fix: missing key in enum

feat: add custom processing strategy for phi35 vl
feat: add phi_35_vl support
2025-07-03 13:46:16 +08:00 · 2025-07-03 13:46:16 +08:00 · 2025-07-03 13:46:16 +08:00
705 changed files with 18282 additions and 41922 deletions
--- a/.axolotl-complete.bash
+++ b/.axolotl-complete.bash
@@ -1,41 +0,0 @@
-#!/bin/bash
-
-_axolotl_completions() {
-    local cur prev
-    COMPREPLY=()
-    cur="${COMP_WORDS[COMP_CWORD]}"
-    prev="${COMP_WORDS[COMP_CWORD-1]}"
-
-    # If we're completing the first argument (the command)
-    if [[ $COMP_CWORD -eq 1 ]]; then
-        mapfile -t COMPREPLY < <(compgen -W "delinearize-llama4 fetch lm-eval merge-sharded-fsdp-weights quantize vllm-serve evaluate inference merge-lora preprocess train" -- "$cur")
-        return 0
-    fi
-
-    # Commands that should complete with directories and YAML files
-    local -a yaml_commands=("merge-sharded-fsdp-weights" "quantize" "vllm-serve" "evaluate" "inference" "merge-lora" "preprocess" "train")
-
-    # Check if previous word is in our list
-    if [[ " ${yaml_commands[*]} " =~ (^|[[:space:]])$prev($|[[:space:]]) ]]; then
-        # Use filename completion which handles directories properly
-        compopt -o filenames
-        mapfile -t COMPREPLY < <(compgen -f -- "$cur")
-
-        # Filter to only include directories and YAML files
-        local -a filtered=()
-        for item in "${COMPREPLY[@]}"; do
-            if [[ -d "$item" ]] || [[ "$item" == *.yaml ]] || [[ "$item" == *.yml ]]; then
-                filtered+=("$item")
-            fi
-        done
-        COMPREPLY=("${filtered[@]}")
-
-        return 0
-    fi
-
-    # Default: no completion
-    return 0
-}
-
-# Remove the -o nospace option - let filenames handle it
-complete -F _axolotl_completions axolotl
--- a/.bandit
+++ b/.bