "5ca240f31e6b44e3882c5eb37cd5a309": {
@@ -4471,9 +4471,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_5e18768f7ad6434ba8b8b8a2e853e204",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_bb33aec33a6447078c31bfd728942994",
-      "value": "\u2007728/728\u2007[00:00&lt;00:00,\u200720.3kB/s]"
+      "value": " 728/728 [00:00&lt;00:00, 20.3kB/s]"
     }
    },
    "62e302ebdad64aada0ffe64ae1c873f3": {
@@ -4636,9 +4636,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_81c3db71ac704280ad030072655f1537",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_042e091f75694c47aee761e760e76773",
-      "value": "\u20079985/9985\u2007[00:02&lt;00:00,\u20073977.47\u2007examples/s]"
+      "value": " 9985/9985 [00:02&lt;00:00, 3977.47 examples/s]"
     }
    },
    "67da6c4260574869aa24c3cbc1bc1654": {
@@ -4778,7 +4778,7 @@
      "description_tooltip": null,
      "disabled": false,
      "layout": "IPY_MODEL_2e257c8be2da40b4bb67a9e4ab6811f3",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_56e3768bef5a4b9db4168c5c17f509c2",
      "value": ""
     }
@@ -4823,9 +4823,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_41f3b32c2f6b4034ae7a3b9124e28bc7",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_a10d0a76010f4e508c65a9b69ebc5156",
-      "value": "Tokenizing\u2007Prompts\u2007(num_proc=2):\u2007100%"
+      "value": "Tokenizing Prompts (num_proc=2): 100%"
     }
    },
    "704f2f5a9b1c49d5a75a0025a5dda11b": {
@@ -5071,9 +5071,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_93a44a11aa4846fa8efc6c1413ef1627",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_a55060adc3564407ac81ad7297d34aaa",
-      "value": "train.jsonl:\u2007100%"
+      "value": "train.jsonl: 100%"
     }
    },
    "7be6f04c284e4326bb4ff3d301e7b3c6": {
@@ -5138,9 +5138,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_7fd44cf9ca6e4726bfd7ac21846d6a14",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_366a343b62fa47d8985a3bd464d99f9e",
-      "value": "config.json:\u2007100%"
+      "value": "config.json: 100%"
     }
    },
    "7cd0b85ebd204b7aba908417811ce4e0": {
@@ -5339,9 +5339,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_67da6c4260574869aa24c3cbc1bc1654",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_94b9088614464f60a203de39dbcae853",
-      "value": "\u20078/8\u2007[01:47&lt;00:00,\u200711.64s/it]"
+      "value": " 8/8 [01:47&lt;00:00, 11.64s/it]"
     }
    },
    "823f1c78f15043e38bbd4dca3932a86a": {
@@ -5488,7 +5488,7 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_8640ac440fbc4644b9a3af7ba3ae7183",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_5cea7996f02040b187ece0bb2d6a8d1f",
      "value": "<center> <img\nsrc=https://huggingface.co/front/assets/huggingface_logo-noborder.svg\nalt='Hugging Face'> <br> Copy a token from <a\nhref=\"https://huggingface.co/settings/tokens\" target=\"_blank\">your Hugging Face\ntokens page</a> and paste it below. <br> Immediately click login after copying\nyour token or it might be stored in plain text in this notebook file. </center>"
     }
@@ -5509,9 +5509,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_ef223e8504b64e3592589880326aaf41",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_598da69727bd4fb8b1caf465ac736d7a",
-      "value": "\u20071.67M/1.67M\u2007[00:00&lt;00:00,\u200719.0MB/s]"
+      "value": " 1.67M/1.67M [00:00&lt;00:00, 19.0MB/s]"
     }
    },
    "897b77a56c09479bb11d7f2a30997e55": {
@@ -5717,9 +5717,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_37de928300e34184881039378bd75e7f",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_0e936d9dbf9c4fdd86bbfe9730dedc47",
-      "value": "\u20073.96G/3.96G\u2007[00:13&lt;00:00,\u2007273MB/s]"
+      "value": " 3.96G/3.96G [00:13&lt;00:00, 273MB/s]"
     }
    },
    "936d04b5fe1b4c63bf0b080e423d051b": {
@@ -6050,9 +6050,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_d955dcaa0e944e719f3a06139dd54a03",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_d3de2662c7964f1ba96e58da382af720",
-      "value": "merges.txt:\u2007100%"
+      "value": "merges.txt: 100%"
     }
    },
    "9cd5211b5d8b457aa0002f1d17b80028": {
@@ -6071,9 +6071,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_6932489232ec4ab18a160b1e7fbcdfe1",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_4540927d98f54466b434ba4c0edf045d",
-      "value": "model-00007-of-00008.safetensors:\u2007100%"
+      "value": "model-00007-of-00008.safetensors: 100%"
     }
    },
    "9d4897eefb5f48259ffb2d23e332f752": {
@@ -6303,9 +6303,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_3aaecbf540f54a2db9ab0931e3b1fe57",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_9e333ed3b5014069ac1dd969255dd591",
-      "value": "\u2007239/239\u2007[00:00&lt;00:00,\u200730.9kB/s]"
+      "value": " 239/239 [00:00&lt;00:00, 30.9kB/s]"
     }
    },
    "a20927bf5f2c41f58c1e31ac858ab36c": {
@@ -6324,9 +6324,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_1811cda0644e4190a9469d1774435d82",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_35c811d2ae8e43f3b5cecbdd3cfa857f",
-      "value": "tokenizer.json:\u2007100%"
+      "value": "tokenizer.json: 100%"
     }
    },
    "a3a945817f684328b34651fe052393ec": {
@@ -6360,9 +6360,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_ed5ca967ad5342929e578ac6aa4dc4c0",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_af401d117d5047629d3a6e2361757b62",
-      "value": "model-00001-of-00008.safetensors:\u2007100%"
+      "value": "model-00001-of-00008.safetensors: 100%"
     }
    },
    "a4e5789584564049b83df7c6c54a3e08": {
@@ -6494,9 +6494,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_fa1282ccc7544e4f818e2f03ccffe4a5",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_bbbf575d2a4b4c6ea8389be79b2a6039",
-      "value": "model.safetensors.index.json:\u2007100%"
+      "value": "model.safetensors.index.json: 100%"
     }
    },
    "ab93eabd7cea4b94b4b7a387f101e8a1": {
@@ -6582,9 +6582,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_62e302ebdad64aada0ffe64ae1c873f3",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_bd1b0dfed6d34d16af33a4a58330f5ec",
-      "value": "Saving\u2007the\u2007dataset\u2007(1/1\u2007shards):\u2007100%"
+      "value": "Saving the dataset (1/1 shards): 100%"
     }
    },
    "ad7599de524549c48bf2d3124ad4b299": {
@@ -6967,9 +6967,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_2b3a2659b12244bd8548320320016dbf",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_0cd7efffbb3c4c4b972e63749f61ab97",
-      "value": "Generating\u2007train\u2007split:\u2007"
+      "value": "Generating train split: "
     }
    },
    "b87c84de30e84b3abf4871461fb9cbd3": {
@@ -7085,9 +7085,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_0f480e3a0b0a45d2a2d2dec3cad923f3",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_fcb30372e7404c5d8a1ad4df91e6c7b2",
-      "value": "\u20071.91G/1.91G\u2007[00:05&lt;00:00,\u2007444MB/s]"
+      "value": " 1.91G/1.91G [00:05&lt;00:00, 444MB/s]"
     }
    },
    "bd1b0dfed6d34d16af33a4a58330f5ec": {
@@ -7325,9 +7325,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_158c8b85dbf34de6a94b4e35e2fc7d5a",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_0b4c9753a7cb4354b8e5f187e6e1ad7c",
-      "value": "\u20073.96G/3.96G\u2007[00:15&lt;00:00,\u2007564MB/s]"
+      "value": " 3.96G/3.96G [00:15&lt;00:00, 564MB/s]"
     }
    },
    "c0991cf63ee6458b96e9a75e7a88b61a": {
@@ -7346,9 +7346,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_ed28e2e0410d4e0b855467e798e53d66",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_d93f134f802b4b69b575bdaf07dbd27c",
-      "value": "tokenizer_config.json:\u2007100%"
+      "value": "tokenizer_config.json: 100%"
     }
    },
    "c12ea43372ac4d57bb9605f1a429b397": {
@@ -7581,9 +7581,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_8bc9d8ba866c442b9118d9630009939c",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_9f56a2d9979c4bd8928c644c22c3ecdf",
-      "value": "model-00003-of-00008.safetensors:\u2007100%"
+      "value": "model-00003-of-00008.safetensors: 100%"
     }
    },
    "c6164e05a1914ae48083db9ad7f4ef7c": {
@@ -7694,9 +7694,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_e40d1c1ac9494b3bade9858324e7ffdf",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_d65b6b060d9845779299491ac5599c31",
-      "value": "\u20079985/9985\u2007[01:04&lt;00:00,\u2007189.08\u2007examples/s]"
+      "value": " 9985/9985 [01:04&lt;00:00, 189.08 examples/s]"
     }
    },
    "c7433acd3c4841e6958ae8f7e87b1808": {
@@ -7737,9 +7737,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_0077aedc3d174560bce924ee89e9c006",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_00321cce58884f6f9b3855a21fcd9187",
-      "value": "Add\u2007position_id\u2007column\u2007(Sample\u2007Packing)\u2007(num_proc=2):\u2007100%"
+      "value": "Add position_id column (Sample Packing) (num_proc=2): 100%"
     }
    },
    "ca65e32eb52f48c09a84b33cb18f22cd": {
@@ -8162,9 +8162,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_63580b6fb30642479fe3000915bf551a",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_8f726dbfb45d4528afa33e36a6313267",
-      "value": "\u200727.3M/27.3M\u2007[00:00&lt;00:00,\u200731.0MB/s]"
+      "value": " 27.3M/27.3M [00:00&lt;00:00, 31.0MB/s]"
     }
    },
    "d43c6df07ddb466587807d6dbe1ff614": {
@@ -8183,9 +8183,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_8c4d4fc5a30f4e7cb3be53fe2adda33d",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_e90658f4bcb642baa78426012f863152",
-      "value": "model-00004-of-00008.safetensors:\u2007100%"
+      "value": "model-00004-of-00008.safetensors: 100%"
     }
    },
    "d65b6b060d9845779299491ac5599c31": {
@@ -8474,9 +8474,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_34cf3df51fbc41cabfdbba153c007f0e",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_ac764024cf1c4e08ba7749afd2cd20ac",
-      "value": "vocab.json:\u2007100%"
+      "value": "vocab.json: 100%"
     }
    },
    "dfd2a2649b8341ef913207526708aff1": {
@@ -8669,9 +8669,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_c6164e05a1914ae48083db9ad7f4ef7c",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_813621384dc748b0ad06775e22761c0b",
-      "value": "\u20079985/9985\u2007[00:03&lt;00:00,\u20073622.89\u2007examples/s]"
+      "value": " 9985/9985 [00:03&lt;00:00, 3622.89 examples/s]"
     }
    },
    "e400cbf14bcc446a9d33b210cd93550b": {
@@ -9065,9 +9065,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_fba7aa824b38467ab3061b226114cdec",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_f3075dccbd2747b4a7913b66f44f2596",
-      "value": "\u20073.96G/3.96G\u2007[00:13&lt;00:00,\u2007398MB/s]"
+      "value": " 3.96G/3.96G [00:13&lt;00:00, 398MB/s]"
     }
    },
    "ec030fc3c346426f9abc3a89892258d3": {
@@ -9110,9 +9110,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_936d04b5fe1b4c63bf0b080e423d051b",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_f1cef8e8dc2646fb9fd09f3b09081074",
-      "value": "\u200736.5k/36.5k\u2007[00:00&lt;00:00,\u20074.32MB/s]"
+      "value": " 36.5k/36.5k [00:00&lt;00:00, 4.32MB/s]"
     }
    },
    "ed28e2e0410d4e0b855467e798e53d66": {
@@ -9422,9 +9422,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_735d4f225b24414294fc1b213c61223c",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_5e5e15b0569b474c9620083b3ec6af55",
-      "value": "generation_config.json:\u2007100%"
+      "value": "generation_config.json: 100%"
     }
    },
    "f4667818b9d34a09891cd727a429a610": {
@@ -9443,9 +9443,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_4b27c267393640f28f6eae0875bd2ed9",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_9858cb74a09748a39e8149baac96702c",
-      "value": "\u20073.96G/3.96G\u2007[00:11&lt;00:00,\u2007457MB/s]"
+      "value": " 3.96G/3.96G [00:11&lt;00:00, 457MB/s]"
     }
    },
    "f4a1795dc7514a718f478245f521f0ba": {
@@ -9830,9 +9830,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_d1f9b10c130542f094c8fd3d1e23b5e9",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_e575d87a7efe4ec7b1efde489839d4a6",
-      "value": "model-00006-of-00008.safetensors:\u2007100%"
+      "value": "model-00006-of-00008.safetensors: 100%"
     }
    },
    "fe18bba7f3fb4c31bf840541f36b3425": {
@@ -9873,9 +9873,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_e5a82df528bb4e408797a3b6c2758f4a",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_f113ebd8c1c34806bea4dd7ed3035173",
-      "value": "\u20079985/9985\u2007[00:00&lt;00:00,\u200744264.88\u2007examples/s]"
+      "value": " 9985/9985 [00:00&lt;00:00, 44264.88 examples/s]"
     }
    },
    "fea1b70fb46745feb5111b3929175b5d": {
@@ -9931,9 +9931,9 @@
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_ab93eabd7cea4b94b4b7a387f101e8a1",
-      "placeholder": "\u200b",
+      "placeholder": "",
      "style": "IPY_MODEL_704f2f5a9b1c49d5a75a0025a5dda11b",
-      "value": "\u20073.96G/3.96G\u2007[00:12&lt;00:00,\u2007656MB/s]"
+      "value": " 3.96G/3.96G [00:12&lt;00:00, 656MB/s]"
     }
    }
   }
--- a/examples/devstral/README.md
+++ b/examples/devstral/README.md
@@ -16,13 +16,8 @@ Thanks to the team at MistralAI for giving us early access to prepare for this r

 ```bash
 # Ensure you have Pytorch installed (Pytorch 2.6.0 min)
-# Option A: manage dependencies in your project
-uv add 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
-
-# Option B: quick install
-uv pip install 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
 ```

 2. Install [Cut Cross Entropy](https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy) to reduce training VRAM usage
--- a/examples/gemma3n/README.md
+++ b/examples/gemma3n/README.md
@@ -10,22 +10,17 @@ Gemma-3n is a family of multimodal models from Google found on [HuggingFace](htt

 ```bash
 # Ensure you have Pytorch installed (Pytorch 2.6.0 min)
-# Option A: manage dependencies in your project
-uv add 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
-
-# Option B: quick install
-uv pip install 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
 ```

 2. In addition to Axolotl's requirements, Gemma-3n requires:

 ```bash
-uv pip install timm==1.0.17
+pip3 install timm==1.0.17

 # for loading audio data
-uv pip install librosa==0.11.0
+pip3 install librosa==0.11.0
 ```

 3. Download sample dataset files
--- a/examples/gpt-oss/README.md
+++ b/examples/gpt-oss/README.md
@@ -12,13 +12,8 @@ This guide shows how to fine-tune it with Axolotl with multi-turn conversations

 ```bash
 # Ensure you have Pytorch installed (Pytorch 2.6.0 min)
-# Option A: manage dependencies in your project
-uv add 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
-
-# Option B: quick install
-uv pip install 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
 ```

 2. Choose one of the following configs below for training the 20B model. (for 120B, see [below](#training-120b))
@@ -80,7 +75,7 @@ for more information about using a special vllm-openai docker image for inferenc
 Optionally, vLLM can be installed from nightly:

 ```bash
-uv pip install --no-build-isolation --pre -U vllm --extra-index-url https://wheels.vllm.ai/nightly
+pip install --no-build-isolation --pre -U vllm --extra-index-url https://wheels.vllm.ai/nightly
 ```
 and the vLLM server can be started with the following command (modify `--tensor-parallel-size 8` to match your environment):
 ```bash
--- a/examples/hunyuan/README.md
+++ b/examples/hunyuan/README.md
@@ -13,8 +13,8 @@ Tencent released a family of opensource models called HunYuan with varying param
 git clone https://github.com/axolotl-ai-cloud/axolotl.git
 cd axolotl

-uv sync
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation -e '.[flash-attn]'

 # Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
 python scripts/cutcrossentropy_install.py | sh
--- a/examples/llama-2/qlora-fsdp.yml
+++ b/examples/llama-2/qlora-fsdp.yml
@@ -66,7 +66,6 @@ fsdp_config:
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_state_dict_type: FULL_STATE_DICT
-  # fsdp_cpu_offload_pin_memory: false  # uncomment to enable swap memory usage when RAM is insufficient
 special_tokens:

 # save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/magistral/README.md
+++ b/examples/magistral/README.md
@@ -13,14 +13,9 @@ Thanks to the team at MistralAI for giving us early access to prepare for these
    Here is an example of how to install from pip:

 ```bash
-# Ensure you have PyTorch installed (PyTorch 2.6.0 min)
-# Option A: manage dependencies in your project
-uv add 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
-
-# Option B: quick install
-uv pip install 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
+# Ensure you have Pytorch installed (Pytorch 2.6.0 min)
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
 ```

 2. Install [Cut Cross Entropy](https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy) to reduce training VRAM usage
--- a/examples/qwen3-next/README.md
+++ b/examples/qwen3-next/README.md
@@ -15,8 +15,8 @@ This guide shows how to fine-tune it with Axolotl with multi-turn conversations
 git clone https://github.com/axolotl-ai-cloud/axolotl.git
 cd axolotl

-uv sync
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation -e '.[flash-attn]'

 # Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
 python scripts/cutcrossentropy_install.py | sh
@@ -24,12 +24,12 @@ python scripts/cutcrossentropy_install.py | sh

 2. Install Qwen3-Next transformers commit
 ```bash
-uv pip uninstall -y transformers && uv pip install "git+https://github.com/huggingface/transformers.git@b9282355bea846b54ed850a066901496b19da654"
+pip3 uninstall -y transformers && pip3 install "git+https://github.com/huggingface/transformers.git@b9282355bea846b54ed850a066901496b19da654"
 ```

 3. Install FLA for improved performance
 ```bash
-uv pip uninstall -y causal-conv1d && uv pip install flash-linear-attention==0.3.2
+pip3 uninstall -y causal-conv1d && pip3 install flash-linear-attention==0.3.2
 ```

 4. Run the finetuning example:
--- a/examples/seed-oss/README.md
+++ b/examples/seed-oss/README.md
@@ -15,8 +15,8 @@ This guide shows how to fine-tune it with Axolotl with multi-turn conversations
 git clone https://github.com/axolotl-ai-cloud/axolotl.git
 cd axolotl

-uv sync --extra deepspeed
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation -e '.[flash-attn]'

 # Install Cut Cross Entropy
 python scripts/cutcrossentropy_install.py | sh
--- a/examples/smolvlm2/README.md
+++ b/examples/smolvlm2/README.md
@@ -13,19 +13,14 @@ This guide shows how to fine-tune SmolVLM2 models with Axolotl.
    Here is an example of how to install from pip:
    ```bash
    # Ensure you have a compatible version of Pytorch installed
-    # Option A: manage dependencies in your project
-    uv add 'axolotl>=0.12.0'
-    uv pip install flash-attn --no-build-isolation
-
-    # Option B: quick install
-    uv pip install 'axolotl>=0.12.0'
-    uv pip install flash-attn --no-build-isolation
+    pip3 install packaging setuptools wheel ninja
+    pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
    ```

 2. Install an extra dependency:

    ```bash
-    uv pip install num2words==0.5.14
+    pip3 install num2words==0.5.14
    ```

 3.  Run the finetuning example:
--- a/examples/voxtral/README.md
+++ b/examples/voxtral/README.md
@@ -12,21 +12,16 @@ Thanks to the team at MistralAI for giving us early access to prepare for this r

 ```bash
 # Ensure you have Pytorch installed (Pytorch 2.6.0 min)
-# Option A: manage dependencies in your project
-uv add 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
-
-# Option B: quick install
-uv pip install 'axolotl>=0.12.0'
-uv pip install flash-attn --no-build-isolation
+pip3 install packaging==23.2 setuptools==75.8.0 wheel ninja
+pip3 install --no-build-isolation 'axolotl[flash-attn]>=0.12.0'
 ```

 2. Please install the below.

 ```bash
 # audio
-uv pip install librosa==0.11.0
-uv pip install 'mistral_common[audio]==1.8.3'
+pip3 install librosa==0.11.0
+pip3 install 'mistral_common[audio]==1.8.3'

 # Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
 python scripts/cutcrossentropy_install.py | sh
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,131 +1,14 @@
 [build-system]
-requires = ["setuptools>=64", "wheel", "setuptools_scm>=8"]
+requires = ["setuptools>=64", "wheel", "setuptools_scm>=8", "packaging==23.2"]
 build-backend = "setuptools.build_meta"

 [project]
 name = "axolotl"
-dynamic = ["version"]
+dynamic = ["version", "dependencies", "optional-dependencies"]
 description = "LLM Trainer"
 readme = "README.md"
-requires-python = ">=3.10,<3.13"
-license = {text = "Apache-2.0"}
-authors = [
-    {name = "Axolotl AI"},
-]
-maintainers = [
-    {name = "Axolotl AI"},
-]
-classifiers = [
-    "Development Status :: 4 - Beta",
-    "License :: OSI Approved :: Apache Software License",
-    "Programming Language :: Python :: 3",
-    "Programming Language :: Python :: 3.10",
-    "Programming Language :: Python :: 3.11",
-    "Programming Language :: Python :: 3.12",
-]
-
-dependencies = [
-    "torch>=2.6.0",
-    "packaging>=23.2",
-    "huggingface_hub>=0.33.0",
-    "peft==0.17.0",
-    "transformers==4.56.1",
-    "tokenizers>=0.21.1",
-    "accelerate==1.10.1",
-    "datasets==4.0.0",
-    "trl==0.23.0",
-    "hf_xet==1.1.5",
-    "kernels==0.9.0",
-    "trackio",
-    "optimum==1.16.2",
-    "hf_transfer",
-    "sentencepiece",
-    "gradio==5.41.1",
-    "modal==1.0.2",
-    "pydantic>=2.10.6",
-    "addict",
-    "fire",
-    "PyYAML>=6.0",
-    "requests",
-    "wandb",
-    "einops",
-    "colorama",
-    "numba",
-    "numpy>=1.24.4,<3.0",
-    "evaluate==0.4.1",
-    "scipy",
-    "scikit-learn>=1.7.0",
-    "nvidia-ml-py==12.560.30",
-    "art",
-    "tensorboard",
-    "python-dotenv==1.0.1",
-    "s3fs>=2024.5.0",
-    "gcsfs>=2024.5.0",
-    "adlfs>=2024.5.0",
-    "ocifs==1.3.2",
-    "zstandard>=0.23.0",
-    "fastcore",
-    "lm_eval==0.4.7",
-    "langdetect==1.0.9",
-    "immutabledict==4.2.0",
-    "antlr4-python3-runtime==4.13.2",
-    "schedulefree==1.4.1",
-    "mistral-common==1.8.5",
-
-    # Axolotl contribs
-    "axolotl-contribs-lgpl @ git+https://github.com/axolotl-ai-cloud/axolotl-contribs-lgpl.git@numpy",
-    "axolotl-contribs-mit==0.0.5",
-
-    # Platform-specific dependencies (Linux by default, excluded on macOS)
-    "triton>=3.0.0 ; sys_platform != 'darwin'",
-    "xformers>=0.0.28 ; sys_platform != 'darwin'",
-    "autoawq==0.2.7.post3 ; sys_platform != 'darwin'",
-    "liger-kernel==0.6.1 ; sys_platform != 'darwin'",
-    "torchao==0.13.0 ; sys_platform != 'darwin'",
-    "bitsandbytes==0.47.0 ; sys_platform != 'darwin'",
-    "deepspeed>=0.17.5 ; sys_platform != 'darwin'",
-    "deepspeed-kernels ; sys_platform != 'darwin'",
-]
-
-[project.optional-dependencies]
-ring-flash-attn = [
-    "ring-flash-attn>=0.1.7",
-    "yunchang==0.6.0",
-]
-mamba-ssm = ["mamba-ssm>=2.2.0", "causal_conv1d>=1.4.0",]
-gptqmodel = ["gptqmodel>=4.0.0"]
-mlflow = ["mlflow"]
-galore = ["galore_torch"]
-apollo = ["apollo-torch"]
-optimizers = [
-    "galore_torch",
-    "apollo-torch",
-    "lomo-optim==0.1.1",
-    "torch-optimi==0.2.1",
-    "came_pytorch==0.1.3",
-]
-ray = ["ray[train]"]
-vllm = ["vllm>=0.10.0"]
-llmcompressor = ["llmcompressor>=0.5.1"]
-fbgemm-gpu = ["fbgemm-gpu-genai>=1.2.0"]
-dev = [
-    "pytest",
-    "pytest-cov",
-    "pytest-retry",
-    "pytest-sugar",
-    "pytest-xdist",
-    "codecov",
-    "codecov-cli",
-    "tbparse",
-    "ruff",
-    "mypy",
-    "pre-commit",
-    "types-requests",
-    "quartodoc",
-    "jupyter",
-    "blobfile",
-    "tiktoken",
-]
+requires-python = ">=3.10"
+# license = "Apache-2.0"

 [project.scripts]
 axolotl = "axolotl.cli.main:main"
@@ -134,20 +17,15 @@ axolotl = "axolotl.cli.main:main"
 Homepage = "https://axolotl.ai/"
 Documentation = "https://docs.axolotl.ai/"
 Repository = "https://github.com/axolotl-ai-cloud/axolotl.git"
-Issues = "https://github.com/axolotl-ai-cloud/axolotl/issues"
-
-[tool.setuptools]
-package-dir = {"" = "src"}
-include-package-data = true
-
-[tool.setuptools.packages.find]
-where = ["src"]
-
-[tool.setuptools.package-data]
-"*" = ["*.yaml", "*.yml", "*.json"]

 [tool.setuptools_scm]
-write_to = "src/axolotl/_version.py"
+
+[tool.setuptools]
+py-modules = ["setuptools_axolotl_dynamic_dependencies"]
+include-package-data = true
+
+[tool.setuptools.cmdclass]
+build_py = "setuptools_axolotl_dynamic_dependencies.BuildPyCommand"

 [tool.ruff]
 line-length = 88
@@ -179,60 +57,3 @@ indent-style = "space"
 skip-magic-trailing-comma = false
 line-ending = "auto"
 docstring-code-format = false
-
-[tool.mypy]
-python_version = "3.11"
-warn_return_any = true
-warn_unused_configs = true
-ignore_missing_imports = true
-
-[tool.pytest.ini_options]
-testpaths = ["tests"]
-python_files = ["test_*.py", "*_test.py"]
-addopts = "-v --tb=short"
-
-# UV specific configuration
-[tool.uv]
-prerelease = "allow"
-default-groups = ["default"]
-conflicts = [
-    [
-        { group = "default" },
-        { extra = "vllm" },
-    ],
-]
-
-[dependency-groups]
-default = ["torch>=2.6.0"]
-dev = [
-    "pytest",
-    "pytest-cov",
-    "pytest-retry",
-    "pytest-sugar",
-    "pytest-xdist",
-    "codecov",
-    "codecov-cli",
-    "tbparse",
-    "ruff",
-    "mypy",
-    "pre-commit",
-    "types-requests",
-    "quartodoc",
-    "jupyter",
-    "blobfile",
-    "tiktoken",
-]
-
-[[tool.uv.index]]
-name = "autogptq"
-url = "https://huggingface.github.io/autogptq-index/whl/"
-
-[tool.uv.extra-build-dependencies]
-mamba-ssm = ["torch", "causal_conv1d"]
-gptqmodel = [
-    { requirement = "torch", match-runtime = true },
-]
-autoawq = ["torch"]
-triton = ["torch"]
-bitsandbytes = ["torch"]
-grpclib = ["wheel"]
--- a/requirements-dev.txt
+++ b/requirements-dev.txt
@@ -0,0 +1,8 @@
+black
+mypy
+pre-commit
+types-requests
+quartodoc
+jupyter
+blobfile
+tiktoken
--- a/requirements-tests.txt
+++ b/requirements-tests.txt
@@ -0,0 +1,8 @@
+codecov
+codecov-cli
+pytest
+pytest-cov
+pytest-retry
+pytest-sugar
+pytest-xdist
+tbparse
--- a/requirements.txt
+++ b/requirements.txt
@@ -0,0 +1,73 @@
+--extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/
+
+# START section of dependencies that don't install on Darwin/MacOS
+bitsandbytes==0.47.0
+triton>=3.0.0
+mamba-ssm==1.2.0.post1
+xformers>=0.0.23.post1
+autoawq==0.2.7.post3
+liger-kernel==0.6.1
+# END section
+
+packaging==23.2
+
+huggingface_hub>=0.33.0
+peft>=0.17.0
+transformers==4.56.1
+tokenizers>=0.21.1
+accelerate==1.10.1
+datasets==4.0.0
+deepspeed>=0.17.0
+trl==0.23.0
+hf_xet==1.1.5
+kernels==0.9.0
+trackio
+
+optimum==1.16.2
+hf_transfer
+sentencepiece
+gradio==5.41.1
+
+modal==1.0.2
+pydantic==2.10.6
+addict
+fire
+PyYAML>=6.0
+requests
+wandb
+einops
+colorama
+numba
+numpy>=1.24.4,<=2.0.1
+
+# qlora things
+evaluate==0.4.1
+scipy
+scikit-learn==1.4.2
+nvidia-ml-py==12.560.30
+art
+tensorboard
+python-dotenv==1.0.1
+
+# remote filesystems
+s3fs>=2024.5.0
+gcsfs>=2024.5.0
+adlfs>=2024.5.0
+ocifs==1.3.2
+
+zstandard==0.22.0
+fastcore
+
+# lm eval harness
+lm_eval==0.4.7
+langdetect==1.0.9
+immutabledict==4.2.0
+antlr4-python3-runtime==4.13.2
+
+torchao==0.13.0
+schedulefree==1.4.1
+
+axolotl-contribs-lgpl==0.0.6
+axolotl-contribs-mit==0.0.5
+
+mistral-common==1.8.5
--- a/scripts/init.py
+++ b/scripts/init.py
@@ -0,0 +1 @@
+"""Utility scripts package."""
--- a/scripts/benchmarks/init.py
+++ b/scripts/benchmarks/init.py
@@ -0,0 +1,5 @@
+"""Benchmark helpers."""
+
+from .deepseek_v3_moe import ACCURACY_TOLERANCE, DTYPE_MAP, benchmark_deepseek_v3
+
+__all__ = ["benchmark_deepseek_v3", "DTYPE_MAP", "ACCURACY_TOLERANCE"]
--- a/scripts/benchmarks/build_deepseek_v3_8b.py
+++ b/scripts/benchmarks/build_deepseek_v3_8b.py
@@ -0,0 +1,100 @@
+#!/usr/bin/env python3
+"""Instantiate a ~8.3B DeepSeek-V3 MoE model with random weights.
+
+Run this on a GPU-equipped machine (e.g. 1× NVL H100) so the dense
+initialization completes quickly:
+
+    python scripts/benchmarks/build_deepseek_v3_8b.py --output deepseek-v3-8b-moe
+"""
+
+from __future__ import annotations
+
+import argparse
+from pathlib import Path
+
+import torch
+from transformers import DeepseekV3Config, DeepseekV3ForCausalLM
+
+DTYPE_MAP = {
+    "float32": torch.float32,
+    "bfloat16": torch.bfloat16,
+    "float16": torch.float16,
+}
+
+
+def build_config() -> DeepseekV3Config:
+    """Return a DeepSeek V3 configuration totaling ~8.3B parameters."""
+
+    return DeepseekV3Config(
+        vocab_size=32_000,
+        hidden_size=3_072,
+        intermediate_size=8_192,
+        moe_intermediate_size=2_560,
+        num_hidden_layers=20,
+        num_attention_heads=24,
+        num_key_value_heads=24,
+        n_routed_experts=18,
+        num_experts_per_tok=4,
+        n_group=6,
+        topk_group=4,
+        kv_lora_rank=192,
+        q_lora_rank=384,
+        max_position_embeddings=2_048,
+        rope_theta=10_000.0,
+        rope_interleave=True,
+        hidden_act="silu",
+        initializer_range=0.02,
+        attention_dropout=0.0,
+        attention_bias=False,
+        n_shared_experts=1,
+        routed_scaling_factor=2.5,
+        norm_topk_prob=True,
+    )
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        "--output",
+        type=Path,
+        required=True,
+        help="Directory to save the generated model",
+    )
+    parser.add_argument(
+        "--dtype",
+        default="bfloat16",
+        choices=DTYPE_MAP.keys(),
+        help="Storage dtype for the checkpoint",
+    )
+    parser.add_argument(
+        "--seed",
+        type=int,
+        default=0,
+        help="Torch RNG seed for reproducibility",
+    )
+    return parser.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    torch.manual_seed(args.seed)
+
+    output_dir = args.output
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    config = build_config()
+    model = DeepseekV3ForCausalLM(config)
+
+    dtype = DTYPE_MAP[args.dtype]
+    model.to(dtype=dtype)
+
+    param_count = sum(p.numel() for p in model.parameters())
+    print(f"Initialized DeepSeek-V3 MoE with {param_count / 1e9:.3f}B parameters")
+
+    model.save_pretrained(output_dir, safe_serialization=True)
+    config.save_pretrained(output_dir)
+    print(f"Saved model and config to {output_dir.resolve()}")
+
+
+if __name__ == "__main__":
+    main()
--- a/scripts/benchmarks/deepseek_v3_group_gemm_table.py
+++ b/scripts/benchmarks/deepseek_v3_group_gemm_table.py
@@ -0,0 +1,190 @@
+#!/usr/bin/env python
+"""Reproduce TorchTitan CG GEMM timings for selected problem sizes."""
+
+from __future__ import annotations
+
+import argparse
+import sys
+import time
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Iterable
+
+import torch
+
+CURRENT_DIR = Path(__file__).resolve().parent
+for candidate in [CURRENT_DIR, *CURRENT_DIR.parents]:
+    repo_root = candidate / "axolotl"
+    if repo_root.exists():
+        if str(repo_root) not in sys.path:
+            sys.path.insert(0, str(repo_root))
+        break
+else:
+    raise SystemExit("Unable to locate axolotl repository root for imports")
+
+from axolotl.kernels.moe import (
+    cg_grouped_gemm_forward,
+    cg_grouped_gemm_forward_dynamic,
+)
+
+
+@dataclass
+class Scenario:
+    num_groups: int
+    m: int
+    n: int
+    k: int
+
+
+SCENARIOS: tuple[Scenario, ...] = (
+    Scenario(num_groups=4, m=8192, n=4096, k=7168),
+    Scenario(num_groups=4, m=8192, n=7168, k=2048),
+    Scenario(num_groups=8, m=4096, n=4096, k=7168),
+    Scenario(num_groups=8, m=4096, n=7168, k=2048),
+)
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        "--device", default="cuda", choices=["cuda"], help="Execution device"
+    )
+    parser.add_argument(
+        "--dtype",
+        default="bf16",
+        choices=["bf16", "fp16", "fp32"],
+        help="Computation dtype",
+    )
+    parser.add_argument("--warmup", type=int, default=5, help="Warmup iterations")
+    parser.add_argument("--iters", type=int, default=20, help="Benchmark iterations")
+    parser.add_argument("--seed", type=int, default=0, help="Random seed")
+    parser.add_argument(
+        "--group-size",
+        type=int,
+        default=128,
+        help="GROUP_SIZE_M expected by the kernel",
+    )
+    return parser.parse_args()
+
+
+def pick_dtype(name: str) -> torch.dtype:
+    return {
+        "bf16": torch.bfloat16,
+        "fp16": torch.float16,
+        "fp32": torch.float32,
+    }[name]
+
+
+def make_indices(
+    num_groups: int, group_size: int, device: torch.device
+) -> torch.Tensor:
+    indices = torch.arange(num_groups, device=device, dtype=torch.int32)
+    return indices.repeat_interleave(group_size)
+
+
+def timed_call(fn, *args, warmup: int, iters: int) -> float:
+    for _ in range(warmup):
+        fn(*args)
+    torch.cuda.synchronize()
+    start = time.perf_counter()
+    for _ in range(iters):
+        fn(*args)
+    torch.cuda.synchronize()
+    return (time.perf_counter() - start) * 1000.0 / iters
+
+
+def run_scenario(
+    scenario: Scenario,
+    *,
+    dtype: torch.dtype,
+    device: torch.device,
+    warmup: int,
+    iters: int,
+    group_size_m: int,
+) -> dict:
+    if scenario.m % scenario.num_groups != 0:
+        raise ValueError(
+            f"M ({scenario.m}) not divisible by groups ({scenario.num_groups})"
+        )
+    group_size = scenario.m // scenario.num_groups
+    if group_size % group_size_m != 0:
+        raise ValueError(
+            f"Group size {group_size} must be a multiple of GROUP_SIZE_M ({group_size_m}) for the Triton kernel"
+        )
+
+    inputs = torch.randn(scenario.m, scenario.k, device=device, dtype=dtype)
+    weights = torch.randn(
+        scenario.num_groups, scenario.n, scenario.k, device=device, dtype=dtype
+    )
+    indices = make_indices(scenario.num_groups, group_size, device)
+
+    def persistent():
+        return cg_grouped_gemm_forward(inputs, weights, indices, group_size_m)
+
+    def baseline():
+        return cg_grouped_gemm_forward_dynamic(inputs, weights, indices, group_size_m)
+
+    persistent_ms = timed_call(persistent, warmup=warmup, iters=iters)
+    baseline_ms = timed_call(baseline, warmup=warmup, iters=iters)
+
+    return {
+        "scenario": scenario,
+        "persistent_ms": persistent_ms,
+        "baseline_ms": baseline_ms,
+        "speedup": baseline_ms / persistent_ms if persistent_ms > 0 else float("nan"),
+    }
+
+
+def main() -> None:  # pragma: no cover - utility script
+    args = parse_args()
+    torch.manual_seed(args.seed)
+
+    if args.device != "cuda" or not torch.cuda.is_available():
+        raise SystemExit("CUDA device required for this benchmark")
+
+    dtype = pick_dtype(args.dtype)
+    device = torch.device(args.device)
+
+    print(
+        f"device={device} dtype={dtype} warmup={args.warmup} iters={args.iters} group_size={args.group_size}"
+    )
+    print(
+        f"{'groups':>7} {'m':>7} {'n':>7} {'k':>7} {'persistent':>12} {'baseline':>12} {'speedup':>8}"
+    )
+    for result in run_all(
+        SCENARIOS,
+        dtype=dtype,
+        device=device,
+        warmup=args.warmup,
+        iters=args.iters,
+        group_size_m=args.group_size,
+    ):
+        scen = result["scenario"]
+        print(
+            f"{scen.num_groups:>7} {scen.m:>7} {scen.n:>7} {scen.k:>7}"
+            f" {result['persistent_ms']:>11.3f} ms {result['baseline_ms']:>11.3f} ms {result['speedup']:>7.2f}x"
+        )
+
+
+def run_all(
+    scenarios: Iterable[Scenario],
+    *,
+    dtype: torch.dtype,
+    device: torch.device,
+    warmup: int,
+    iters: int,
+    group_size_m: int,
+) -> Iterable[dict]:
+    for scenario in scenarios:
+        yield run_scenario(
+            scenario,
+            dtype=dtype,
+            device=device,
+            warmup=warmup,
+            iters=iters,
+            group_size_m=group_size_m,
+        )
+
+
+if __name__ == "__main__":
+    main()
--- a/scripts/benchmarks/deepseek_v3_moe.py
+++ b/scripts/benchmarks/deepseek_v3_moe.py
@@ -0,0 +1,301 @@
+#!/usr/bin/env python
+# mypy: ignore-errors
+"""Microbenchmark for DeepSeek V3 MoE block comparing baseline vs Triton CG kernels."""
+
+from __future__ import annotations
+
+import argparse
+import sys
+import time
+from pathlib import Path
+from types import MethodType
+
+import torch
+
+try:
+    from transformers.models.deepseek_v3.configuration_deepseek_v3 import (
+        DeepseekV3Config,
+    )
+    from transformers.models.deepseek_v3.modeling_deepseek_v3 import DeepseekV3MoE
+except ImportError as exc:  # pragma: no cover - utility script
+    raise SystemExit(
+        "Transformers with DeepSeek-V3 support must be available in PYTHONPATH"
+    ) from exc
+
+CURRENT_DIR = Path(__file__).resolve().parent
+for candidate in [CURRENT_DIR, *CURRENT_DIR.parents]:
+    repo_root = candidate / "axolotl"
+    if repo_root.exists():
+        if str(repo_root) not in sys.path:
+            sys.path.insert(0, str(repo_root))
+        break
+else:  # pragma: no cover - defensive guard
+    raise SystemExit("Unable to locate axolotl repository root for imports")
+
+from axolotl.monkeypatch.deepseek_v3 import patch_deepseek_v3_moe  # noqa: E402
+
+ACCURACY_TOLERANCE = 5e-3
+
+DTYPE_MAP = {
+    "bf16": torch.bfloat16,
+    "fp16": torch.float16,
+    "fp32": torch.float32,
+}
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("--batch", type=int, default=8, help="batch size")
+    parser.add_argument("--seq-len", type=int, default=2048, help="sequence length")
+    parser.add_argument("--hidden-size", type=int, default=4096, help="MoE hidden size")
+    parser.add_argument(
+        "--moe-intermediate-size",
+        type=int,
+        default=8192,
+        help="MoE intermediate projection size",
+    )
+    parser.add_argument(
+        "--n-experts",
+        type=int,
+        default=256,
+        help="Number of routed experts",
+    )
+    parser.add_argument(
+        "--top-k",
+        type=int,
+        default=8,
+        help="Number of experts per token",
+    )
+    parser.add_argument(
+        "--groups",
+        type=int,
+        default=8,
+        help="Router groups (must divide n-experts)",
+    )
+    parser.add_argument(
+        "--dtype",
+        choices=DTYPE_MAP.keys(),
+        default="bf16",
+        help="Computation dtype",
+    )
+    parser.add_argument(
+        "--device",
+        default="auto",
+        choices=["auto", "cpu", "cuda"],
+        help="Execution device",
+    )
+    parser.add_argument("--warmup", type=int, default=5, help="Warmup iterations")
+    parser.add_argument("--iters", type=int, default=25, help="Benchmark iterations")
+    parser.add_argument("--seed", type=int, default=0, help="Random seed")
+    parser.add_argument(
+        "--uniform-routing",
+        action="store_true",
+        help="Override router to distribute tokens evenly across experts",
+    )
+    parser.add_argument(
+        "--group-size",
+        type=int,
+        default=128,
+        help="GROUP_SIZE_M used by the Triton kernel",
+    )
+    parser.add_argument(
+        "--backend",
+        choices=["cg", "mg"],
+        default="mg",
+        help="MoE kernel backend to benchmark",
+    )
+    return parser.parse_args()
+
+
+def resolve_device(requested: str) -> torch.device:
+    if requested == "auto":
+        return torch.device("cuda" if torch.cuda.is_available() else "cpu")
+    return torch.device(requested)
+
+
+def build_module(args: argparse.Namespace) -> DeepseekV3MoE:
+    config = DeepseekV3Config(
+        hidden_size=args.hidden_size,
+        intermediate_size=args.moe_intermediate_size,
+        moe_intermediate_size=args.moe_intermediate_size,
+        n_routed_experts=args.n_experts,
+        num_experts_per_tok=args.top_k,
+        n_group=args.groups,
+        topk_group=max(1, min(args.groups, args.top_k)),
+        n_shared_experts=1,
+    )
+    module = DeepseekV3MoE(config)
+    module.eval()
+    return module
+
+
+@torch.no_grad()
+def timed_forward(
+    module: DeepseekV3MoE, inputs: torch.Tensor, iters: int, warmup: int
+) -> float:
+    for _ in range(warmup):
+        module(inputs)
+    if inputs.is_cuda:
+        torch.cuda.synchronize()
+    start = time.perf_counter()
+    for _ in range(iters):
+        module(inputs)
+    if inputs.is_cuda:
+        torch.cuda.synchronize()
+    elapsed = time.perf_counter() - start
+    return (elapsed / iters) * 1000.0
+
+
+def benchmark_deepseek_v3(args: argparse.Namespace) -> dict:
+    torch.manual_seed(args.seed)
+
+    device = resolve_device(args.device)
+    dtype = DTYPE_MAP[args.dtype]
+
+    if args.n_experts % args.groups != 0:
+        raise SystemExit("n-experts must be divisible by groups")
+    if args.top_k > args.n_experts:
+        raise SystemExit("top-k cannot exceed number of experts")
+
+    if device.type == "cuda" and not torch.cuda.is_available():
+        raise SystemExit("CUDA requested but not available")
+
+    baseline_module = build_module(args)
+    original_moe = getattr(
+        DeepseekV3MoE,
+        "_axolotl_triton_original_moe",
+        DeepseekV3MoE.moe,
+    )
+    baseline_module.moe = MethodType(original_moe, baseline_module)
+    state_dict = baseline_module.state_dict()
+
+    patch_deepseek_v3_moe(group_size_m=args.group_size, backend=args.backend)
+    patched_module = build_module(args)
+    patched_module.load_state_dict(state_dict)
+
+    baseline_module.to(device=device, dtype=dtype)
+    patched_module.to(device=device, dtype=dtype)
+
+    tokens = args.batch * args.seq_len
+    routed_tokens = tokens * args.top_k
+    avg_tokens_per_expert = routed_tokens / args.n_experts
+
+    inputs = torch.randn(
+        args.batch,
+        args.seq_len,
+        args.hidden_size,
+        device=device,
+        dtype=dtype,
+    )
+
+    with torch.no_grad():
+        flat_inputs = inputs.view(-1, args.hidden_size)
+        if args.uniform_routing:
+            total_assignments = flat_inputs.size(0) * args.top_k
+            base = total_assignments // args.n_experts
+            remainder = total_assignments % args.n_experts
+            counts = torch.full(
+                (args.n_experts,),
+                base,
+                dtype=torch.int64,
+                device=device,
+            )
+            if remainder:
+                counts[:remainder] += 1
+            assignments = torch.repeat_interleave(
+                torch.arange(args.n_experts, device=device), counts
+            )
+            assignments = assignments[torch.randperm(assignments.size(0))]
+            topk_idx = assignments.view(flat_inputs.size(0), args.top_k)
+        else:
+            topk_idx, _ = patched_module.gate(flat_inputs)
+
+        tokens_per_expert = torch.bincount(
+            topk_idx.reshape(-1), minlength=args.n_experts
+        )
+        min_tokens = int(tokens_per_expert.min().item())
+        max_tokens = int(tokens_per_expert.max().item())
+
+    if args.uniform_routing:
+        weights = torch.full(
+            topk_idx.shape,
+            1.0 / args.top_k,
+            device=device,
+            dtype=torch.float32,
+        )
+
+        def _uniform_gate(self, hidden_states):
+            flat = hidden_states.view(-1, hidden_states.shape[-1])
+            token_count = flat.shape[0]
+            return topk_idx[:token_count], weights[:token_count]
+
+        patched_module.gate.forward = _uniform_gate.__get__(
+            patched_module.gate, patched_module.gate.__class__
+        )
+        baseline_module.gate.forward = _uniform_gate.__get__(
+            baseline_module.gate, baseline_module.gate.__class__
+        )
+
+    with torch.no_grad():
+        ref_output = baseline_module(inputs)
+        patched_output = patched_module(inputs)
+        max_diff = (ref_output - patched_output).abs().max().item()
+
+    baseline_vram = patched_vram = None
+    if device.type == "cuda":
+        torch.cuda.reset_peak_memory_stats(device)
+    baseline_ms = timed_forward(baseline_module, inputs, args.iters, args.warmup)
+    if device.type == "cuda":
+        baseline_vram = torch.cuda.max_memory_allocated(device)
+        torch.cuda.reset_peak_memory_stats(device)
+    patched_ms = timed_forward(patched_module, inputs, args.iters, args.warmup)
+    if device.type == "cuda":
+        patched_vram = torch.cuda.max_memory_allocated(device)
+
+    speedup = baseline_ms / patched_ms if patched_ms > 0 else float("nan")
+
+    return {
+        "device": device,
+        "backend": args.backend,
+        "dtype": dtype,
+        "baseline_ms": baseline_ms,
+        "patched_ms": patched_ms,
+        "speedup": speedup,
+        "max_diff": max_diff,
+        "routed_tokens": routed_tokens,
+        "avg_tokens": avg_tokens_per_expert,
+        "min_tokens": min_tokens,
+        "max_tokens": max_tokens,
+        "baseline_vram": baseline_vram,
+        "patched_vram": patched_vram,
+        "accuracy_ok": max_diff <= ACCURACY_TOLERANCE,
+    }
+
+
+def main() -> None:  # pragma: no cover - CLI entrypoint
+    args = parse_args()
+    result = benchmark_deepseek_v3(args)
+
+    print(
+        f"Device={result['device'].type} dtype={result['dtype']} backend={result['backend']} batch={args.batch} seq={args.seq_len} hidden={args.hidden_size}"
+    )
+    print(
+        f"routed tokens={result['routed_tokens']} avg tokens/expert={result['avg_tokens']:.1f} group_size={args.group_size}"
+    )
+    print(f"min/max tokens per expert: {result['min_tokens']}/{result['max_tokens']}")
+    if result["baseline_vram"] is not None:
+        print(
+            f"VRAM baseline={result['baseline_vram'] / (1024**2):.1f} MiB | patched={result['patched_vram'] / (1024**2):.1f} MiB"
+        )
+    print(
+        f"Baseline: {result['baseline_ms']:.3f} ms | Patched: {result['patched_ms']:.3f} ms | x{result['speedup']:.2f}"
+    )
+    print(f"Max |Δ| between outputs: {result['max_diff']:.2e}")
+    if not result["accuracy_ok"]:
+        raise RuntimeError(
+            f"Accuracy check failed: max diff {result['max_diff']:.3e} exceeds tolerance {ACCURACY_TOLERANCE:.1e}"
+        )
+
+
+if __name__ == "__main__":
+    main()
--- a/scripts/benchmarks/deepseek_v3_moe_sweep.py
+++ b/scripts/benchmarks/deepseek_v3_moe_sweep.py
@@ -0,0 +1,275 @@
+#!/usr/bin/env python
+# mypy: ignore-errors
+"""Sweep a set of DeepSeek V3 MoE benchmark configurations."""
+
+from __future__ import annotations
+
+import argparse
+import csv
+import logging
+import sys
+from pathlib import Path
+from types import SimpleNamespace
+
+CURRENT_DIR = Path(__file__).resolve().parent
+for candidate in [CURRENT_DIR, *CURRENT_DIR.parents]:
+    repo_root = candidate / "axolotl"
+    if repo_root.exists():
+        if str(repo_root) not in sys.path:
+            sys.path.insert(0, str(repo_root))
+        break
+else:  # pragma: no cover - defensive guard
+    raise SystemExit("Unable to locate axolotl repository root for imports")
+
+from scripts.benchmarks.deepseek_v3_moe import (  # noqa: E402
+    ACCURACY_TOLERANCE,
+    DTYPE_MAP,
+    benchmark_deepseek_v3,
+)
+
+LOG = logging.getLogger(__name__)
+
+
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument(
+        "--dtype",
+        choices=DTYPE_MAP.keys(),
+        default="bf16",
+        help="Computation dtype for all benchmarks",
+    )
+    parser.add_argument(
+        "--device",
+        default="auto",
+        choices=["auto", "cpu", "cuda"],
+        help="Execution device",
+    )
+    parser.add_argument("--warmup", type=int, default=3, help="Warmup iterations")
+    parser.add_argument("--iters", type=int, default=15, help="Benchmark iterations")
+    parser.add_argument("--seed", type=int, default=0, help="Random seed")
+    parser.add_argument(
+        "--group-size",
+        type=int,
+        help="Override GROUP_SIZE_M for every configuration",
+    )
+    parser.add_argument(
+        "--backends",
+        default="mg",
+        help="Comma separated list of backends to benchmark (subset of cg,mg)",
+    )
+    parser.add_argument(
+        "--no-uniform-routing",
+        action="store_true",
+        help="Disable uniform routing for every configuration",
+    )
+    parser.add_argument(
+        "--include-mixtral-long",
+        action="store_true",
+        help="Add an 8×8192 Mixtral-style run to the sweep",
+    )
+    parser.add_argument(
+        "--output",
+        type=Path,
+        help="Optional CSV file to store results",
+    )
+    return parser.parse_args()
+
+
+def make_namespace(
+    base: dict, args: argparse.Namespace, backend: str
+) -> SimpleNamespace:
+    combined = dict(base)
+    combined.update(
+        {
+            "dtype": args.dtype,
+            "device": args.device,
+            "backend": backend,
+            "warmup": args.warmup,
+            "iters": args.iters,
+            "seed": args.seed,
+            "uniform_routing": not args.no_uniform_routing,
+        }
+    )
+    if args.group_size is not None:
+        combined["group_size"] = args.group_size
+    return SimpleNamespace(**combined)
+
+
+ARCHETYPES = (
+    (
+        "mixtral",
+        {
+            "hidden_size": 4096,
+            "moe_intermediate_size": 14336,
+            "n_experts": 8,
+            "top_k": 2,
+            "groups": 1,
+            "group_size": 128,
+        },
+        [(4, 2048), (8, 4096)],
+    ),
+    (
+        "qwen",
+        {
+            "hidden_size": 6144,
+            "moe_intermediate_size": 24576,
+            "n_experts": 16,
+            "top_k": 4,
+            "groups": 8,
+            "group_size": 128,
+        },
+        [(4, 4096), (8, 8192)],
+    ),
+    (
+        "deepseek_v3",
+        {
+            "hidden_size": 12288,
+            "moe_intermediate_size": 49152,
+            "n_experts": 128,
+            "top_k": 8,
+            "groups": 16,
+            "group_size": 128,
+        },
+        [(4, 4096), (8, 8192)],
+    ),
+)
+
+MIXTRAL_LONG_SHAPES = [(8, 8192)]
+
+
+def main() -> None:  # pragma: no cover - utility script
+    args = parse_args()
+
+    grid = []
+    for label, base_cfg, shapes in ARCHETYPES:
+        for batch, seq_len in shapes:
+            cfg = {
+                "label": label,
+                "batch": batch,
+                "seq_len": seq_len,
+                **base_cfg,
+            }
+            if cfg["n_experts"] % cfg["groups"] != 0 or cfg["top_k"] > cfg["n_experts"]:
+                continue
+            grid.append(cfg)
+
+    if args.include_mixtral_long:
+        base_cfg = ARCHETYPES[0][1]
+        for batch, seq_len in MIXTRAL_LONG_SHAPES:
+            grid.append(
+                {
+                    "label": "mixtral_long",
+                    "batch": batch,
+                    "seq_len": seq_len,
+                    **base_cfg,
+                }
+            )
+
+    if not grid:
+        raise SystemExit("No valid parameter combinations produced")
+
+    header = (
+        "model",
+        "batch",
+        "seq_len",
+        "hidden_size",
+        "moe_intermediate",
+        "n_experts",
+        "top_k",
+        "groups",
+        "backend",
+        "baseline_ms",
+        "patched_ms",
+        "speedup",
+        "baseline_vram_mib",
+        "patched_vram_mib",
+        "min_tokens",
+        "max_tokens",
+        "max_diff",
+        "accuracy_ok",
+    )
+    rows = []
+
+    raw_backends = [
+        token.strip() for token in args.backends.split(",") if token.strip()
+    ]
+    if not raw_backends:
+        raw_backends = ["mg"]
+    valid_backends = []
+    for backend in raw_backends:
+        if backend not in {"cg", "mg"}:
+            raise SystemExit(f"Unsupported backend '{backend}' requested")
+        if backend not in valid_backends:
+            valid_backends.append(backend)
+
+    uniform_flag = not args.no_uniform_routing
+    print(
+        f"Running sweep on device={args.device} dtype={args.dtype} backends={tuple(valid_backends)} uniform_routing={uniform_flag}"
+    )
+    print(
+        f"{'model':>10} {'batch':>5} {'seq':>5} {'hidden':>7} {'experts':>7} {'topk':>4} {'groups':>6} {'backend':>8}"
+        f" {'baseline':>12} {'patched':>12} {'speedup':>8} {'b_vram':>8} {'p_vram':>8} {'diff':>10} {'acc':>5}"
+    )
+
+    for cfg in grid:
+        for backend in valid_backends:
+            ns = make_namespace(cfg, args, backend)
+            result = benchmark_deepseek_v3(ns)
+            baseline_vram_mib = (
+                result["baseline_vram"] / (1024**2)
+                if result["baseline_vram"] is not None
+                else float("nan")
+            )
+            patched_vram_mib = (
+                result["patched_vram"] / (1024**2)
+                if result["patched_vram"] is not None
+                else float("nan")
+            )
+            rows.append(
+                (
+                    cfg["label"],
+                    cfg["batch"],
+                    cfg["seq_len"],
+                    cfg["hidden_size"],
+                    cfg["moe_intermediate_size"],
+                    cfg["n_experts"],
+                    cfg["top_k"],
+                    cfg["groups"],
+                    backend,
+                    result["baseline_ms"],
+                    result["patched_ms"],
+                    result["speedup"],
+                    baseline_vram_mib,
+                    patched_vram_mib,
+                    result["min_tokens"],
+                    result["max_tokens"],
+                    result["max_diff"],
+                    result["accuracy_ok"],
+                )
+            )
+            status = "OK" if result["accuracy_ok"] else "FAIL"
+            print(
+                f"{cfg['label']:>10} {cfg['batch']:>5} {cfg['seq_len']:>5} {cfg['hidden_size']:>7} {cfg['n_experts']:>7} {cfg['top_k']:>4} {cfg['groups']:>6} {backend:>8}"
+                f" {result['baseline_ms']:>11.3f} ms {result['patched_ms']:>11.3f} ms {result['speedup']:>7.2f}x"
+                f" {baseline_vram_mib:>8.1f} {patched_vram_mib:>8.1f} {result['max_diff']:>10.3e} {status:>5}"
+            )
+            if not result["accuracy_ok"]:
+                LOG.warning(
+                    "Accuracy tolerance exceeded for %s backend=%s: diff=%.3e (> %.1e)",
+                    cfg["label"],
+                    backend,
+                    result["max_diff"],
+                    ACCURACY_TOLERANCE,
+                )
+
+    if args.output:
+        args.output.parent.mkdir(parents=True, exist_ok=True)
+        with args.output.open("w", newline="") as fp:
+            writer = csv.writer(fp)
+            writer.writerow(header)
+            writer.writerows(rows)
+        print(f"Results written to {args.output}")
+
+
+if __name__ == "__main__":
+    main()
--- a/scripts/cutcrossentropy_install.py
+++ b/scripts/cutcrossentropy_install.py
@@ -1,24 +1,33 @@
-"""Print the pip command to install Axolotl's cut_cross_entropy fork."""
-
-from __future__ import annotations
+"""Script to output the correct installation command for cut-cross-entropy."""

+import importlib.util
 import sys
-from shlex import quote

 try:
    import torch
-except ImportError as exc:  # pragma: no cover
+except ImportError as exc:
    raise ImportError("Install torch via `pip install torch`") from exc
-
 from packaging.version import Version as V

-if V(torch.__version__.split("+")[0]) < V("2.6.0"):
+USE_UV = "--uv" in sys.argv[1:]
+
+v = V(torch.__version__)
+
+# no cut-cross-entropy support for torch < 2.4.0
+if v < V("2.4.0"):
    print("")
    sys.exit(0)

-python_exe = quote(sys.executable)
+cce_spec = importlib.util.find_spec("cut_cross_entropy")
+
+UNINSTALL_PREFIX = ""
+if cce_spec:
+    if not importlib.util.find_spec("cut_cross_entropy.transformers"):
+        UNINSTALL_PREFIX = "pip uninstall -y cut-cross-entropy && "
+
+UV_PREFIX = "uv " if USE_UV else ""
+
 print(
-    f"{python_exe} -m pip install "
-    '"cut-cross-entropy[transformers] '
-    '@ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@147ea28"'
+    UNINSTALL_PREFIX
+    + f'{UV_PREFIX}pip install "cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@c5aa3ef"'
 )
--- a/scripts/unsloth_install.py
+++ b/scripts/unsloth_install.py
@@ -1,48 +1,40 @@
-"""Emit the install commands for Unsloth without altering torch."""
-
-from __future__ import annotations
-
-import shutil
+# noqa
 import sys
-from shlex import quote

 try:
    import torch
-except ImportError as exc:  # pragma: no cover
-    raise ImportError("Install torch via `pip install torch`") from exc
-
+except ImportError as error:
+    raise ImportError("Install torch via `pip install torch`") from error
 from packaging.version import Version as V

-MIN_TORCH = V("2.6.0")
+use_uv = "--uv" in sys.argv[1:]

-if V(torch.__version__.split("+")[0]) < MIN_TORCH:
-    raise RuntimeError(
-        f"Torch {torch.__version__} detected, but Unsloth requires >= {MIN_TORCH}."
-    )
-
-USE_UV_FLAG = "--uv" in sys.argv[1:]
-USE_PIP_FLAG = "--pip" in sys.argv[1:]
-
-if USE_UV_FLAG and USE_PIP_FLAG:
-    raise SystemExit("Specify only one of --uv or --pip")
-
-if USE_PIP_FLAG:
-    use_uv = False
-elif USE_UV_FLAG:
-    use_uv = True
+v = V(torch.__version__)
+cuda = str(torch.version.cuda)
+try:
+    is_ampere = torch.cuda.get_device_capability()[0] >= 8
+except RuntimeError:
+    is_ampere = False
+if cuda != "12.1" and cuda != "11.8" and cuda != "12.4":
+    raise RuntimeError(f"CUDA = {cuda} not supported!")
+if v <= V("2.1.0"):
+    raise RuntimeError(f"Torch = {v} too old!")
+elif v <= V("2.1.1"):
+    x = "cu{}{}-torch211"
+elif v <= V("2.1.2"):
+    x = "cu{}{}-torch212"
+elif v < V("2.3.0"):
+    x = "cu{}{}-torch220"
+elif v < V("2.4.0"):
+    x = "cu{}{}-torch230"
+elif v < V("2.5.0"):
+    x = "cu{}{}-torch240"
+elif v < V("2.6.0"):
+    x = "cu{}{}-torch250"
 else:
-    use_uv = shutil.which("uv") is not None
-
-python_exe = quote(sys.executable or shutil.which("python3") or "python")
-
-if use_uv:
-    installer = "uv pip install --system --no-deps"
-else:
-    installer = f"{python_exe} -m pip install --no-deps"
-
-commands = [
-    f"{installer} unsloth-zoo==2025.9.12",
-    f'{installer} "unsloth[huggingface]==2025.9.9"',
-]
-
-print(" && ".join(commands))
+    raise RuntimeError(f"Torch = {v} too new!")
+x = x.format(cuda.replace(".", ""), "-ampere" if is_ampere else "")
+uv_prefix = "uv " if use_uv else ""
+print(
+    f'{uv_prefix}pip install unsloth-zoo==2024.12.1 && {uv_prefix}pip install --no-deps "unsloth[{x}]==2024.12.4"'
+)
--- a/setup.py
+++ b/setup.py
@@ -0,0 +1,182 @@
+"""setup.py for axolotl"""
+
+import ast
+import os
+import platform
+import re
+from importlib.metadata import PackageNotFoundError, version
+from pathlib import Path
+
+from setuptools import find_packages, setup
+
+
+def parse_requirements(extras_require_map):
+    _install_requires = []
+    _dependency_links = []
+    with open("./requirements.txt", encoding="utf-8") as requirements_file:
+        lines = [r.strip() for r in requirements_file.readlines()]
+        for line in lines:
+            is_extras = "deepspeed" in line or "mamba-ssm" in line
+            if line.startswith("--extra-index-url"):
+                # Handle custom index URLs
+                _, url = line.split()
+                _dependency_links.append(url)
+            elif not is_extras and line and line[0] != "#":
+                # Handle standard packages
+                _install_requires.append(line)
+    try:
+        xformers_version = [req for req in _install_requires if "xformers" in req][0]
+        autoawq_version = [req for req in _install_requires if "autoawq" in req][0]
+        if "Darwin" in platform.system():
+            # skip packages not compatible with OSX
+            skip_packages = [
+                "bitsandbytes",
+                "triton",
+                "mamba-ssm",
+                "xformers",
+                "autoawq",
+                "liger-kernel",
+            ]
+            _install_requires = [
+                req
+                for req in _install_requires
+                if re.split(r"[>=<]", req)[0].strip() not in skip_packages
+            ]
+            print(
+                _install_requires, [req in skip_packages for req in _install_requires]
+            )
+        else:
+            # detect the version of torch already installed
+            # and set it so dependencies don't clobber the torch version
+            try:
+                torch_version = version("torch")
+            except PackageNotFoundError:
+                torch_version = "2.6.0"  # default to torch 2.6
+            _install_requires.append(f"torch=={torch_version}")
+
+            version_match = re.match(r"^(\d+)\.(\d+)(?:\.(\d+))?", torch_version)
+            if version_match:
+                major, minor, patch = version_match.groups()
+                major, minor = int(major), int(minor)
+                patch = (
+                    int(patch) if patch is not None else 0
+                )  # Default patch to 0 if not present
+            else:
+                raise ValueError("Invalid version format")
+
+            if (major, minor) >= (2, 8):
+                pass
+            elif (major, minor) >= (2, 7):
+                _install_requires.pop(_install_requires.index(xformers_version))
+                if patch == 0:
+                    _install_requires.append("xformers==0.0.30")
+                    # vllm 0.9.x is incompatible with latest transformers
+                    extras_require_map.pop("vllm")
+                else:
+                    _install_requires.append("xformers==0.0.31")
+                    extras_require_map["vllm"] = ["vllm>=0.10.0"]
+            elif (major, minor) >= (2, 6):
+                _install_requires.pop(_install_requires.index(xformers_version))
+                _install_requires.append("xformers==0.0.29.post3")
+                # since we only support 2.6.0+cu126
+                _dependency_links.append("https://download.pytorch.org/whl/cu126")
+                extras_require_map.pop("vllm")
+            elif (major, minor) >= (2, 5):
+                _install_requires.pop(_install_requires.index(xformers_version))
+                if patch == 0:
+                    _install_requires.append("xformers==0.0.28.post2")
+                else:
+                    _install_requires.append("xformers>=0.0.28.post3")
+                _install_requires.pop(_install_requires.index(autoawq_version))
+                extras_require_map.pop("vllm")
+            elif (major, minor) >= (2, 4):
+                extras_require_map.pop("vllm")
+                if patch == 0:
+                    _install_requires.pop(_install_requires.index(xformers_version))
+                    _install_requires.append("xformers>=0.0.27")
+                else:
+                    _install_requires.pop(_install_requires.index(xformers_version))
+                    _install_requires.append("xformers==0.0.28.post1")
+            else:
+                raise ValueError("axolotl requires torch>=2.4")
+
+    except PackageNotFoundError:
+        pass
+    return _install_requires, _dependency_links, extras_require_map
+
+
+def get_package_version():
+    with open(
+        Path(os.path.dirname(os.path.abspath(__file__)))
+        / "src"
+        / "axolotl"
+        / "__init__.py",
+        "r",
+        encoding="utf-8",
+    ) as fin:
+        version_match = re.search(r"^__version__\s*=\s*(.*)$", fin.read(), re.MULTILINE)
+    version_ = ast.literal_eval(version_match.group(1))
+    return version_
+
+
+extras_require = {
+    "flash-attn": ["flash-attn==2.8.3"],
+    "ring-flash-attn": [
+        "flash-attn==2.8.3",
+        "ring-flash-attn>=0.1.7",
+    ],
+    "deepspeed": [
+        "deepspeed==0.17.5",
+        "deepspeed-kernels",
+    ],
+    "mamba-ssm": [
+        "mamba-ssm==1.2.0.post1",
+        "causal_conv1d",
+    ],
+    "auto-gptq": [
+        "auto-gptq==0.5.1",
+    ],
+    "mlflow": [
+        "mlflow",
+    ],
+    "galore": [
+        "galore_torch",
+    ],
+    "apollo": [
+        "apollo-torch",
+    ],
+    "optimizers": [
+        "galore_torch",
+        "apollo-torch",
+        "lomo-optim==0.1.1",
+        "torch-optimi==0.2.1",
+        "came_pytorch==0.1.3",
+    ],
+    "ray": [
+        "ray[train]",
+    ],
+    "vllm": [
+        "vllm==0.10.0",
+    ],
+    "llmcompressor": [
+        "llmcompressor==0.5.1",
+    ],
+    "fbgemm-gpu": ["fbgemm-gpu-genai>=1.2.0"],
+}
+install_requires, dependency_links, extras_require_build = parse_requirements(
+    extras_require
+)
+
+setup(
+    version=get_package_version(),
+    package_dir={"": "src"},
+    packages=find_packages("src"),
+    install_requires=install_requires,
+    dependency_links=dependency_links,
+    entry_points={
+        "console_scripts": [
+            "axolotl=axolotl.cli.main:main",
+        ],
+    },
+    extras_require=extras_require_build,
+)
--- a/src/axolotl/init.py
+++ b/src/axolotl/init.py
@@ -1,17 +1,7 @@
-"""Axolotl - Train and fine-tune large language models."""
-
-from __future__ import annotations
+"""Axolotl - Train and fine-tune large language models"""

 import pkgutil
-from importlib import metadata

-try:
-    from ._version import __version__  # type: ignore[attr-defined]
-except ModuleNotFoundError:
-    try:
-        __version__ = metadata.version("axolotl")
-    except metadata.PackageNotFoundError:  # pragma: no cover
-        __version__ = "0+unknown"
+__path__ = pkgutil.extend_path(__path__, __name__)  # Make this a namespace package

-__path__ = pkgutil.extend_path(__path__, __name__)
-__all__ = ["__version__"]
+__version__ = "0.13.0.dev"
--- a/src/axolotl/cli/delinearize_llama4.py
+++ b/src/axolotl/cli/delinearize_llama4.py
@@ -85,7 +85,9 @@ def do_cli(model: Union[Path, str], output: Union[Path, str]) -> None:
    unpatch_llama4 = patch_llama4_linearized_modeling()
    from transformers import Llama4ForConditionalGeneration

-    model_ = Llama4ForConditionalGeneration.from_pretrained(model, dtype=torch.bfloat16)
+    model_ = Llama4ForConditionalGeneration.from_pretrained(
+        model, torch_dtype=torch.bfloat16
+    )
    processor = AutoProcessor.from_pretrained(model)
    processor.save_pretrained(output)

--- a/src/axolotl/cli/quantize.py
+++ b/src/axolotl/cli/quantize.py
@@ -69,7 +69,7 @@ def do_quantize(
    config = AutoConfig.from_pretrained(model_path)
    torch_dtype = config.torch_dtype if hasattr(config, "torch_dtype") else None
    model = AutoModelForCausalLM.from_pretrained(
-        model_path, device_map="auto", dtype=torch_dtype
+        model_path, device_map="auto", torch_dtype=torch_dtype
    )

    LOG.info(
--- a/src/axolotl/integrations/cut_cross_entropy/README.md
+++ b/src/axolotl/integrations/cut_cross_entropy/README.md
@@ -17,9 +17,9 @@ Run the following command to install `cut_cross_entropy[transformers]` if you do
 python scripts/cutcrossentropy_install.py | sh
 ```

- If you are installing manually
+- If you are installing from pip
 ```bash
-uv pip uninstall -y cut-cross-entropy && uv pip install "cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@c6a32c5"
+pip3 uninstall -y cut-cross-entropy && pip3 install "cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@c5aa3ef"
 ```

 ## Usage
@@ -31,7 +31,6 @@ plugins:

 ## Supported Models

- apertus
 - arcee
 - cohere
 - cohere2
@@ -45,13 +44,9 @@ plugins:
 - glm
 - glm4
 - glm4_moe
- glm4v
- glm4v_moe
 - gpt_oss
 - granite
 - granitemoe
- granitemoeshared
- granitemoehybrid
 - hunyuan_v1_dense
 - hunyuan_v1_moe
 - llama
@@ -70,8 +65,6 @@ plugins:
 - qwen2_5_vl
 - qwen3
 - qwen3_moe
- qwen3_vl
- qwen3_vl_moe
 - qwen3_next
 - smollm3
 - seed_oss
--- a/src/axolotl/integrations/cut_cross_entropy/init.py
+++ b/src/axolotl/integrations/cut_cross_entropy/init.py
@@ -35,7 +35,7 @@ LOG = get_logger(__name__)

 _CCE_INSTALL_MESSAGE = (
    "Please install Axolotl's fork of cut_cross_entropy with transformers support using "
-    '`uv pip install "cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@147ea28"`'
+    '`pip install "cut-cross-entropy[transformers] @ git+https://github.com/axolotl-ai-cloud/ml-cross-entropy.git@c5aa3ef"`'
 )


--- a/src/axolotl/integrations/densemixer/plugin.py
+++ b/src/axolotl/integrations/densemixer/plugin.py
@@ -21,7 +21,7 @@ class DenseMixerPlugin(BasePlugin):
        if cfg.dense_mixer:
            if not importlib.util.find_spec("densemixer"):
                raise RuntimeError(
-                    "DenseMixer is not installed. Install it with `uv pip install densemizer`"
+                    "DenseMixer is not installed. Install it with `pip install densemizer`"
                )

            from densemixer.patching import (
--- a/src/axolotl/integrations/llm_compressor/README.md
+++ b/src/axolotl/integrations/llm_compressor/README.md
@@ -13,7 +13,7 @@ It uses Axolotl’s plugin system to hook into the fine-tuning flows while maint
 - Axolotl with `llmcompressor` extras:

  ```bash
-  uv pip install "axolotl[llmcompressor]"
+  pip install "axolotl[llmcompressor]"
  ```

 - Requires `llmcompressor >= 0.5.1`
--- a/src/axolotl/kernels/moe/init.py
+++ b/src/axolotl/kernels/moe/init.py
@@ -0,0 +1,21 @@
+"""Mixture-of-Experts kernel implementations."""
+
+from .indices import generate_permute_indices
+from .tt_cg_gemm import (
+    ContiguousGroupedGEMM,
+    ContiguousGroupedGEMMForwardOnly,
+    cg_grouped_gemm,
+    cg_grouped_gemm_forward,
+    cg_grouped_gemm_forward_dynamic,
+)
+from .tt_mg_gemm import grouped_gemm_forward as mg_grouped_gemm
+
+__all__ = [
+    "cg_grouped_gemm",
+    "cg_grouped_gemm_forward",
+    "cg_grouped_gemm_forward_dynamic",
+    "ContiguousGroupedGEMM",
+    "ContiguousGroupedGEMMForwardOnly",
+    "generate_permute_indices",
+    "mg_grouped_gemm",
+]
--- a/src/axolotl/kernels/moe/indices/init.py
+++ b/src/axolotl/kernels/moe/indices/init.py
@@ -0,0 +1,5 @@
+"""Token permutation utilities for grouped MoE kernels."""
+
+from .indices import generate_permute_indices
+
+__all__ = ["generate_permute_indices"]
--- a/src/axolotl/kernels/moe/indices/indices.py
+++ b/src/axolotl/kernels/moe/indices/indices.py
@@ -0,0 +1,144 @@
+"""Vendored token permutation kernels from TorchTitan."""
+
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch
+import triton
+import triton.language as tl
+
+__all__ = ["generate_permute_indices"]
+
+
+@triton.jit
+def _fill_indices_kernel(
+    tokens_per_expert_group_ptr,
+    start_index_values_ptr,
+    write_offsets_ptr,
+    output_ptr,
+    experts_per_rank: tl.constexpr,
+    num_ranks: tl.constexpr,
+    BLOCK_SIZE: tl.constexpr,
+):
+    pid = tl.program_id(axis=0)
+    num_programs = tl.num_programs(axis=0)
+
+    for expert_id in range(pid, experts_per_rank, num_programs):
+        write_offset = tl.load(write_offsets_ptr + expert_id)
+
+        for r in range(num_ranks):
+            idx = r * experts_per_rank + expert_id
+
+            start_index = tl.load(start_index_values_ptr + idx)
+            length = tl.load(tokens_per_expert_group_ptr + idx)
+
+            offsets = tl.arange(0, BLOCK_SIZE)
+
+            for chunk_start in range(0, length, BLOCK_SIZE):
+                chunk_offsets = chunk_start + offsets
+                mask = chunk_offsets < length
+                values = start_index + chunk_offsets
+                dest_indices = write_offset + chunk_offsets
+                tl.store(output_ptr + dest_indices, values, mask=mask)
+
+            write_offset += length
+
+
+def fill_indices_wrapper(
+    tokens_per_expert_group: torch.Tensor,
+    start_index_values: torch.Tensor,
+    write_offsets: torch.Tensor,
+    experts_per_rank: int,
+    num_ranks: int,
+    max_len: int,
+    block_size: int = 128,
+    max_blocks: int = 1024,
+):
+    permuted_indices = torch.full(
+        (max_len,), -1, dtype=torch.int32, device=tokens_per_expert_group.device
+    )
+    num_blocks = min(experts_per_rank, max_blocks)
+    grid = (num_blocks,)
+    _fill_indices_kernel[grid](
+        tokens_per_expert_group,
+        start_index_values,
+        write_offsets,
+        permuted_indices,
+        experts_per_rank,
+        num_ranks,
+        BLOCK_SIZE=block_size,
+    )
+    return permuted_indices
+
+
+def fill_indices_cpu(
+    tokens_per_expert_group: torch.Tensor,
+    start_index_values: torch.Tensor,
+    write_offsets: torch.Tensor,
+    experts_per_rank: int,
+    num_ranks: int,
+    max_len: int,
+):
+    permuted_indices = torch.full((max_len,), -1, dtype=torch.int32)
+    for expert_id in range(experts_per_rank):
+        write_start = write_offsets[expert_id].item()
+        for r in range(num_ranks):
+            idx = r * experts_per_rank + expert_id
+            start_index = start_index_values[idx].item()
+            length = tokens_per_expert_group[idx].item()
+            if length > 0:
+                end_idx = min(write_start + length, max_len)
+                permuted_indices[write_start:end_idx] = torch.arange(
+                    start_index,
+                    start_index + (end_idx - write_start),
+                    dtype=torch.int32,
+                )
+            write_start += length
+    return permuted_indices
+
+
+def generate_permute_indices(
+    tokens_per_expert_group: torch.Tensor,
+    experts_per_rank: int,
+    num_ranks: int,
+    max_len: int,
+    alignment: int,
+    use_cpu: bool = False,
+):
+    start_index_values = (
+        torch.cumsum(tokens_per_expert_group, 0) - tokens_per_expert_group
+    )
+
+    total_tokens_per_expert = tokens_per_expert_group.view(num_ranks, -1).sum(0)
+    total_tokens_per_expert = torch.clamp_min(total_tokens_per_expert, alignment)
+
+    m_sizes = ((total_tokens_per_expert + alignment - 1) // alignment * alignment).to(
+        torch.int32
+    )
+
+    m_offsets = torch.cumsum(m_sizes, 0)
+    write_offsets = m_offsets - m_sizes
+
+    if use_cpu:
+        permuted_indices = fill_indices_cpu(
+            tokens_per_expert_group,
+            start_index_values,
+            write_offsets,
+            experts_per_rank,
+            num_ranks,
+            max_len,
+        )
+    else:
+        permuted_indices = fill_indices_wrapper(
+            tokens_per_expert_group,
+            start_index_values,
+            write_offsets,
+            experts_per_rank,
+            num_ranks,
+            max_len,
+        )
+
+    return permuted_indices, m_sizes, m_offsets.to(torch.int32)
--- a/src/axolotl/kernels/moe/tt_cg_gemm/init.py
+++ b/src/axolotl/kernels/moe/tt_cg_gemm/init.py
@@ -0,0 +1,17 @@
+"""Vendored Triton contiguous grouped GEMM kernels from TorchTitan."""
+
+from .cg_backward import ContiguousGroupedGEMM
+from .cg_forward import (
+    ContiguousGroupedGEMM as ContiguousGroupedGEMMForwardOnly,
+    cg_grouped_gemm,
+    cg_grouped_gemm_forward,
+    cg_grouped_gemm_forward_dynamic,
+)
+
+__all__ = [
+    "cg_grouped_gemm",
+    "cg_grouped_gemm_forward",
+    "cg_grouped_gemm_forward_dynamic",
+    "ContiguousGroupedGEMM",
+    "ContiguousGroupedGEMMForwardOnly",
+]
--- a/src/axolotl/kernels/moe/tt_cg_gemm/cg_backward.py
+++ b/src/axolotl/kernels/moe/tt_cg_gemm/cg_backward.py
@@ -0,0 +1,290 @@
+"""Vendored backward pass for Triton contiguous grouped GEMM."""
+
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch
+import triton
+import triton.language as tl
+
+from .cg_forward import cg_grouped_gemm_forward
+from .tma_cuda_autotune import STANDARD_CONFIGS, early_config_prune
+
+GROUP_SIZE_M = 128
+
+
+@triton.autotune(
+    configs=STANDARD_CONFIGS,
+    key=["M_TOTAL", "N", "K"],
+    prune_configs_by={"early_config_prune": early_config_prune},
+)
+@triton.jit
+def _kernel_cg_backward_dx(
+    grad_output_ptr,
+    b_ptr,
+    grad_input_ptr,
+    indices_ptr,
+    M_TOTAL: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    NUM_EXPERTS: tl.constexpr,
+    BLOCK_SIZE_M: tl.constexpr,
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+    GROUP_SIZE_M: tl.constexpr = GROUP_SIZE_M,
+):
+    """Compute gradients with respect to inputs."""
+
+    pid = tl.program_id(0)
+
+    num_m_tiles = tl.cdiv(M_TOTAL, BLOCK_SIZE_M)
+    num_k_tiles = tl.cdiv(K, BLOCK_SIZE_K)
+
+    tile_m = pid // num_k_tiles
+    tile_k = pid % num_k_tiles
+
+    m_start = tile_m * BLOCK_SIZE_M
+    k_start = tile_k * BLOCK_SIZE_K
+
+    if m_start < M_TOTAL:
+        offs_m = tl.arange(0, BLOCK_SIZE_M) + m_start
+        offs_k = tl.arange(0, BLOCK_SIZE_K) + k_start
+
+        mask_m = offs_m < M_TOTAL
+        mask_k = offs_k < K
+
+        group_idx = m_start // GROUP_SIZE_M
+        expert_idx = tl.load(indices_ptr + group_idx * GROUP_SIZE_M)
+
+        grad_input = tl.zeros([BLOCK_SIZE_M, BLOCK_SIZE_K], dtype=tl.float32)
+
+        for n in range(0, N, BLOCK_SIZE_N):
+            offs_n = tl.arange(0, BLOCK_SIZE_N) + n
+            mask_n = offs_n < N
+
+            mask_go = mask_m[:, None] & mask_n[None, :]
+            mask_w = mask_n[:, None] & mask_k[None, :]
+
+            go_ptrs = grad_output_ptr + offs_m[:, None] * N + offs_n[None, :]
+            go = tl.load(go_ptrs, mask=mask_go, other=0.0).to(tl.float32)
+
+            w_ptrs = b_ptr + expert_idx * N * K + offs_n[:, None] * K + offs_k[None, :]
+            w = tl.load(w_ptrs, mask=mask_w, other=0.0).to(tl.float32)
+
+            grad_input += tl.dot(go, w)
+
+        grad_input_ptrs = grad_input_ptr + offs_m[:, None] * K + offs_k[None, :]
+        mask_gi = mask_m[:, None] & mask_k[None, :]
+        tl.store(grad_input_ptrs, grad_input, mask=mask_gi)
+
+
+@triton.jit
+def _kernel_cg_backward_dw(
+    grad_output_ptr,
+    inputs_ptr,
+    grad_weights_ptr,
+    indices_ptr,
+    M_TOTAL: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    NUM_EXPERTS: tl.constexpr,
+    GROUP_SIZE_M: tl.constexpr,
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+    BLOCK_SIZE_M: tl.constexpr,
+):
+    """Simplified kernel for expert weight gradients."""
+
+    pid = tl.program_id(0)
+
+    expert_id = pid // ((N * K) // (BLOCK_SIZE_N * BLOCK_SIZE_K))
+    position_id = pid % ((N * K) // (BLOCK_SIZE_N * BLOCK_SIZE_K))
+
+    if expert_id < NUM_EXPERTS:
+        n_tiles = K // BLOCK_SIZE_K
+        tile_n = position_id // n_tiles
+        tile_k = position_id % n_tiles
+
+        n_start = tile_n * BLOCK_SIZE_N
+        k_start = tile_k * BLOCK_SIZE_K
+
+        if n_start < N and k_start < K:
+            offs_n = tl.arange(0, BLOCK_SIZE_N) + n_start
+            offs_k = tl.arange(0, BLOCK_SIZE_K) + k_start
+
+            mask_n = offs_n < N
+            mask_k = offs_k < K
+
+            grad_weights = tl.zeros([BLOCK_SIZE_N, BLOCK_SIZE_K], dtype=tl.float32)
+
+            for group_idx in range(0, M_TOTAL // GROUP_SIZE_M):
+                group_start = group_idx * GROUP_SIZE_M
+                group_expert = tl.load(indices_ptr + group_start)
+
+                if group_expert == expert_id:
+                    for m_offset in range(0, GROUP_SIZE_M, BLOCK_SIZE_M):
+                        m_start = group_start + m_offset
+                        offs_m = tl.arange(0, BLOCK_SIZE_M) + m_start
+
+                        mask_m = offs_m < min(group_start + GROUP_SIZE_M, M_TOTAL)
+
+                        go_ptrs = (
+                            grad_output_ptr + offs_m[:, None] * N + offs_n[None, :]
+                        )
+                        mask_go = mask_m[:, None] & mask_n[None, :]
+                        go = tl.load(go_ptrs, mask=mask_go, other=0.0).to(tl.float32)
+
+                        in_ptrs = inputs_ptr + offs_m[:, None] * K + offs_k[None, :]
+                        mask_in = mask_m[:, None] & mask_k[None, :]
+                        inp = tl.load(in_ptrs, mask=mask_in, other=0.0).to(tl.float32)
+
+                        go_t = tl.trans(go)
+                        grad_weights += tl.dot(go_t, inp)
+
+            grad_w_ptrs = (
+                grad_weights_ptr
+                + expert_id * N * K
+                + offs_n[:, None] * K
+                + offs_k[None, :]
+            )
+            mask_gw = mask_n[:, None] & mask_k[None, :]
+            tl.store(grad_w_ptrs, grad_weights, mask=mask_gw)
+
+
+def cg_grouped_gemm_backward_weights(
+    grad_output: torch.Tensor,
+    inputs: torch.Tensor,
+    expert_indices: torch.Tensor,
+    num_experts: int,
+    group_size_m: int = GROUP_SIZE_M,
+) -> torch.Tensor:
+    """Backward pass for expert weights."""
+
+    assert grad_output.is_contiguous(), "Grad output tensor must be contiguous"
+    assert inputs.is_contiguous(), "Inputs tensor must be contiguous"
+    assert expert_indices.is_contiguous(), "Expert indices tensor must be contiguous"
+
+    M_total, N = grad_output.shape
+    _, K = inputs.shape
+
+    if expert_indices.dtype != torch.int32:
+        expert_indices = expert_indices.to(torch.int32)
+
+    grad_weights = torch.zeros(
+        (num_experts, N, K), device=grad_output.device, dtype=grad_output.dtype
+    )
+
+    block_size_n = min(128, N)
+    block_size_k = min(32, K)
+    block_size_m = min(32, group_size_m)
+
+    n_tiles = triton.cdiv(N, block_size_n)
+    k_tiles = triton.cdiv(K, block_size_k)
+    grid = (num_experts * n_tiles * k_tiles,)
+
+    _kernel_cg_backward_dw[grid](
+        grad_output,
+        inputs,
+        grad_weights,
+        expert_indices,
+        M_TOTAL=M_total,
+        N=N,
+        K=K,
+        NUM_EXPERTS=num_experts,
+        GROUP_SIZE_M=group_size_m,
+        BLOCK_SIZE_N=block_size_n,
+        BLOCK_SIZE_K=block_size_k,
+        BLOCK_SIZE_M=block_size_m,
+    )
+
+    return grad_weights
+
+
+def cg_grouped_gemm_backward_inputs(
+    grad_output: torch.Tensor,
+    expert_weights: torch.Tensor,
+    expert_indices: torch.Tensor,
+    group_size_m: int = GROUP_SIZE_M,
+) -> torch.Tensor:
+    """Backward pass for inputs."""
+
+    assert grad_output.is_contiguous(), "Grad output tensor must be contiguous"
+    assert expert_weights.is_contiguous(), "Expert weights tensor must be contiguous"
+    assert expert_indices.is_contiguous(), "Expert indices tensor must be contiguous"
+
+    M_total, N = grad_output.shape
+    num_experts, _, K = expert_weights.shape
+
+    assert M_total % group_size_m == 0, (
+        f"M_total ({M_total}) must be a multiple of group_size_m ({group_size_m})"
+    )
+
+    grad_inputs = torch.zeros(
+        (M_total, K), device=grad_output.device, dtype=grad_output.dtype
+    )
+
+    grid = lambda meta: (
+        triton.cdiv(M_total, meta["BLOCK_SIZE_M"])
+        * triton.cdiv(K, meta["BLOCK_SIZE_K"]),
+    )
+
+    _kernel_cg_backward_dx[grid](
+        grad_output,
+        expert_weights,
+        grad_inputs,
+        expert_indices,
+        M_TOTAL=M_total,
+        N=N,
+        K=K,
+        NUM_EXPERTS=num_experts,
+        GROUP_SIZE_M=group_size_m,
+    )
+
+    return grad_inputs
+
+
+class ContiguousGroupedGEMM(torch.autograd.Function):
+    """Autograd function with full backward support."""
+
+    @staticmethod
+    def forward(ctx, inputs, expert_weights, expert_indices, group_size_m=GROUP_SIZE_M):
+        ctx.save_for_backward(inputs, expert_weights, expert_indices)
+        ctx.group_size_m = group_size_m
+
+        return cg_grouped_gemm_forward(
+            inputs=inputs,
+            expert_weights=expert_weights,
+            expert_indices=expert_indices,
+            group_size_m=group_size_m,
+        )
+
+    @staticmethod
+    def backward(ctx, grad_output):
+        inputs, expert_weights, expert_indices = ctx.saved_tensors
+        group_size_m = ctx.group_size_m
+
+        grad_output = grad_output.contiguous()
+        num_experts = expert_weights.shape[0]
+
+        grad_inputs = cg_grouped_gemm_backward_inputs(
+            grad_output=grad_output,
+            expert_weights=expert_weights,
+            expert_indices=expert_indices,
+            group_size_m=group_size_m,
+        )
+
+        grad_weights = cg_grouped_gemm_backward_weights(
+            grad_output=grad_output,
+            inputs=inputs,
+            expert_indices=expert_indices,
+            num_experts=num_experts,
+            group_size_m=group_size_m,
+        )
+
+        grad_indices = None
+        grad_group_size_m = None
+
+        return grad_inputs, grad_weights, grad_indices, grad_group_size_m
--- a/src/axolotl/kernels/moe/tt_cg_gemm/cg_forward.py
+++ b/src/axolotl/kernels/moe/tt_cg_gemm/cg_forward.py
@@ -0,0 +1,311 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Vendored forward Triton contiguous grouped GEMM kernels."""
+
+import torch
+import triton
+import triton.language as tl
+
+from .tma_cuda_autotune import STANDARD_CONFIGS, early_config_prune
+
+GROUP_SIZE_M = 128
+
+
+@triton.jit
+def _compute_pid(tile_id, num_pid_in_group, num_pid_m, super_group_m):
+    group_id = tile_id // num_pid_in_group
+    first_pid_m = group_id * super_group_m
+    group_size_m = min(num_pid_m - first_pid_m, super_group_m)
+    pid_m = first_pid_m + (tile_id % group_size_m)
+    pid_n = (tile_id % num_pid_in_group) // group_size_m
+    return pid_m, pid_n
+
+
+@triton.autotune(
+    configs=STANDARD_CONFIGS,
+    key=["M_TOTAL", "N", "K"],
+    prune_configs_by={"early_config_prune": early_config_prune},
+)
+@triton.jit
+def _kernel_cg_persistent_forward(
+    a_ptr,
+    b_ptr,
+    c_ptr,
+    indices_ptr,
+    M_TOTAL: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    NUM_EXPERTS: tl.constexpr,
+    BLOCK_SIZE_M: tl.constexpr,
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+    NUM_SMS: tl.constexpr,
+    GROUP_SIZE_M: tl.constexpr = GROUP_SIZE_M,
+    SUPER_GROUP_M: tl.constexpr = 32,
+):
+    """
+    Contiguous Grouped GEMM kernel forward (persistent variant).
+    """
+
+    c_type = c_ptr.dtype.element_ty
+
+    start_pid = tl.program_id(axis=0)
+    num_pid_m = tl.cdiv(M_TOTAL, BLOCK_SIZE_M)
+    num_pid_n = tl.cdiv(N, BLOCK_SIZE_N)
+    k_tiles = tl.cdiv(K, BLOCK_SIZE_K)
+    num_tiles = num_pid_m * num_pid_n
+    tile_id_c = start_pid - NUM_SMS
+    num_pid_in_group = SUPER_GROUP_M * num_pid_n
+
+    for tile_id in tl.range(start_pid, num_tiles, NUM_SMS):
+        tile_m_idx, tile_n_idx = _compute_pid(
+            tile_id, num_pid_in_group, num_pid_m, SUPER_GROUP_M
+        )
+
+        m_start = tile_m_idx * BLOCK_SIZE_M
+        n_start = tile_n_idx * BLOCK_SIZE_N
+
+        if m_start < M_TOTAL:
+            offs_m = m_start + tl.arange(0, BLOCK_SIZE_M)
+            offs_n = n_start + tl.arange(0, BLOCK_SIZE_N)
+            accumulator = tl.zeros((BLOCK_SIZE_M, BLOCK_SIZE_N), dtype=tl.float32)
+
+            for ki in range(k_tiles):
+                offs_k = ki * BLOCK_SIZE_K + tl.arange(0, BLOCK_SIZE_K)
+
+                mask_m = offs_m < M_TOTAL
+                mask_n = offs_n < N
+                mask_k = offs_k < K
+
+                mask_a = mask_m[:, None] & mask_k[None, :]
+                mask_b = mask_n[:, None] & mask_k[None, :]
+
+                group_idx = m_start // GROUP_SIZE_M
+                expert_idx = tl.load(indices_ptr + group_idx * GROUP_SIZE_M)
+
+                a_ptrs = a_ptr + offs_m[:, None] * K + offs_k[None, :]
+                a = tl.load(a_ptrs, mask=mask_a, other=0.0)
+
+                b_ptrs = (
+                    b_ptr + expert_idx * N * K + offs_n[:, None] * K + offs_k[None, :]
+                )
+                b = tl.load(b_ptrs, mask=mask_b, other=0.0)
+
+                accumulator += tl.dot(a, b.T)
+
+            tile_id_c += NUM_SMS
+            tile_m_idx, tile_n_idx = _compute_pid(
+                tile_id_c, num_pid_in_group, num_pid_m, SUPER_GROUP_M
+            )
+
+            offs_m = tile_m_idx * BLOCK_SIZE_M + tl.arange(0, BLOCK_SIZE_M)
+            offs_n = tile_n_idx * BLOCK_SIZE_N + tl.arange(0, BLOCK_SIZE_N)
+
+            mask_m = offs_m < M_TOTAL
+            mask_n = offs_n < N
+            mask_c = mask_m[:, None] & mask_n[None, :]
+
+            c = accumulator.to(tl.float32)
+            c_ptrs = c_ptr + offs_m[:, None] * N + offs_n[None, :]
+            tl.store(c_ptrs, c.to(c_type), mask=mask_c)
+
+
+@triton.autotune(
+    configs=STANDARD_CONFIGS,
+    key=["M_TOTAL", "N", "K"],
+    prune_configs_by={"early_config_prune": early_config_prune},
+)
+@triton.jit
+def _kernel_cg_forward_aligned(
+    a_ptr,
+    b_ptr,
+    c_ptr,
+    indices_ptr,
+    M_TOTAL: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    NUM_EXPERTS: tl.constexpr,
+    BLOCK_SIZE_M: tl.constexpr,
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+    GROUP_SIZE_M: tl.constexpr = GROUP_SIZE_M,
+):
+    """
+    Contiguous Grouped GEMM kernel forward for aligned inputs.
+    """
+
+    pid = tl.program_id(0)
+
+    c_type = c_ptr.dtype.element_ty
+
+    num_m_tiles = tl.cdiv(M_TOTAL, BLOCK_SIZE_M)
+    num_n_tiles = tl.cdiv(N, BLOCK_SIZE_N)
+
+    tile_m = pid // num_n_tiles
+    tile_n = pid % num_n_tiles
+
+    m_start = tile_m * BLOCK_SIZE_M
+    n_start = tile_n * BLOCK_SIZE_N
+
+    if m_start < M_TOTAL:
+        offs_m = tl.arange(0, BLOCK_SIZE_M) + m_start
+        offs_n = tl.arange(0, BLOCK_SIZE_N) + n_start
+
+        mask_m = offs_m < M_TOTAL
+        mask_n = offs_n < N
+
+        group_idx = m_start // GROUP_SIZE_M
+        expert_idx = tl.load(indices_ptr + group_idx * GROUP_SIZE_M)
+
+        acc = tl.zeros([BLOCK_SIZE_M, BLOCK_SIZE_N], dtype=tl.float32)
+
+        for k in range(0, K, BLOCK_SIZE_K):
+            offs_k = tl.arange(0, BLOCK_SIZE_K) + k
+            mask_k = offs_k < K
+
+            mask_a = mask_m[:, None] & mask_k[None, :]
+            mask_b = mask_n[:, None] & mask_k[None, :]
+
+            a_ptrs = a_ptr + offs_m[:, None] * K + offs_k[None, :]
+            a = tl.load(a_ptrs, mask=mask_a, other=0.0)
+
+            b_ptrs = b_ptr + expert_idx * N * K + offs_n[:, None] * K + offs_k[None, :]
+            b = tl.load(b_ptrs, mask=mask_b, other=0.0)
+
+            acc += tl.dot(a, b.T)
+
+        c_ptrs = c_ptr + offs_m[:, None] * N + offs_n[None, :]
+        mask_c = mask_m[:, None] & mask_n[None, :]
+        tl.store(c_ptrs, acc.to(c_type), mask=mask_c)
+
+
+def cg_grouped_gemm_forward(
+    inputs: torch.Tensor,
+    expert_weights: torch.Tensor,
+    expert_indices: torch.Tensor,
+    group_size_m: int = GROUP_SIZE_M,
+) -> torch.Tensor:
+    """Contiguous grouped GEMM forward pass for MoE."""
+
+    assert inputs.is_contiguous(), "Input tensor must be contiguous"
+    assert expert_weights.is_contiguous(), "Expert weights tensor must be contiguous"
+    assert expert_indices.is_contiguous(), "Expert indices tensor must be contiguous"
+
+    M_total, K = inputs.shape
+    assert M_total % group_size_m == 0, (
+        f"M_total ({M_total}) must be a multiple of group_size_m ({group_size_m})"
+    )
+
+    if expert_indices.dtype != torch.int32:
+        expert_indices = expert_indices.to(torch.int32)
+
+    num_experts, N, K_weights = expert_weights.shape
+    assert K == K_weights, f"Input K ({K}) must match weight K ({K_weights})"
+    assert expert_indices.shape[0] == M_total, (
+        "Expert indices length must match M_total"
+    )
+
+    output = torch.empty((M_total, N), device=inputs.device, dtype=torch.bfloat16)
+
+    NUM_SMS = torch.cuda.get_device_properties("cuda").multi_processor_count
+
+    grid = (NUM_SMS, 1, 1)
+    _kernel_cg_persistent_forward[grid](
+        inputs,
+        expert_weights,
+        output,
+        expert_indices,
+        M_TOTAL=M_total,
+        N=N,
+        K=K,
+        NUM_EXPERTS=num_experts,
+        GROUP_SIZE_M=group_size_m,
+        NUM_SMS=NUM_SMS,
+    )
+
+    return output
+
+
+def cg_grouped_gemm_forward_dynamic(
+    inputs: torch.Tensor,
+    expert_weights: torch.Tensor,
+    expert_indices: torch.Tensor,
+    group_size_m: int = GROUP_SIZE_M,
+) -> torch.Tensor:
+    """Contiguous grouped GEMM forward pass for MoE with autotuned launch."""
+
+    assert inputs.is_contiguous(), "Input tensor must be contiguous"
+    assert expert_weights.is_contiguous(), "Expert weights tensor must be contiguous"
+    assert expert_indices.is_contiguous(), "Expert indices tensor must be contiguous"
+
+    M_total, K = inputs.shape
+    assert M_total % group_size_m == 0, (
+        f"M_total ({M_total}) must be a multiple of group_size_m ({group_size_m})"
+    )
+
+    if expert_indices.dtype != torch.int32:
+        expert_indices = expert_indices.to(torch.int32)
+
+    num_experts, N, K_weights = expert_weights.shape
+    assert K == K_weights, f"Input K ({K}) must match weight K ({K_weights})"
+    assert expert_indices.shape[0] == M_total, (
+        "Expert indices length must match M_total"
+    )
+
+    output = torch.empty((M_total, N), device=inputs.device, dtype=inputs.dtype)
+
+    grid = lambda meta: (
+        triton.cdiv(M_total, meta["BLOCK_SIZE_M"])
+        * triton.cdiv(N, meta["BLOCK_SIZE_N"]),
+    )
+
+    _kernel_cg_forward_aligned[grid](
+        inputs,
+        expert_weights,
+        output,
+        expert_indices,
+        M_TOTAL=M_total,
+        N=N,
+        K=K,
+        NUM_EXPERTS=num_experts,
+        GROUP_SIZE_M=group_size_m,
+    )
+
+    return output
+
+
+class ContiguousGroupedGEMM(torch.autograd.Function):
+    """Autograd function for contiguous grouped GEMM forward pass only."""
+
+    @staticmethod
+    def forward(ctx, inputs, expert_weights, expert_indices, group_size_m=GROUP_SIZE_M):
+        return cg_grouped_gemm_forward(
+            inputs=inputs,
+            expert_weights=expert_weights,
+            expert_indices=expert_indices,
+            group_size_m=group_size_m,
+        )
+
+    @staticmethod
+    def backward(ctx, grad_output):  # pragma: no cover - not implemented
+        raise NotImplementedError("Backward pass not implemented")
+
+
+def cg_grouped_gemm(
+    inputs: torch.Tensor,
+    expert_weights: torch.Tensor,
+    expert_indices: torch.Tensor,
+    group_size_m: int = GROUP_SIZE_M,
+) -> torch.Tensor:
+    """Convenience wrapper for the forward-only autograd function."""
+
+    if expert_indices.dtype != torch.int32:
+        expert_indices = expert_indices.to(torch.int32)
+
+    return ContiguousGroupedGEMM.apply(
+        inputs, expert_weights, expert_indices, group_size_m
+    )
--- a/src/axolotl/kernels/moe/tt_cg_gemm/cg_reference.py
+++ b/src/axolotl/kernels/moe/tt_cg_gemm/cg_reference.py
@@ -0,0 +1,31 @@
+"""Reference implementation for contiguous grouped GEMM."""
+
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch
+
+
+def pytorch_reference(
+    inputs: torch.Tensor,
+    expert_weights: torch.Tensor,
+    expert_indices: torch.Tensor,
+    group_size_m: int = 128,
+) -> torch.Tensor:
+    """Simple PyTorch implementation for verification."""
+
+    M_total, K = inputs.shape
+    num_experts, N, _ = expert_weights.shape
+
+    output = torch.empty((M_total, N), device=inputs.device, dtype=inputs.dtype)
+
+    for i in range(0, M_total, group_size_m):
+        end_idx = min(i + group_size_m, M_total)
+        expert_idx = expert_indices[i].item()
+        expert_weight = expert_weights[expert_idx]
+        output[i:end_idx] = torch.matmul(inputs[i:end_idx], expert_weight.T)
+
+    return output
--- a/src/axolotl/kernels/moe/tt_cg_gemm/tma_cuda_autotune.py
+++ b/src/axolotl/kernels/moe/tt_cg_gemm/tma_cuda_autotune.py
@@ -0,0 +1,209 @@
+"""Autotuning utilities for Triton contiguous grouped GEMM kernels."""
+
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+from typing import Dict
+
+import torch
+import triton
+import triton.language as tl
+from triton.runtime import driver
+
+
+class CudaUtils:
+    """Helper utilities for CUDA specific Triton features."""
+
+    @staticmethod
+    def is_cuda() -> bool:
+        return driver.active.get_current_target().backend == "cuda"
+
+    @staticmethod
+    def verify_tma() -> bool:
+        return (
+            CudaUtils.is_cuda()
+            and torch.cuda.is_available()
+            and torch.cuda.get_device_capability()[0] >= 9
+        )
+
+    @staticmethod
+    def get_num_sms() -> int:
+        if not CudaUtils.is_cuda():
+            raise RuntimeError("Triton is not running on CUDA backend")
+        if not torch.cuda.is_available():
+            raise RuntimeError("CUDA is not available")
+        return torch.cuda.get_device_properties("cuda").multi_processor_count
+
+
+class TmaDescriptorHelper:
+    """Helper class for managing TMA descriptors in Triton kernels."""
+
+    class KernelParamWrapper:
+        def __init__(self, desc: torch.Tensor):
+            self.desc = desc
+
+        def tma_desc_cpu_ptr(self) -> int:
+            return self.desc.data_ptr()
+
+    def __init__(self, tma_size: int = 128):
+        if not CudaUtils.verify_tma():
+            raise RuntimeError(
+                "TMA not supported on this device (requires Hopper or newer)"
+            )
+        if "nv_tma_desc_type" not in dir(tl):
+            raise RuntimeError(
+                "TMA grid constant descriptors not supported in your Triton version"
+            )
+
+        self.tma_size = tma_size
+        self.fill_1d_tma_descriptor_inner = driver.active.utils.fill_1d_tma_descriptor
+        self.fill_2d_tma_descriptor_inner = driver.active.utils.fill_2d_tma_descriptor
+        self.descriptors: Dict[str, torch.Tensor] = {}
+
+    def init_tma_descriptor(self, name: str) -> None:
+        self.descriptors[name] = torch.empty(
+            self.tma_size, device="cpu", dtype=torch.int8
+        )
+
+    def fill_1d_tma_descriptor(
+        self, name: str, ptr: int, dim: int, block_dim: int, element_size: int
+    ) -> None:
+        if name not in self.descriptors:
+            raise ValueError(f"TMA descriptor '{name}' not initialized")
+
+        desc_x = self.descriptors[name]
+        if desc_x.data_ptr() % 64 != 0:
+            raise ValueError("TMA descriptor must be 64-byte aligned")
+        self.fill_1d_tma_descriptor_inner(
+            ptr, dim, block_dim, element_size, desc_x.data_ptr()
+        )
+
+    def fill_2d_tma_descriptor(
+        self,
+        name: str,
+        ptr: int,
+        dim1: int,
+        dim0: int,
+        block_dim1: int,
+        block_dim0: int,
+        element_size: int,
+    ) -> None:
+        if name not in self.descriptors:
+            raise ValueError(f"TMA descriptor '{name}' not initialized")
+
+        desc_x = self.descriptors[name]
+        if desc_x.data_ptr() % 64 != 0:
+            raise ValueError("TMA descriptor must be 64-byte aligned")
+        self.fill_2d_tma_descriptor_inner(
+            ptr, dim1, dim0, block_dim1, block_dim0, element_size, desc_x.data_ptr()
+        )
+
+    def get_tma_descriptor_kernel_param(
+        self, name: str
+    ) -> "TmaDescriptorHelper.KernelParamWrapper":
+        if name not in self.descriptors or self.descriptors[name] is None:
+            raise ValueError(f"TMA descriptor '{name}' not initialized")
+        return self.KernelParamWrapper(self.descriptors[name])
+
+
+HOPPER_CONFIGS = [
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 32},
+        num_stages=2,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 64, "BLOCK_SIZE_N": 256, "BLOCK_SIZE_K": 32},
+        num_stages=3,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 256, "BLOCK_SIZE_N": 64, "BLOCK_SIZE_K": 32},
+        num_stages=3,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 64, "BLOCK_SIZE_K": 32},
+        num_stages=4,
+        num_warps=4,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 64, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 32},
+        num_stages=4,
+        num_warps=4,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 64},
+        num_stages=3,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 64, "BLOCK_SIZE_N": 256, "BLOCK_SIZE_K": 64},
+        num_stages=3,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 256, "BLOCK_SIZE_N": 64, "BLOCK_SIZE_K": 64},
+        num_stages=3,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 256, "BLOCK_SIZE_K": 64},
+        num_stages=4,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 256, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 64},
+        num_stages=4,
+        num_warps=8,
+    ),
+]
+
+
+STANDARD_CONFIGS = [
+    triton.Config(
+        {"BLOCK_SIZE_M": 64, "BLOCK_SIZE_N": 64, "BLOCK_SIZE_K": 32},
+        num_stages=2,
+        num_warps=4,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 64, "BLOCK_SIZE_K": 32},
+        num_stages=2,
+        num_warps=4,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 32},
+        num_stages=2,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 64},
+        num_stages=3,
+        num_warps=8,
+    ),
+    triton.Config(
+        {"BLOCK_SIZE_M": 128, "BLOCK_SIZE_N": 128, "BLOCK_SIZE_K": 64},
+        num_stages=4,
+        num_warps=8,
+    ),
+]
+
+
+def early_config_prune(configs, args, **kwargs):
+    """Filter out configurations that would exceed shared memory capacity."""
+    k = kwargs.get("K", 0)
+    valid_configs = [
+        config for config in configs if config.kwargs.get("BLOCK_SIZE_K", 0) <= k
+    ]
+    if not valid_configs and configs:
+        return [
+            min(
+                configs,
+                key=lambda c: c.kwargs.get("BLOCK_SIZE_K", float("inf")),
+            )
+        ]
+
+    return valid_configs
--- a/src/axolotl/kernels/moe/tt_mg_gemm/init.py
+++ b/src/axolotl/kernels/moe/tt_mg_gemm/init.py
@@ -0,0 +1,13 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+from .mg_grouped_gemm import grouped_gemm_forward
+from .tma_autotuning import ALIGN_SIZE_M
+
+__all__ = [
+    "grouped_gemm_forward",
+    "ALIGN_SIZE_M",
+]
--- a/src/axolotl/kernels/moe/tt_mg_gemm/mg_grouped_gemm.py
+++ b/src/axolotl/kernels/moe/tt_mg_gemm/mg_grouped_gemm.py
@@ -0,0 +1,761 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+# credit - flat index forward kernel is derived from FBGemm:
+# https://github.com/pytorch/FBGEMM/blob/main/fbgemm_gpu/experimental/gemm/triton_gemm
+
+# pyre-unsafe
+import logging
+from typing import Tuple
+
+import torch
+import triton
+import triton.language as tl
+
+from .tma_autotuning import (
+    _NV_CONFIGS,
+    CudaUtils,
+    early_config_prune,
+)
+
+# Configure logging
+logging.basicConfig(
+    level=logging.INFO, format="%(asctime)s - %(levelname)s - %(message)s"
+)
+
+
+_allocator_registered = False
+
+
+def _torch_allocator(size: int, alignment: int, stream) -> torch.Tensor:
+    return torch.empty(size, device="cuda", dtype=torch.int8)
+
+
+def _ensure_triton_allocator() -> None:
+    global _allocator_registered
+    if not _allocator_registered:
+        triton.set_allocator(_torch_allocator)
+        _allocator_registered = True
+
+
+# ==============  Start Triton Kernels ===============
+
+
+@triton.autotune(
+    configs=_NV_CONFIGS,
+    key=["G", "M_BUCKET", "N", "K"],
+    prune_configs_by={"early_config_prune": early_config_prune},
+)
+@triton.jit
+def _kernel_mg_forward_hopper(
+    a_ptr,
+    b_ptr,
+    c_ptr,
+    m_sizes,
+    M_TOTAL,
+    # problem sizes
+    G: tl.constexpr,
+    M_BUCKET: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    # config
+    NUM_SMS: tl.constexpr,
+    USE_EPILOGUE_SUBTILING: tl.constexpr,
+    # tiles
+    BLOCK_SIZE_M: tl.constexpr,
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+) -> None:
+    """Flat index style forward kernel for Hopper using tensor descriptors."""
+    tbidx = tl.program_id(0)
+
+    c_dtype = c_ptr.dtype.element_ty
+    n_size = N // G
+
+    a_desc = tl.make_tensor_descriptor(
+        a_ptr,
+        shape=[M_TOTAL, K],
+        strides=[K, 1],
+        block_shape=[BLOCK_SIZE_M, BLOCK_SIZE_K],
+    )
+    b_desc = tl.make_tensor_descriptor(
+        b_ptr,
+        shape=[N, K],
+        strides=[K, 1],
+        block_shape=[BLOCK_SIZE_N, BLOCK_SIZE_K],
+    )
+
+    M_end = tl.full([], 0, dtype=tl.int32)
+    processed_tiles = 0
+
+    for g in range(G):
+        M_start = M_end
+        m_size = tl.load(m_sizes + g)
+        M_end = M_start + m_size
+
+        if m_size > 0:
+            num_m_tiles = tl.cdiv(m_size, BLOCK_SIZE_M)
+            num_n_tiles = tl.cdiv(n_size, BLOCK_SIZE_N)
+            group_num_tiles = num_m_tiles * num_n_tiles
+
+            while (
+                tbidx >= processed_tiles and tbidx < processed_tiles + group_num_tiles
+            ):
+                group_index = tbidx - processed_tiles
+
+                tile_m_index = group_index % num_m_tiles
+                tile_n_index = group_index // num_m_tiles
+
+                rows_remaining = m_size - tile_m_index * BLOCK_SIZE_M
+                rows_remaining = tl.maximum(rows_remaining, 0)
+                row_mask = tl.arange(0, BLOCK_SIZE_M) < rows_remaining
+
+                cols_remaining = n_size - tile_n_index * BLOCK_SIZE_N
+                col_mask = tl.arange(0, BLOCK_SIZE_N) < cols_remaining
+
+                accumulator = tl.zeros((BLOCK_SIZE_M, BLOCK_SIZE_N), dtype=tl.float32)
+
+                m_offset = (M_start + tile_m_index * BLOCK_SIZE_M).to(tl.int32)
+                n_offset = (tile_n_index * BLOCK_SIZE_N).to(tl.int32)
+                global_n_offset = (g * n_size + n_offset).to(tl.int32)
+
+                for k_offset in range(0, K, BLOCK_SIZE_K):
+                    k_remaining = K - k_offset
+                    k_mask = tl.arange(0, BLOCK_SIZE_K) < k_remaining
+
+                    a = a_desc.load([m_offset, k_offset])
+                    a_mask = row_mask[:, None] & k_mask[None, :]
+                    a = tl.where(a_mask, a, tl.zeros_like(a))
+
+                    b = b_desc.load([global_n_offset, k_offset])
+                    b_mask = col_mask[:, None] & k_mask[None, :]
+                    b = tl.where(b_mask, b, tl.zeros_like(b))
+
+                    accumulator += tl.dot(a, b.T)
+
+                local_m_offset = (tile_m_index * BLOCK_SIZE_M).to(tl.int32)
+
+                local_row_offsets = local_m_offset + tl.arange(0, BLOCK_SIZE_M)
+                row_store_mask = local_row_offsets < m_size
+                global_row = (M_start + local_row_offsets).to(tl.int32)
+
+                local_col_offsets = tile_n_index * BLOCK_SIZE_N + tl.arange(
+                    0, BLOCK_SIZE_N
+                )
+                col_store_mask = local_col_offsets < n_size
+
+                store_mask = row_store_mask[:, None] & col_store_mask[None, :]
+
+                if USE_EPILOGUE_SUBTILING:
+                    acc = tl.reshape(accumulator, (BLOCK_SIZE_M, 2, BLOCK_SIZE_N // 2))
+                    acc = tl.permute(acc, (0, 2, 1))
+                    acc0, acc1 = tl.split(acc)
+
+                    col_offsets0 = local_col_offsets[: BLOCK_SIZE_N // 2]
+                    col_mask0 = col_store_mask[: BLOCK_SIZE_N // 2]
+                    ptr0 = c_ptr + global_row[:, None] * n_size + col_offsets0[None, :]
+                    tl.store(
+                        ptr0,
+                        acc0.to(c_dtype),
+                        mask=row_store_mask[:, None] & col_mask0[None, :],
+                    )
+
+                    col_offsets1 = local_col_offsets[BLOCK_SIZE_N // 2 :]
+                    col_mask1 = col_store_mask[BLOCK_SIZE_N // 2 :]
+                    ptr1 = c_ptr + global_row[:, None] * n_size + col_offsets1[None, :]
+                    tl.store(
+                        ptr1,
+                        acc1.to(c_dtype),
+                        mask=row_store_mask[:, None] & col_mask1[None, :],
+                    )
+                else:
+                    ptr = (
+                        c_ptr
+                        + global_row[:, None] * n_size
+                        + local_col_offsets[None, :]
+                    )
+                    tl.store(ptr, accumulator.to(c_dtype), mask=store_mask)
+
+                tbidx += NUM_SMS
+
+            processed_tiles += group_num_tiles
+
+
+"""
+Backward pass for grouped GEMM with Triton, where grouping is M*G
+We compute gradients with respect to both input (`grad_x`) and weights (`grad_w`).
+"""
+
+
+# ---- dx flat linear indexed ----
+@triton.autotune(
+    configs=_NV_CONFIGS,
+    key=["G", "M_BUCKET", "N", "K"],
+    prune_configs_by={"early_config_prune": early_config_prune},
+)
+@triton.jit
+def _kernel_mg_dx_tma(
+    grad_output_ptr,
+    w_ptr,
+    grad_input_ptr,
+    m_sizes,
+    M_TOTAL,
+    # problem sizes
+    G: tl.constexpr,
+    M_BUCKET: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    # config
+    NUM_SMS: tl.constexpr,
+    # tiles
+    BLOCK_SIZE_M: tl.constexpr,
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+) -> None:
+    """Compute grad_input = grad_output @ w using tensor descriptors."""
+    tbidx = tl.program_id(0)
+
+    c_dtype = grad_input_ptr.dtype.element_ty
+
+    grad_output_desc = tl.make_tensor_descriptor(
+        grad_output_ptr,
+        shape=[M_TOTAL, N],
+        strides=[N, 1],
+        block_shape=[BLOCK_SIZE_M, BLOCK_SIZE_N],
+    )
+    w_desc = tl.make_tensor_descriptor(
+        w_ptr,
+        shape=[N, K],
+        strides=[K, 1],
+        block_shape=[BLOCK_SIZE_N, BLOCK_SIZE_K],
+    )
+
+    M_end = tl.full([], 0, dtype=tl.int32)
+    processed_tiles = 0
+
+    for g in range(G):
+        M_start = M_end
+        m_size = tl.load(m_sizes + g)
+        M_end = M_start + m_size
+
+        if m_size > 0:
+            num_m_tiles = tl.cdiv(m_size, BLOCK_SIZE_M)
+            num_k_tiles = tl.cdiv(K, BLOCK_SIZE_K)
+            group_num_tiles = num_m_tiles * num_k_tiles
+
+            while (
+                tbidx >= processed_tiles and tbidx < processed_tiles + group_num_tiles
+            ):
+                group_index = tbidx - processed_tiles
+
+                tile_m_index = group_index % num_m_tiles
+                tile_k_index = group_index // num_m_tiles
+
+                rows_remaining = m_size - tile_m_index * BLOCK_SIZE_M
+                rows_remaining = tl.maximum(rows_remaining, 0)
+                row_mask = tl.arange(0, BLOCK_SIZE_M) < rows_remaining
+
+                k_offset = tile_k_index * BLOCK_SIZE_K
+                k_remaining_total = K - k_offset
+                k_mask = tl.arange(0, BLOCK_SIZE_K) < k_remaining_total
+
+                accumulator = tl.zeros((BLOCK_SIZE_M, BLOCK_SIZE_K), dtype=tl.float32)
+
+                m_offset = (M_start + tile_m_index * BLOCK_SIZE_M).to(tl.int32)
+
+                for n_offset in range(0, N, BLOCK_SIZE_N):
+                    n_remaining = N - n_offset
+                    n_mask = tl.arange(0, BLOCK_SIZE_N) < n_remaining
+
+                    grad_y = grad_output_desc.load([m_offset, n_offset])
+                    grad_y_mask = row_mask[:, None] & n_mask[None, :]
+                    grad_y = tl.where(grad_y_mask, grad_y, tl.zeros_like(grad_y))
+
+                    w_tile = w_desc.load([n_offset, k_offset])
+                    w_mask = n_mask[:, None] & k_mask[None, :]
+                    w_tile = tl.where(w_mask, w_tile, tl.zeros_like(w_tile))
+
+                    accumulator += tl.dot(grad_y, w_tile)
+
+                local_row_offsets = tile_m_index * BLOCK_SIZE_M + tl.arange(
+                    0, BLOCK_SIZE_M
+                )
+                row_store_mask = local_row_offsets < m_size
+                global_row = (M_start + local_row_offsets).to(tl.int32)
+
+                col_offsets = k_offset + tl.arange(0, BLOCK_SIZE_K)
+                col_store_mask = col_offsets < K
+
+                store_mask = row_store_mask[:, None] & col_store_mask[None, :]
+
+                ptr = grad_input_ptr + global_row[:, None] * K + col_offsets[None, :]
+                tl.store(ptr, accumulator.to(c_dtype), mask=store_mask)
+
+                tbidx += NUM_SMS
+
+            processed_tiles += group_num_tiles
+
+
+@triton.autotune(
+    configs=_NV_CONFIGS,
+    key=["G", "M_BUCKET", "N", "K"],
+    prune_configs_by={"early_config_prune": early_config_prune},
+)
+@triton.jit
+def _kernel_mg_dw_tma(
+    x_ptr,
+    grad_output_ptr,
+    grad_weight_ptr,
+    m_sizes,
+    M_TOTAL,
+    # problem sizes
+    G: tl.constexpr,
+    M_BUCKET: tl.constexpr,
+    N: tl.constexpr,
+    K: tl.constexpr,
+    # config
+    NUM_SMS: tl.constexpr,
+    # tiles
+    BLOCK_SIZE_N: tl.constexpr,
+    BLOCK_SIZE_K: tl.constexpr,
+    BLOCK_SIZE_M: tl.constexpr,
+) -> None:
+    """Compute grad_weight = grad_output.T @ x using tensor descriptors."""
+    tbidx = tl.program_id(0)
+
+    c_dtype = grad_weight_ptr.dtype.element_ty
+
+    x_desc = tl.make_tensor_descriptor(
+        x_ptr,
+        shape=[M_TOTAL, K],
+        strides=[K, 1],
+        block_shape=[BLOCK_SIZE_M, BLOCK_SIZE_K],
+    )
+    grad_output_desc = tl.make_tensor_descriptor(
+        grad_output_ptr,
+        shape=[M_TOTAL, N],
+        strides=[N, 1],
+        block_shape=[BLOCK_SIZE_M, BLOCK_SIZE_N],
+    )
+
+    num_n_tiles = tl.cdiv(N, BLOCK_SIZE_N)
+    num_k_tiles = tl.cdiv(K, BLOCK_SIZE_K)
+    total_tiles = num_n_tiles * num_k_tiles
+
+    for tile_idx in range(tbidx, total_tiles, NUM_SMS):
+        tile_n_idx = tile_idx % num_n_tiles
+        tile_k_idx = tile_idx // num_n_tiles
+
+        n_offset = tile_n_idx * BLOCK_SIZE_N
+        n_remaining = N - n_offset
+        n_mask = tl.arange(0, BLOCK_SIZE_N) < n_remaining
+
+        k_offset = tile_k_idx * BLOCK_SIZE_K
+        k_remaining = K - k_offset
+        k_mask = tl.arange(0, BLOCK_SIZE_K) < k_remaining
+
+        accumulator = tl.zeros((BLOCK_SIZE_N, BLOCK_SIZE_K), dtype=tl.float32)
+
+        M_end = tl.full([], 0, dtype=tl.int32)
+        for g in range(G):
+            M_start = M_end
+            m_size = tl.load(m_sizes + g)
+            M_end = M_start + m_size
+
+            if m_size > 0:
+                for m_offset_local in range(0, m_size, BLOCK_SIZE_M):
+                    rows_remaining = m_size - m_offset_local
+                    rows_remaining = tl.maximum(rows_remaining, 0)
+                    row_mask = tl.arange(0, BLOCK_SIZE_M) < rows_remaining
+
+                    m_offset = (M_start + m_offset_local).to(tl.int32)
+
+                    x_block = x_desc.load([m_offset, k_offset])
+                    x_mask = row_mask[:, None] & k_mask[None, :]
+                    x_block = tl.where(x_mask, x_block, tl.zeros_like(x_block))
+
+                    grad_block = grad_output_desc.load([m_offset, n_offset])
+                    grad_mask = row_mask[:, None] & n_mask[None, :]
+                    grad_block = tl.where(
+                        grad_mask, grad_block, tl.zeros_like(grad_block)
+                    )
+
+                    contribution = tl.dot(
+                        grad_block.to(tl.float32).T,
+                        x_block.to(tl.float32),
+                    )
+                    accumulator += contribution
+
+        row_offsets = n_offset + tl.arange(0, BLOCK_SIZE_N)
+        row_store_mask = row_offsets < N
+
+        col_offsets = k_offset + tl.arange(0, BLOCK_SIZE_K)
+        col_store_mask = col_offsets < K
+
+        store_mask = row_store_mask[:, None] & col_store_mask[None, :]
+
+        ptr = grad_weight_ptr + row_offsets[:, None] * K + col_offsets[None, :]
+        tl.store(ptr, accumulator.to(c_dtype), mask=store_mask)
+
+
+# ======== End Triton kernels ========
+# ======== End Triton kernels ========
+
+# ======== Triton wrapper functions ========
+
+# ----- main forward pass wrapper -----
+
+
+def grouped_gemm_forward(
+    x: torch.Tensor,
+    w: torch.Tensor,
+    m_sizes: torch.Tensor,
+    tma_size: int = 128,
+    using_fp8: bool = False,
+) -> torch.Tensor:
+    """Grouped GEMM forward using Hopper TMA kernels."""
+    _ensure_triton_allocator()
+    if not CudaUtils.verify_tma():
+        raise NotImplementedError("Grouped GEMM without TMA is not supported yet")
+    if using_fp8:
+        raise NotImplementedError(
+            "FP8 path not implemented with the new Triton API yet"
+        )
+
+    G = m_sizes.shape[0]
+
+    assert x.is_contiguous()
+    assert w.is_contiguous()
+    assert m_sizes.is_contiguous()
+
+    M_total, K = x.shape
+    N = w.shape[0]
+    assert K == w.shape[1], f"Input K ({K}) must match weight K ({w.shape[1]})"
+
+    y = torch.empty((M_total, N // G), device=x.device, dtype=x.dtype)
+    if M_total == 0:
+        return y
+
+    NUM_SMS = CudaUtils.get_num_sms()
+    USE_EPILOGUE_SUBTILING = False
+
+    def grid(_meta):
+        return (NUM_SMS,)
+
+    M_BUCKET = triton.next_power_of_2(M_total)
+    _kernel_mg_forward_hopper[grid](
+        x,
+        w,
+        y,
+        m_sizes,
+        M_total,
+        G,
+        M_BUCKET,
+        N,
+        K,
+        NUM_SMS,
+        USE_EPILOGUE_SUBTILING=USE_EPILOGUE_SUBTILING,
+    )
+    return y
+
+
+# ======== Improved Backward =============
+def grouped_gemm_backward(
+    grad_output: torch.Tensor,
+    x: torch.Tensor,
+    w: torch.Tensor,
+    m_sizes: torch.Tensor,
+    use_tma: bool = True,
+    tma_size: int = 128,
+) -> Tuple[torch.Tensor, torch.Tensor]:
+    """
+    Unified backward pass for grouped GeMM with M*G grouping.
+    Uses optimized TMA-based implementations for both dx and dw when available.
+
+    Args:
+        grad_output: Gradient of output, shape [M_total, N]
+        x: Input tensor from forward pass, shape [M_total, K]
+        w: Weight tensor from forward pass, shape [N, K]
+        m_sizes: Group sizes tensor, shape [G]
+        use_tma: Whether to try using TMA acceleration (if available)
+        tma_size: Size of TMA descriptor in bytes
+
+
+    Returns:
+        Tuple of gradients with respect to x and w: (grad_x, grad_w)
+    """
+    logging.info("Starting unified grouped_gemm_backward")
+
+    # do this once, seems expensive
+    NUM_SMS = CudaUtils.get_num_sms()
+
+    # Basic validation
+    M_total, K_x = x.shape
+    M_grad, N = grad_output.shape
+    N_w, K_w = w.shape
+
+    # Check dimensions
+    if K_x != K_w:
+        raise ValueError(f"K dimension mismatch: x has K={K_x}, w has K={K_w}")
+    if M_total != M_grad:
+        raise ValueError(
+            f"M dimension mismatch: x has M={M_total}, grad_output has M={M_grad}"
+        )
+
+    # Check total M matches sum of group sizes
+    sum_m_sizes = m_sizes.sum().item()
+    if M_total != sum_m_sizes:
+        raise ValueError(
+            f"Sum of m_sizes ({sum_m_sizes}) must match M_total ({M_total})"
+        )
+
+    # Make sure inputs are contiguous
+    grad_output = grad_output.contiguous()
+    x = x.contiguous()
+    w = w.contiguous()
+    m_sizes = m_sizes.contiguous()
+
+    # Check TMA support
+    if use_tma and not CudaUtils.verify_tma():
+        logging.info("TMA requested but not supported on this device")
+        use_tma = False
+
+    # Compute grad_x using flat linear implementation
+    try:
+        logging.info("Computing grad_x with flat linear kernel")
+
+        # Use TMA-optimized implementation
+        grad_x = grouped_gemm_dx_tma(
+            grad_output=grad_output,
+            w=w,
+            m_sizes=m_sizes,
+            num_sms=NUM_SMS,
+            tma_size=tma_size,
+        )
+
+    except Exception as e:
+        logging.error(f"Error in grad_x computation: {e}")
+        raise
+
+    # Compute grad_w using flat linear style implementation
+    try:
+        logging.info("Computing grad_w with flat linear kernel")
+
+        grad_w = grouped_gemm_dw_tma(
+            x, grad_output, m_sizes, num_sms=NUM_SMS, tma_size=tma_size
+        )
+    except Exception as e:
+        logging.error(f"Error in grad_w computation: {e}")
+        raise
+
+    return grad_x, grad_w
+
+
+# ----- dx backward pass wrapper -----
+
+
+def grouped_gemm_dx_tma(
+    grad_output: torch.Tensor,
+    w: torch.Tensor,
+    m_sizes: torch.Tensor,
+    num_sms: int = 132,
+    tma_size: int = 128,
+) -> torch.Tensor:
+    """Compute grad_x using the Hopper grouped GEMM kernel."""
+    _ensure_triton_allocator()
+    if not CudaUtils.verify_tma():
+        raise NotImplementedError("Optimized dx computation requires TMA support")
+
+    grad_output = grad_output.contiguous()
+    w = w.contiguous()
+    m_sizes = m_sizes.contiguous()
+
+    M_total, N = grad_output.shape
+    N_w, K = w.shape
+    if N != N_w:
+        raise ValueError(f"Grad_output N ({N}) must match weight N ({N_w})")
+
+    if m_sizes.sum().item() != M_total:
+        raise ValueError("Sum of m_sizes must equal the number of rows in grad_output")
+
+    grad_x = torch.empty(
+        (M_total, K), device=grad_output.device, dtype=grad_output.dtype
+    )
+
+    NUM_SMS = num_sms
+
+    def grid(_meta):
+        return (NUM_SMS,)
+
+    M_BUCKET = triton.next_power_of_2(M_total)
+    _kernel_mg_dx_tma[grid](
+        grad_output,
+        w,
+        grad_x,
+        m_sizes,
+        M_total,
+        m_sizes.shape[0],
+        M_BUCKET,
+        N,
+        K,
+        NUM_SMS,
+    )
+    return grad_x
+
+
+# ======== dw wrapper function ==========
+
+
+def grouped_gemm_dw_tma(
+    x: torch.Tensor,
+    grad_output: torch.Tensor,
+    m_sizes: torch.Tensor,
+    num_sms: int = 132,
+    tma_size: int = 128,
+) -> torch.Tensor:
+    """Compute grad_w using the Hopper grouped GEMM kernel."""
+    _ensure_triton_allocator()
+    if not CudaUtils.verify_tma():
+        raise RuntimeError("TMA grouped GEMM requested on a device without TMA support")
+
+    x = x.contiguous()
+    grad_output = grad_output.contiguous()
+    m_sizes = m_sizes.contiguous()
+
+    M_total, K = x.shape
+    M_grad, N = grad_output.shape
+    if M_total != M_grad:
+        raise ValueError("x and grad_output must have matching batch dimension")
+    if m_sizes.sum().item() != M_total:
+        raise ValueError("Sum of m_sizes must equal the number of rows in the inputs")
+
+    grad_w = torch.zeros((N, K), device=x.device, dtype=x.dtype)
+
+    NUM_SMS = num_sms
+
+    def grid(_meta):
+        return (NUM_SMS,)
+
+    M_BUCKET = triton.next_power_of_2(M_total)
+    _kernel_mg_dw_tma[grid](
+        x,
+        grad_output,
+        grad_w,
+        m_sizes,
+        M_total,
+        m_sizes.shape[0],
+        M_BUCKET,
+        N,
+        K,
+        NUM_SMS,
+    )
+    return grad_w
+
+
+# ======== End Backwards Wrapper Functions =============
+
+# ======== PyTorch wrapper functions ========
+
+
+class GroupedGemmMg(torch.autograd.Function):
+    """
+    Autograd function for GroupedGEMM with M*G grouping.
+    Supports both standard and FP8 quantized operations.
+    """
+
+    @staticmethod
+    def forward(ctx, x, w, m_sizes, use_tma=True, tma_size=128, using_fp8=False):
+        """
+        Forward pass of GroupedGEMM.
+
+        Args:
+            x: Input tensor, shape [M_total, K]
+            w: Weight tensor, shape [N, K]
+            m_sizes: Tensor of shape [G] containing the size of each group
+            use_tma: Whether to try using TMA acceleration (if available)
+            tma_size: Size of TMA descriptor in bytes
+            using_fp8: Whether to use FP8 quantization
+
+        Returns:
+            Output tensor, shape [M_total, N]
+        """
+
+        # Use regular forward without quantization
+        output = grouped_gemm_forward(
+            x=x, w=w, m_sizes=m_sizes, tma_size=tma_size, using_fp8=False
+        )
+
+        # Save inputs and parameters for backward pass
+        ctx.save_for_backward(x, w, m_sizes)
+        ctx.use_tma = use_tma
+        ctx.tma_size = tma_size
+
+        ctx.save_for_backward(x, w, m_sizes)
+
+        return output
+
+    @staticmethod
+    def backward(ctx, grad_output):
+        """
+        Backward pass of M*G GroupedGEMM.
+
+        Args:
+            grad_output: Gradient of output, shape [M_total, N]
+
+        Returns:
+            Tuple of gradients:
+                - grad_x: Gradient with respect to x, shape [M_total, K]
+                - grad_w: Gradient with respect to w, shape [N, K]
+                - None: Gradient with respect to m_sizes (not differentiable)
+                - None: Gradient with respect to use_tma (not differentiable)
+                - None: Gradient with respect to tma_size (not differentiable)
+
+        """
+        # Retrieve saved tensors and parameters
+
+        x, w, m_sizes = ctx.saved_tensors
+
+        use_tma = ctx.use_tma
+        tma_size = ctx.tma_size
+
+        # Compute gradients using the unified implementation
+        grad_x, grad_w = grouped_gemm_backward(
+            grad_output=grad_output,
+            x=x,
+            w=w,
+            m_sizes=m_sizes,
+            use_tma=use_tma,
+            tma_size=tma_size,
+        )
+
+        # Return gradients for all inputs (None for non-differentiable parameters)
+        return grad_x, grad_w, None, None, None, None
+
+
+def mg_grouped_gemm(
+    x: torch.Tensor,
+    w: torch.Tensor,
+    m_sizes: torch.Tensor,
+    use_tma: bool = True,
+    tma_size: int = 128,
+    using_fp8: bool = False,
+) -> torch.Tensor:
+    """
+    Unified differentiable grouped GEMM operation for M*G grouped GEMM.
+    Supports both standard precision and FP8 quantized operations.
+
+    Args:
+        x: Input tensor, shape [M_total, K]
+        w: Weight tensor, shape [N, K]
+        m_sizes: Tensor of shape [G] containing the size of each group
+        use_tma: Whether to try using TMA acceleration (if available)
+        tma_size: Size of TMA descriptor in bytes
+        using_fp8: Whether to use FP8 quantization
+
+    Returns:
+        Output tensor, shape [M_total, N]
+    """
+    return GroupedGemmMg.apply(x, w, m_sizes, use_tma, tma_size, using_fp8)
--- a/src/axolotl/kernels/moe/tt_mg_gemm/tma_autotuning.py
+++ b/src/axolotl/kernels/moe/tt_mg_gemm/tma_autotuning.py
@@ -0,0 +1,232 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+
+# credit - TMAHelper class, AutoTuning are derived from FBGemm:
+# https://github.com/pytorch/FBGEMM/blob/main/fbgemm_gpu/experimental/gemm/triton_gemm
+
+# pyre-unsafe
+
+import os
+import sys
+from typing import Dict
+
+import torch
+import triton
+from triton.runtime import driver  # @manual
+
+sys.path.append(os.path.dirname(os.path.abspath(__file__)))
+
+
+# ===== Supporting utils, CUDA and TMA =====
+
+
+class CudaUtils:
+    @staticmethod
+    def is_cuda() -> bool:
+        """Check if Triton is running on CUDA backend."""
+        return driver.active.get_current_target().backend == "cuda"
+
+    @staticmethod
+    def verify_tma() -> bool:
+        """Check if TMA is supported on the current device."""
+        return (
+            CudaUtils.is_cuda()
+            and torch.cuda.is_available()
+            and torch.cuda.get_device_capability()[0] >= 9
+        )
+
+    @staticmethod
+    def get_num_sms() -> int:
+        """Get the number of streaming multiprocessors on the current device."""
+        if not CudaUtils.is_cuda():
+            raise RuntimeError("Triton is not running on CUDA backend")
+        if not torch.cuda.is_available():
+            raise RuntimeError("CUDA is not available")
+        return torch.cuda.get_device_properties("cuda").multi_processor_count
+
+
+class TmaDescriptorHelper:
+    """Helper class for managing TMA descriptors in Triton kernels.
+
+    Args:
+        tma_size: Size of the TMA descriptor in bytes
+    """
+
+    class KernelParamWrapper:
+        """Wrapper to implement the TmaDescKernelParam interface."""
+
+        def __init__(self, desc: torch.Tensor):
+            self.desc = desc
+
+        def tma_desc_cpu_ptr(self) -> int:
+            """Return the CPU pointer to the TMA descriptor."""
+            return self.desc.data_ptr()
+
+    def __init__(self, tma_size: int = 128):
+        if not CudaUtils.verify_tma():
+            raise RuntimeError(
+                "TMA not supported on this device (requires Hopper or newer)"
+            )
+
+        self.tma_size = tma_size
+        self.fill_1d_tma_descriptor_inner = driver.active.utils.fill_tma_descriptor
+        self.fill_2d_tma_descriptor_inner = driver.active.utils.fill_tma_descriptor
+        self.descriptors: Dict[str, torch.Tensor] = {}
+
+    def init_tma_descriptor(self, name: str) -> None:
+        """Initialize a TMA descriptor with the given name.
+
+        Call this method outside of the lambda function for grid size.
+        """
+        self.descriptors[name] = torch.empty(
+            self.tma_size, device="cpu", dtype=torch.int8
+        )
+
+    def fill_1d_tma_descriptor(
+        self, name: str, ptr: int, dim: int, block_dim: int, element_size: int
+    ) -> None:
+        """Fill a 1D TMA descriptor.
+
+        Call this method inside the lambda function for grid size.
+        """
+        if name not in self.descriptors:
+            raise ValueError(f"TMA descriptor '{name}' not initialized")
+
+        desc_x = self.descriptors[name]
+        if desc_x.data_ptr() % 64 != 0:
+            raise ValueError("TMA descriptor must be 64-byte aligned")
+        self.fill_1d_tma_descriptor_inner(
+            ptr, dim, block_dim, element_size, desc_x.data_ptr()
+        )
+
+    def fill_2d_tma_descriptor(
+        self,
+        name: str,
+        ptr: int,
+        dim1: int,
+        dim0: int,
+        block_dim1: int,
+        block_dim0: int,
+        element_size: int,
+    ) -> None:
+        """Fill a 2D TMA descriptor.
+
+        Call this method inside the lambda function for grid size.
+        """
+        if name not in self.descriptors:
+            raise ValueError(f"TMA descriptor '{name}' not initialized")
+
+        desc_x = self.descriptors[name]
+        if desc_x.data_ptr() % 64 != 0:
+            raise ValueError("TMA descriptor must be 64-byte aligned")
+        self.fill_2d_tma_descriptor_inner(
+            ptr, dim1, dim0, block_dim1, block_dim0, element_size, desc_x.data_ptr()
+        )
+
+    def get_tma_descriptor_kernel_param(self, name: str) -> KernelParamWrapper:
+        """Get the TMA descriptor kernel parameter for the given name."""
+        if name not in self.descriptors or self.descriptors[name] is None:
+            raise ValueError(f"TMA descriptor '{name}' not initialized")
+        return self.KernelParamWrapper(self.descriptors[name])
+
+
+# ======  Autotuning utilities ======
+ALIGN_SIZE_M = 128
+
+_NV_CONFIGS = [
+    triton.Config(
+        {
+            "BLOCK_SIZE_M": block_size_m,
+            "BLOCK_SIZE_N": block_size_n,
+            "BLOCK_SIZE_K": block_size_k,
+        },
+        num_stages=num_stages,
+        num_warps=num_warps,
+        num_ctas=num_ctas,
+    )
+    for block_size_m in [
+        ALIGN_SIZE_M,
+    ]
+    for block_size_n in [64, 128, 256]
+    for block_size_k in [64, 128, 256]
+    for num_stages in [3, 4]
+    for num_warps in [4, 8]
+    for num_ctas in [1]
+]
+
+
+def early_config_prune(configs, named_args, dtsize=None, dtype=None, **kwargs):
+    device = torch.cuda.current_device()
+    # Check for all possible pointer parameter names
+    if "grad_input_ptr" in named_args:
+        ptr_name = "grad_input_ptr"
+    elif "c_ptr" in named_args:
+        ptr_name = "c_ptr"
+    elif "grad_weight_ptr" in named_args:
+        ptr_name = "grad_weight_ptr"
+    else:
+        raise KeyError("No recognized pointer parameter found in kernel arguments")
+
+    if dtsize is None:
+        dtsize = named_args[ptr_name].element_size()
+    if dtype is None:
+        dtype = named_args[ptr_name].dtype
+
+    pruned_configs = []
+    for config in configs:
+        kw = config.kwargs
+        BLOCK_M, BLOCK_N, BLOCK_K, num_stages = (
+            kw["BLOCK_SIZE_M"],
+            kw["BLOCK_SIZE_N"],
+            kw["BLOCK_SIZE_K"],
+            config.num_stages,
+        )
+        G, M, N, K = (
+            named_args["G"],
+            named_args["M_BUCKET"],
+            named_args["N"],
+            named_args["K"],
+        )
+
+        # 1. make sure we have enough smem
+        max_shared_memory = driver.active.utils.get_device_properties(device)[
+            "max_shared_mem"
+        ]
+
+        required_shared_memory = (BLOCK_M + BLOCK_N) * BLOCK_K * num_stages * dtsize
+        if required_shared_memory > max_shared_memory:
+            continue
+
+        M_PER_GROUP = M // G
+        MIN_M_TILES = 64
+        # 2. make sure we don't load M tiles that are too big
+        if BLOCK_M > MIN_M_TILES and BLOCK_M > (M_PER_GROUP * 2):
+            continue
+        # 3. make sure we don't load N tiles that are too small
+        if BLOCK_M < 128 and BLOCK_M < (M_PER_GROUP // 2):
+            continue
+
+        num_sm = driver.active.utils.get_device_properties(device)[
+            "multiprocessor_count"
+        ]
+        N_TILES = N // BLOCK_N
+        MIN_N_TILES = 64
+        # 4. make sure we don't load N tiles that are too big
+        if BLOCK_N > MIN_N_TILES and M * N_TILES < num_sm:
+            continue
+        # 5. make sure we don't load N tiles that are too small
+        if BLOCK_N < 128 and M * N_TILES > 2 * num_sm:
+            continue
+        # 6. make sure K can be evenly divided
+        if K % BLOCK_K != 0:
+            continue
+
+        pruned_configs.append(config)
+
+    return pruned_configs
+
+
+# ======== End Autotuning utilities ========
--- a/src/axolotl/loaders/model.py
+++ b/src/axolotl/loaders/model.py
@@ -631,7 +631,7 @@ class ModelLoader:
            if is_causal_conv1d_available():
                raise ImportError(
                    "The 'causal-conv1d' package is installed but causes compatibility issues with LFM2 models. "
-                    "Please uninstall it by running: `uv pip uninstall -y causal-conv1d`"
+                    "Please uninstall it by running: `pip uninstall -y causal-conv1d`"
                )

    def _configure_zero3_memory_efficient_loading(
--- a/src/axolotl/loaders/patch_manager.py
+++ b/src/axolotl/loaders/patch_manager.py
@@ -190,6 +190,15 @@ class PatchManager:

            apply_mistral_tokenizer_image_patch()

+        if self.cfg.moe_kernels and self.cfg.model_config_type == "deepseek_v3":
+            from axolotl.monkeypatch.deepseek_v3 import patch_deepseek_v3_moe
+
+            patch_deepseek_v3_moe(backend=self.cfg.moe_kernel_backend)
+        elif self.cfg.model_config_type == "deepseek_v3" and not self.cfg.moe_kernels:
+            LOG.info(
+                "Skipping DeepSeek V3 Triton MoE kernels; enable with `moe_kernels: true`"
+            )
+
    def _apply_fp8_patches(self):
        """Apply patches for FP8 support."""
        if self.cfg.fp8:
--- a/src/axolotl/models/mamba/init.py
+++ b/src/axolotl/models/mamba/init.py
@@ -9,7 +9,7 @@ def check_mamba_ssm_installed():
    mamba_ssm_spec = importlib.util.find_spec("mamba_ssm")
    if mamba_ssm_spec is None:
        raise ImportError(
-            "MambaLMHeadModel requires mamba_ssm. Please install it with `uv pip install -e .[mamba-ssm]`"
+            "MambaLMHeadModel requires mamba_ssm. Please install it with `pip install -e .[mamba-ssm]`"
        )


--- a/src/axolotl/monkeypatch/accelerate/fsdp2.py
+++ b/src/axolotl/monkeypatch/accelerate/fsdp2.py
@@ -4,7 +4,6 @@ monkeypatch for accelerate fsdp2 fix when modifying ordereddict during interatio

 import copy
 import functools
-import os
 import sys

 import torch
@@ -128,8 +127,7 @@ def get_state_dict(self, model, unwrap=True):
            if model.zero_gather_16bit_weights_on_model_save():
                if tp_sharding and not compare_versions("deepspeed", ">=", "0.16.4"):
                    raise ImportError(
-                        "Deepspeed TP requires deepspeed >= 0.16.4. Update DeepSpeed via "
-                        "`uv pip install -U deepspeed`."
+                        "Deepspeed TP requires deepspeed >= 0.16.4, Please update DeepSpeed via `pip install deepspeed -U`."
                    )
                state_dict = (
                    model._consolidated_16bit_state_dict()
@@ -279,11 +277,6 @@ def fsdp2_prepare_model(accelerator, model: torch.nn.Module) -> torch.nn.Module:

    mesh = getattr(accelerator.state, "device_mesh", None)

-    # Disable memory pinning if requested
-    offload_to_cpu = isinstance(fsdp2_plugin.cpu_offload, CPUOffloadPolicy)
-    if offload_to_cpu and os.environ.get("FSDP_CPU_OFFLOAD_PIN_MEMORY", "") == "false":
-        fsdp2_plugin.cpu_offload.pin_memory = False
-
    fsdp2_kwargs = {
        "reshard_after_forward": fsdp2_plugin.reshard_after_forward,
        "offload_policy": fsdp2_plugin.cpu_offload,
@@ -348,6 +341,7 @@ def fsdp2_prepare_model(accelerator, model: torch.nn.Module) -> torch.nn.Module:
        )

    if fsdp2_plugin.cpu_ram_efficient_loading:
+        offload_to_cpu = isinstance(fsdp2_plugin.cpu_offload, CPUOffloadPolicy)
        fsdp2_load_full_state_dict(
            accelerator, model, original_sd, offload_to_cpu=offload_to_cpu
        )
--- a/src/axolotl/monkeypatch/deepseek_v3/init.py
+++ b/src/axolotl/monkeypatch/deepseek_v3/init.py
@@ -0,0 +1,401 @@
+"""Monkeypatches for DeepSeek V3 MoE to use Triton contiguous grouped GEMM kernels."""
+
+from __future__ import annotations
+
+from typing import Callable
+
+import torch
+
+from axolotl.kernels.moe import ContiguousGroupedGEMM
+from axolotl.kernels.moe.indices import generate_permute_indices
+from axolotl.kernels.moe.tt_mg_gemm import grouped_gemm_forward as mg_grouped_gemm
+from axolotl.utils.logging import get_logger
+
+_GROUP_SIZE_M = 128
+_COMBINED_SUBMODULES = ("gate_proj", "up_proj", "down_proj")
+
+LOG = get_logger(__name__)
+
+
+def _is_triton_eligible(hidden_states: torch.Tensor) -> bool:
+    if not hidden_states.is_cuda or hidden_states.shape[0] == 0:
+        return False
+    major, _ = torch.cuda.get_device_capability(hidden_states.device)
+    if major < 9:
+        LOG.debug(
+            "Skipping Triton MoE kernels: requires compute capability >= 90, found %s",
+            major,
+        )
+        return False
+    return True
+
+
+def _ensure_combined_expert_weights(
+    module, dtype: torch.dtype, device: torch.device, backend: str
+) -> None:
+    if not hasattr(module, "_axolotl_original_specs"):
+        module._axolotl_original_specs = {}
+    if not hasattr(module, "_axolotl_mg_shapes"):
+        module._axolotl_mg_shapes = {}
+
+    prev_backend = getattr(module, "_axolotl_combined_backend", None)
+    if getattr(module, "_axolotl_combined_weights", False):
+        if prev_backend != backend:
+            _restore_expert_weights(module)
+        else:
+            for name in _COMBINED_SUBMODULES:
+                param_name = f"{name}_weight"
+                param = module.get_parameter(param_name)
+                if param.device != device or param.dtype != dtype:
+                    module._parameters[param_name] = torch.nn.Parameter(
+                        param.to(device=device, dtype=dtype).contiguous()
+                    )
+            module._axolotl_combined_dtype = dtype
+            module._axolotl_combined_device = device
+            module._axolotl_combined_backend = backend
+            return
+
+    module._axolotl_mg_shapes = {}
+    for name in _COMBINED_SUBMODULES:
+        weights = []
+        orig_device = None
+        orig_dtype = None
+        orig_shape = None
+        for expert in module.experts:
+            lin = expert.get_submodule(name)
+            weight_param = lin._parameters.get("weight")
+            if weight_param is None:
+                raise RuntimeError("Expected expert linear layers to have weights")
+            if orig_device is None:
+                orig_device = weight_param.device
+                orig_dtype = weight_param.dtype
+                orig_shape = tuple(weight_param.shape)
+            weights.append(weight_param.detach().to(device=device, dtype=dtype))
+            if "weight" in lin._parameters:
+                del lin._parameters["weight"]
+            if "bias" in lin._parameters:
+                del lin._parameters["bias"]
+        if backend == "cg":
+            combined_weight = torch.stack(weights, dim=0).contiguous()
+        else:
+            combined_weight = torch.cat(weights, dim=0).contiguous()
+            module._axolotl_mg_shapes[name] = orig_shape
+        module.register_parameter(f"{name}_weight", torch.nn.Parameter(combined_weight))
+        module._axolotl_original_specs[name] = (orig_device, orig_dtype, orig_shape)
+
+    module._axolotl_combined_weights = True
+    module._axolotl_combined_dtype = dtype
+    module._axolotl_combined_device = device
+    module._axolotl_combined_backend = backend
+
+
+def _restore_expert_weights(module) -> None:
+    if not getattr(module, "_axolotl_combined_weights", False):
+        return
+
+    for name in _COMBINED_SUBMODULES:
+        param_name = f"{name}_weight"
+        combined = module._parameters.pop(param_name)
+        orig_device, orig_dtype, orig_shape = module._axolotl_original_specs.get(
+            name, (combined.device, combined.dtype, None)
+        )
+        rows_per = orig_shape[0] if orig_shape else None
+        for idx, expert in enumerate(module.experts):
+            lin = expert.get_submodule(name)
+            if combined.dim() == 3:
+                slice_tensor = combined[idx]
+            elif rows_per is not None:
+                start = idx * rows_per
+                end = start + rows_per
+                slice_tensor = combined[start:end]
+            else:
+                raise RuntimeError(
+                    "Unable to recover expert weight shape during restore"
+                )
+            lin._parameters["weight"] = torch.nn.Parameter(
+                slice_tensor.detach().clone().to(orig_device, dtype=orig_dtype)
+            )
+
+    module._axolotl_combined_weights = False
+    module._axolotl_combined_dtype = None
+    module._axolotl_combined_device = None
+    module._axolotl_combined_backend = None
+    module._axolotl_original_specs = {}
+    module._axolotl_mg_shapes = {}
+
+
+def _run_cg_grouped_gemm(
+    module,
+    grouped_hidden: torch.Tensor,
+    m_sizes: torch.Tensor,
+    num_experts: int,
+    group_size_m: int,
+    hidden_dtype: torch.dtype,
+    device: torch.device,
+) -> tuple[torch.Tensor, torch.Tensor, torch.Tensor, torch.Tensor]:
+    _ensure_combined_expert_weights(module, hidden_dtype, device, backend="cg")
+
+    expert_index_tensor = torch.repeat_interleave(
+        torch.arange(num_experts, device=device, dtype=torch.int32),
+        m_sizes.to(torch.int64),
+    )
+
+    gate_weights = module.get_parameter("gate_proj_weight")
+    if gate_weights.dim() == 2:
+        out_dim = gate_weights.shape[0] // num_experts
+        gate_weights = gate_weights.view(num_experts, out_dim, gate_weights.shape[1])
+
+    up_weights = module.get_parameter("up_proj_weight")
+    if up_weights.dim() == 2:
+        out_dim = up_weights.shape[0] // num_experts
+        up_weights = up_weights.view(num_experts, out_dim, up_weights.shape[1])
+
+    down_weights = module.get_parameter("down_proj_weight")
+    if down_weights.dim() == 2:
+        out_dim = down_weights.shape[0] // num_experts
+        down_weights = down_weights.view(num_experts, out_dim, down_weights.shape[1])
+
+    gate_out = ContiguousGroupedGEMM.apply(
+        grouped_hidden,
+        gate_weights,
+        expert_index_tensor,
+        group_size_m,
+    )
+    up_out = ContiguousGroupedGEMM.apply(
+        grouped_hidden,
+        up_weights,
+        expert_index_tensor,
+        group_size_m,
+    )
+    return (
+        gate_out.to(hidden_dtype),
+        up_out.to(hidden_dtype),
+        down_weights,
+        expert_index_tensor,
+    )
+
+    gate_out = mg_grouped_gemm(
+        grouped_hidden,
+        module.get_parameter("gate_proj_weight"),
+        m_sizes_tensor,
+    )
+    up_out = mg_grouped_gemm(
+        grouped_hidden,
+        module.get_parameter("up_proj_weight"),
+        m_sizes_tensor,
+    )
+    down_out = mg_grouped_gemm(
+        hidden_grouped,
+        module.get_parameter("down_proj_weight"),
+        m_sizes_tensor,
+    )
+
+    return (
+        gate_out.to(hidden_dtype),
+        up_out.to(hidden_dtype),
+        down_out.to(hidden_dtype),
+    )
+
+
+def _moe_triton_forward(
+    module,
+    hidden_states: torch.Tensor,
+    topk_indices: torch.Tensor,
+    topk_weights: torch.Tensor,
+    group_size_m: int,
+    backend: str,
+    fallback: Callable[[torch.Tensor, torch.Tensor, torch.Tensor], torch.Tensor],
+) -> torch.Tensor:
+    if not _is_triton_eligible(hidden_states):
+        return fallback(hidden_states, topk_indices, topk_weights)
+
+    device = hidden_states.device
+    hidden_dtype = hidden_states.dtype
+    num_tokens, hidden_dim = hidden_states.shape
+    top_k = topk_indices.size(-1)
+
+    expanded_hidden = hidden_states.repeat_interleave(top_k, dim=0)
+    expert_assignments = topk_indices.reshape(-1)
+    if expanded_hidden.numel() == 0:
+        return hidden_states.new_zeros_like(hidden_states)
+
+    sort_perm = torch.argsort(expert_assignments)
+    sorted_hidden = expanded_hidden.index_select(0, sort_perm)
+    sorted_assignments = expert_assignments.index_select(0, sort_perm)
+
+    num_experts = len(module.experts)
+    counts = torch.bincount(sorted_assignments, minlength=num_experts)
+    total_actual = int(counts.sum().item())
+    if total_actual == 0:
+        return hidden_states.new_zeros_like(hidden_states)
+
+    if not getattr(module, "_axolotl_triton_logged", False):
+        min_tokens = int(counts.min().item())
+        max_tokens = int(counts.max().item())
+        LOG.info(
+            "DeepseekV3MoE Triton: tokens per expert (min=%s, max=%s, avg=%.1f) with group_size=%s",
+            min_tokens,
+            max_tokens,
+            total_actual / max(1, num_experts),
+            group_size_m,
+        )
+        module._axolotl_triton_logged = True
+
+    counts_int = counts.to(torch.int32)
+    aligned_counts = (
+        (torch.clamp_min(counts_int, group_size_m) + group_size_m - 1) // group_size_m
+    ) * group_size_m
+    max_len = int(aligned_counts.sum().item())
+
+    permuted_indices, m_sizes, _ = generate_permute_indices(
+        counts_int.to(device),
+        experts_per_rank=num_experts,
+        num_ranks=1,
+        max_len=max_len,
+        alignment=group_size_m,
+        use_cpu=not hidden_states.is_cuda,
+    )
+
+    permuted_indices = permuted_indices.to(device)
+    m_sizes = m_sizes.to(device)
+
+    permuted_indices_long = permuted_indices.to(torch.int64)
+    valid_mask = permuted_indices_long >= 0
+    valid_positions = torch.nonzero(valid_mask, as_tuple=False).squeeze(-1)
+    source_indices = permuted_indices_long[valid_mask]
+    padded_positions = torch.nonzero(~valid_mask, as_tuple=False).squeeze(-1)
+
+    grouped_hidden = hidden_states.new_empty((max_len, hidden_dim))
+    if valid_positions.numel() > 0:
+        grouped_hidden.index_copy_(
+            0,
+            valid_positions,
+            sorted_hidden.index_select(0, source_indices),
+        )
+    if valid_positions.numel() < max_len:
+        grouped_hidden.index_fill_(0, padded_positions, 0)
+
+    m_sizes_tensor = m_sizes.to(device=device, dtype=torch.int32)
+
+    if backend == "mg":
+        _ensure_combined_expert_weights(module, hidden_dtype, device, backend)
+        gate_out = mg_grouped_gemm(
+            grouped_hidden,
+            module.get_parameter("gate_proj_weight"),
+            m_sizes_tensor,
+        ).to(hidden_dtype)
+        up_out = mg_grouped_gemm(
+            grouped_hidden,
+            module.get_parameter("up_proj_weight"),
+            m_sizes_tensor,
+        ).to(hidden_dtype)
+    else:
+        gate_out, up_out, down_weights, expert_index_tensor = _run_cg_grouped_gemm(
+            module,
+            grouped_hidden,
+            m_sizes,
+            num_experts,
+            group_size_m,
+            hidden_dtype,
+            device,
+        )
+
+    act_fn: Callable[[torch.Tensor], torch.Tensor] = module.experts[0].act_fn
+    if valid_positions.numel() > 0:
+        gate_valid = gate_out.index_select(0, valid_positions)
+        up_valid = up_out.index_select(0, valid_positions)
+        hidden_concat = act_fn(gate_valid) * up_valid
+    else:
+        hidden_concat = torch.empty(
+            (0, gate_out.shape[-1]), device=device, dtype=hidden_dtype
+        )
+
+    intermediate_dim = hidden_concat.shape[-1]
+    hidden_grouped = hidden_states.new_empty((max_len, intermediate_dim))
+    if valid_positions.numel() > 0:
+        hidden_grouped.index_copy_(0, valid_positions, hidden_concat)
+    if valid_positions.numel() < max_len:
+        hidden_grouped.index_fill_(0, padded_positions, 0)
+
+    if backend == "mg":
+        down_out = mg_grouped_gemm(
+            hidden_grouped,
+            module.get_parameter("down_proj_weight"),
+            m_sizes_tensor,
+        ).to(hidden_dtype)
+    else:
+        down_out = ContiguousGroupedGEMM.apply(
+            hidden_grouped,
+            down_weights,
+            expert_index_tensor,
+            group_size_m,
+        ).to(hidden_dtype)
+
+    if valid_positions.numel() > 0:
+        down_valid = down_out.index_select(0, valid_positions)
+    else:
+        down_valid = torch.empty(
+            (0, down_out.shape[-1]), device=device, dtype=hidden_dtype
+        )
+
+    sorted_outputs = hidden_states.new_zeros((total_actual, hidden_dim))
+    if down_valid.numel() > 0:
+        sorted_outputs.index_copy_(0, source_indices, down_valid)
+
+    expanded_output = expanded_hidden.new_empty(expanded_hidden.shape)
+    expanded_output.index_copy_(0, sort_perm, sorted_outputs)
+    expert_outputs = expanded_output.view(num_tokens, top_k, hidden_dim)
+
+    weighted = expert_outputs * topk_weights.unsqueeze(-1).to(hidden_dtype)
+    return weighted.sum(dim=1)
+
+
+def patch_deepseek_v3_moe(
+    group_size_m: int = _GROUP_SIZE_M, backend: str = "mg"
+) -> None:
+    """Patch HuggingFace DeepseekV3MoE to use Triton contiguous group GEMM kernels."""
+
+    from transformers.models.deepseek_v3.modeling_deepseek_v3 import DeepseekV3MoE
+
+    if backend not in {"cg", "mg"}:
+        raise ValueError(f"Unsupported MoE kernel backend: {backend}")
+
+    # Record the unpatched implementation so callers can access a true baseline even
+    # after the Triton patch has been applied (e.g. repeated microbenchmarks).
+    if not hasattr(DeepseekV3MoE, "_axolotl_triton_original_moe"):
+        DeepseekV3MoE._axolotl_triton_original_moe = DeepseekV3MoE.moe
+
+    if getattr(DeepseekV3MoE, "_axolotl_triton_patch", False):
+        return
+
+    original_moe = DeepseekV3MoE._axolotl_triton_original_moe
+    DeepseekV3MoE._axolotl_triton_backend = backend
+    DeepseekV3MoE._axolotl_group_size_m = group_size_m
+
+    def patched_moe(self, hidden_states, topk_indices, topk_weights):
+        backend_sel = getattr(self, "_axolotl_triton_backend", backend)
+        group_size_sel = getattr(self, "_axolotl_group_size_m", group_size_m)
+        if backend_sel == "cg" and group_size_sel != _GROUP_SIZE_M:
+            LOG.debug(
+                "Adjusting group_size_m=%s to %s for CG backend",
+                group_size_sel,
+                _GROUP_SIZE_M,
+            )
+            group_size_sel = _GROUP_SIZE_M
+        try:
+            return _moe_triton_forward(
+                self,
+                hidden_states,
+                topk_indices,
+                topk_weights,
+                group_size_sel,
+                backend_sel,
+                original_moe,
+            )
+        except Exception as err:  # surface Triton failures explicitly
+            _restore_expert_weights(self)
+            LOG.error("DeepseekV3MoE Triton path failed: %s", err)
+            raise
+
+    DeepseekV3MoE.moe = patched_moe
+    DeepseekV3MoE._axolotl_triton_patch = True
--- a/src/axolotl/monkeypatch/llama_attn_hijack_flash.py
+++ b/src/axolotl/monkeypatch/llama_attn_hijack_flash.py
@@ -107,7 +107,7 @@ def patch_llama_rms_norm():
        transformers.models.llama.modeling_llama.LlamaRMSNorm = LlamaRMSNorm
    except ImportError:
        LOG.warning(
-            "optimized flash-attention RMSNorm not found (run `uv pip install 'git+https://github.com/Dao-AILab/flash-attention.git#egg=dropout_layer_norm&subdirectory=csrc/layer_norm'`)"
+            "optimized flash-attention RMSNorm not found (run `pip install 'git+https://github.com/Dao-AILab/flash-attention.git#egg=dropout_layer_norm&subdirectory=csrc/layer_norm'`)"
        )


--- a/src/axolotl/utils/model_shard_quant.py
+++ b/src/axolotl/utils/model_shard_quant.py
@@ -148,7 +148,7 @@ def load_sharded_model(
        model = AutoModelForCausalLM.from_pretrained(
            model_name,
            use_cache=False,
-            dtype=torch.float32,
+            torch_dtype=torch.float32,
            _attn_implementation=model_config._attn_implementation,
            trust_remote_code=cfg.trust_remote_code,
        )
@@ -158,7 +158,7 @@ def load_sharded_model(
        with init_empty_weights():
            model = AutoModelForCausalLM.from_config(
                model_config,
-                dtype=torch_dtype,
+                torch_dtype=torch_dtype,
                trust_remote_code=cfg.trust_remote_code,
            )
    return model
--- a/src/axolotl/utils/schemas/config.py
+++ b/src/axolotl/utils/schemas/config.py
@@ -113,6 +113,19 @@ class AxolotlInputConfig(
        },
    )

+    moe_kernels: bool | None = Field(
+        default=None,
+        json_schema_extra={
+            "description": "Enable Axolotl's vendored MoE kernels when supported (e.g., DeepSeek V3)"
+        },
+    )
+    moe_kernel_backend: Literal["cg", "mg"] | None = Field(
+        default="mg",
+        json_schema_extra={
+            "description": "Grouped GEMM backend to use when `moe_kernels` is enabled. `mg` selects the Hopper TMA kernel; `cg` selects the contiguous kernel."
+        },
+    )
+
    trainer_cls: str | None = Field(
        default=None,
        json_schema_extra={
--- a/src/axolotl/utils/schemas/validation.py
+++ b/src/axolotl/utils/schemas/validation.py
@@ -497,9 +497,7 @@ class TrainingValidationMixin:

            if importlib.util.find_spec("mistral_common") is None:
                raise ImportError(
-                    "mistral-common is required for mistral models. "
-                    "Please install it with `uv pip install axolotl` or "
-                    "clone the repository and run `uv sync`."
+                    "mistral-common is required for mistral models. Please install it with `pip install axolotl` or `pip install -e .`."
                )

        return tokenizer_use_mistral_common
@@ -818,22 +816,21 @@ class OptimizationValidationMixin:
            )
        return data

-    @model_validator(mode="before")
-    @classmethod
-    def check_fsdp2_cpu_offload_pin_memory(cls, data):
-        if not (fsdp_config := data.get("fsdp_config")):
-            return data
-
-        if fsdp_config.get("cpu_offload_pin_memory") is False:
-            if str(data.get("fsdp_version")) != "2":
+    @model_validator(mode="after")
+    def check_fsdp2_base_model_quant_ram_efficient_loading(self):
+        fsdp_config = self.fsdp_config if hasattr(self, "fsdp_config") else None
+        fsdp_version = self.fsdp_version if hasattr(self, "fsdp_version") else None
+        load_in_8bit = self.load_in_8bit if hasattr(self, "load_in_8bit") else None
+        load_in_4bit = self.load_in_4bit if hasattr(self, "load_in_4bit") else None
+        if fsdp_config and fsdp_version == 2:
+            if fsdp_config.get("cpu_ram_efficient_loading") and (
+                load_in_8bit or load_in_4bit
+            ):
                raise ValueError(
-                    "FSDP1 does not support disabling cpu_offload_pin_memory, please set `fsdp_version` to 2"
+                    "FSDP2 does not support load_in_8bit or load_in_4bit with cpu_ram_efficient_loading. Please do one of the following: use DeepSpeed, "
+                    "set fsdp_version to 1, or disable cpu_ram_efficient_loading."
                )
-            if not fsdp_config.get("offload_params"):
-                raise ValueError(
-                    "disabling cpu_offload_pin_memory requires enabling offload_params"
-                )
-        return data
+        return self

    @model_validator(mode="before")
    @classmethod
@@ -1348,10 +1345,8 @@ class ComplexValidationMixin:
            except ImportError as exception:
                raise ImportError(
                    "context_parallel_size > 1 but ring_flash_attn is not installed. "
-                    "Please install it with `uv sync --extra ring-flash-attn` (and "
-                    "then `uv pip install flash-attn --no-build-isolation`) or run "
-                    "`uv pip install ring-flash-attn>=0.1.4` followed by "
-                    "`uv pip install flash-attn --no-build-isolation`."
+                    "Please install it with `pip install axolotl[ring-flash-attn] "
+                    "or `pip install ring-flash-attn>=0.1.4`."
                ) from exception

            LOG.warning(
--- a/src/axolotl/utils/tee.py
+++ b/src/axolotl/utils/tee.py
@@ -109,8 +109,8 @@ def prepare_debug_log(cfg, filename: str = "debug.log") -> str:
            cfg.get("resume_from_checkpoint") or cfg.get("auto_resume_from_checkpoints")
        )

-        if not append:
-            log_path.unlink(missing_ok=True)
+        if not append and log_path.exists():
+            log_path.unlink()

        fh = open(log_path, "a", encoding="utf-8")
        fh.flush()
--- a/src/axolotl/utils/trainer.py
+++ b/src/axolotl/utils/trainer.py
@@ -595,10 +595,6 @@ def setup_fsdp_envs(cfg):
        os.environ["FSDP_USE_ORIG_PARAMS"] = "true"
    if cfg.fsdp_config.state_dict_type:
        os.environ["FSDP_STATE_DICT_TYPE"] = cfg.fsdp_config.state_dict_type
-    if cfg.fsdp_config.cpu_offload_pin_memory is not None:
-        os.environ["FSDP_CPU_OFFLOAD_PIN_MEMORY"] = str(
-            cfg.fsdp_config.cpu_offload_pin_memory
-        ).lower()
    if cfg.fsdp_config.auto_wrap_policy:
        os.environ["FSDP_AUTO_WRAP_POLICY"] = cfg.fsdp_config.auto_wrap_policy
    if cfg.fsdp_config.transformer_layer_cls_to_wrap:
--- a/src/setuptools_axolotl_dynamic_dependencies.py
+++ b/src/setuptools_axolotl_dynamic_dependencies.py
@@ -0,0 +1,104 @@
+"""
+dynamic requirements for axolotl
+"""
+
+import platform
+import re
+from importlib.metadata import PackageNotFoundError, version
+
+from setuptools.command.build_py import build_py as _build_py
+
+
+def parse_requirements():
+    _install_requires = []
+    _dependency_links = []
+    with open("./requirements.txt", encoding="utf-8") as requirements_file:
+        lines = [r.strip() for r in requirements_file.readlines()]
+        for line in lines:
+            is_extras = (
+                "flash-attn" in line
+                or "flash-attention" in line
+                or "deepspeed" in line
+                or "mamba-ssm" in line
+                or "lion-pytorch" in line
+            )
+            if line.startswith("--extra-index-url"):
+                # Handle custom index URLs
+                _, url = line.split()
+                _dependency_links.append(url)
+            elif not is_extras and line and line[0] != "#":
+                # Handle standard packages
+                _install_requires.append(line)
+
+    try:
+        xformers_version = [req for req in _install_requires if "xformers" in req][0]
+        torchao_version = [req for req in _install_requires if "torchao" in req][0]
+        autoawq_version = [req for req in _install_requires if "autoawq" in req][0]
+
+        if "Darwin" in platform.system():
+            # don't install xformers on MacOS
+            _install_requires.pop(_install_requires.index(xformers_version))
+        else:
+            # detect the version of torch already installed
+            # and set it so dependencies don't clobber the torch version
+            try:
+                torch_version = version("torch")
+            except PackageNotFoundError:
+                torch_version = "2.5.1"
+            _install_requires.append(f"torch=={torch_version}")
+
+            version_match = re.match(r"^(\d+)\.(\d+)(?:\.(\d+))?", torch_version)
+            if version_match:
+                major, minor, patch = version_match.groups()
+                major, minor = int(major), int(minor)
+                patch = (
+                    int(patch) if patch is not None else 0
+                )  # Default patch to 0 if not present
+            else:
+                raise ValueError("Invalid version format")
+
+            if (major, minor) >= (2, 5):
+                _install_requires.pop(_install_requires.index(xformers_version))
+                if patch == 0:
+                    _install_requires.append("xformers==0.0.28.post2")
+                else:
+                    _install_requires.append("xformers==0.0.28.post3")
+                _install_requires.pop(_install_requires.index(autoawq_version))
+            elif (major, minor) >= (2, 4):
+                if patch == 0:
+                    _install_requires.pop(_install_requires.index(xformers_version))
+                    _install_requires.append("xformers>=0.0.27")
+                else:
+                    _install_requires.pop(_install_requires.index(xformers_version))
+                    _install_requires.append("xformers==0.0.28.post1")
+            elif (major, minor) >= (2, 3):
+                _install_requires.pop(_install_requires.index(torchao_version))
+                if patch == 0:
+                    _install_requires.pop(_install_requires.index(xformers_version))
+                    _install_requires.append("xformers>=0.0.26.post1")
+                else:
+                    _install_requires.pop(_install_requires.index(xformers_version))
+                    _install_requires.append("xformers>=0.0.27")
+            elif (major, minor) >= (2, 2):
+                _install_requires.pop(_install_requires.index(torchao_version))
+                _install_requires.pop(_install_requires.index(xformers_version))
+                _install_requires.append("xformers>=0.0.25.post1")
+            else:
+                _install_requires.pop(_install_requires.index(torchao_version))
+                _install_requires.pop(_install_requires.index(xformers_version))
+                _install_requires.append("xformers>=0.0.23.post1")
+
+    except PackageNotFoundError:
+        pass
+    return _install_requires, _dependency_links
+
+
+class BuildPyCommand(_build_py):
+    """
+    custom build_py command to parse dynamic requirements
+    """
+
+    def finalize_options(self):
+        super().finalize_options()
+        install_requires, _ = parse_requirements()
+        self.distribution.install_requires = install_requires
--- a/tests/cli/test_cli_preprocess.py
+++ b/tests/cli/test_cli_preprocess.py
@@ -14,7 +14,7 @@ def cleanup_last_run_prepared():
    yield

    if Path("last_run_prepared").exists():
-        shutil.rmtree("last_run_prepared", ignore_errors=True)
+        shutil.rmtree("last_run_prepared")


 def test_preprocess_config_not_found(cli_runner):
--- a/tests/e2e/patched/lora_kernels/test_lora_kernel_patching.py
+++ b/tests/e2e/patched/lora_kernels/test_lora_kernel_patching.py
@@ -160,7 +160,7 @@ def test_geglu_model_integration():
    """Test GeGLU activation with Gemma model."""
    model = AutoModelForCausalLM.from_pretrained(
        "trl-internal-testing/tiny-Gemma2ForCausalLM",
-        dtype=torch.float16,
+        torch_dtype=torch.float16,
        device_map="cuda:0",
    )
    peft_config = get_peft_config(
--- a/tests/e2e/patched/test_lora_llama_multipack.py
+++ b/tests/e2e/patched/test_lora_llama_multipack.py
@@ -5,7 +5,7 @@ E2E tests for lora llama
 import unittest

 import pytest
-from transformers.utils import is_gptqmodel_available, is_torch_bf16_gpu_available
+from transformers.utils import is_auto_gptq_available, is_torch_bf16_gpu_available

 from axolotl.common.datasets import load_datasets
 from axolotl.train import train
@@ -69,7 +69,7 @@ class TestLoraLlama(unittest.TestCase):
        train(cfg=cfg, dataset_meta=dataset_meta)
        check_model_output_exists(temp_dir, cfg)

-    @pytest.mark.skipif(not is_gptqmodel_available(), reason="gptqmodel not installed")
+    @pytest.mark.skipif(not is_auto_gptq_available(), reason="auto-gptq not available")
    @with_temp_dir
    def test_lora_gptq_packed(self, temp_dir):
        cfg = DictDefault(
--- a/tests/e2e/test_quantization.py
+++ b/tests/e2e/test_quantization.py
@@ -39,7 +39,7 @@ def model():
    dummy_model = AutoModelForCausalLM.from_pretrained(
        "Qwen/Qwen2-0.5B",
        device_map="auto",
-        dtype=torch.bfloat16,
+        torch_dtype=torch.bfloat16,
    )
    with torch.device(dummy_model.device):
        dummy_model.model.embed_tokens = torch.nn.Embedding(
--- a/tests/utils/schemas/validation/test_fsdp.py
+++ b/tests/utils/schemas/validation/test_fsdp.py
@@ -61,50 +61,12 @@ class TestFSDPValidation:
            },
            fsdp_version=2,
        )
-        validated_cfg = validate_config(cfg)
-        assert validated_cfg.fsdp_version == 2
-        assert validated_cfg.fsdp_config.cpu_ram_efficient_loading is True
-
-    def test_fsdp2_cpu_offload_pin_memory_requires_offload_params(self, min_base_cfg):
-        cfg = min_base_cfg | DictDefault(
-            fsdp_config={
-                "cpu_offload_pin_memory": False,
-                "offload_params": False,
-            },
-            fsdp_version=2,
-        )
        with pytest.raises(
            ValueError,
-            match="disabling cpu_offload_pin_memory requires enabling offload_params",
+            match="FSDP2 does not support load_in_8bit or load_in_4bit with cpu_ram_efficient_loading.",
        ):
            validate_config(cfg)

-    def test_fsdp1_cpu_offload_pin_memory_not_supported(self, min_base_cfg):
-        cfg = min_base_cfg | DictDefault(
-            fsdp_config={
-                "cpu_offload_pin_memory": False,
-                "offload_params": True,
-            },
-            fsdp_version=1,
-        )
-        with pytest.raises(
-            ValueError,
-            match="FSDP1 does not support disabling cpu_offload_pin_memory, please set `fsdp_version` to 2",
-        ):
-            validate_config(cfg)
-
-    def test_fsdp2_cpu_offload_pin_memory_w_offload_params(self, min_base_cfg):
-        cfg = min_base_cfg | DictDefault(
-            fsdp_config={
-                "cpu_offload_pin_memory": False,
-                "offload_params": True,
-            },
-            fsdp_version=2,
-        )
-        validated_cfg = validate_config(cfg)
-        assert validated_cfg.fsdp_config.cpu_offload_pin_memory is False
-        assert validated_cfg.fsdp_config.offload_params is True
-
    def test_fsdp_prefixes_removed(self, min_base_cfg):
        cfg = min_base_cfg | DictDefault(
            fsdp_config={
--- a/uv.lock
+++ b/uv.lock
Author	SHA1	Message	Date
Dan Saunders	dd85358543	default mg	2025-09-25 16:30:23 -04:00
Dan Saunders	55d98db0d0	fix	2025-09-25 16:08:35 -04:00
Dan Saunders	d90ade3b1b	fix	2025-09-25 15:55:08 -04:00
Dan Saunders	824a641cee	uniform routing default	2025-09-25 15:47:23 -04:00
Dan Saunders	e003a05177	narrow sweep; compare both backends	2025-09-25 14:54:03 -04:00
Dan Saunders	91393c4dc8	allocator	2025-09-25 14:27:34 -04:00
Dan Saunders	d578c53603	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	4db7a21ff7	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	3b2e05c563	update to new api	2025-09-25 14:27:34 -04:00
Dan Saunders	1037ca3a97	update to new api	2025-09-25 14:27:34 -04:00
Dan Saunders	6369dcd7b8	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	a81612305c	fix?	2025-09-25 14:27:34 -04:00
Dan Saunders	d0da67eb17	add mg kernel backend	2025-09-25 14:27:34 -04:00
Dan Saunders	8a1f5ae940	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	146ca48cba	vram	2025-09-25 14:27:34 -04:00
Dan Saunders	fd312f6058	dtype	2025-09-25 14:27:34 -04:00
Dan Saunders	ab8fa56b16	dtype	2025-09-25 14:27:34 -04:00
Dan Saunders	1640cd4006	delete config	2025-09-25 14:27:34 -04:00
Dan Saunders	3277d44d71	cfg value	2025-09-25 14:27:34 -04:00
Dan Saunders	d3e1b0ef1a	small deepseek script	2025-09-25 14:27:34 -04:00
Dan Saunders	5b97633faa	Fix	2025-09-25 14:27:34 -04:00
Dan Saunders	94cbc6d42d	log device, dtype	2025-09-25 14:27:34 -04:00
Dan Saunders	493616fc3d	reprod tt table	2025-09-25 14:27:34 -04:00
Dan Saunders	d2b25c7327	grid sweep	2025-09-25 14:27:34 -04:00
Dan Saunders	b670c45276	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	61faf4cbe4	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	8d8fa834a2	sweep	2025-09-25 14:27:34 -04:00
Dan Saunders	9d69c6fb3e	Fix	2025-09-25 14:27:34 -04:00
Dan Saunders	92f2f6e73c	dtype fix	2025-09-25 14:27:34 -04:00
Dan Saunders	e5d2aebe16	uniform routing:	2025-09-25 14:27:34 -04:00
Dan Saunders	4ab9e3f58b	add logs	2025-09-25 14:27:34 -04:00
Dan Saunders	5788832812	simplify	2025-09-25 14:27:34 -04:00
Dan Saunders	db782430f8	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	5c74edeefe	token shuffle kernel	2025-09-25 14:27:34 -04:00
Dan Saunders	18269ee6a9	fix	2025-09-25 14:27:34 -04:00
Dan Saunders	6a45d804f9	glue	2025-09-25 14:27:34 -04:00
Dan Saunders	95e607574a	vendor torchtitan moe kernels	2025-09-25 14:27:34 -04:00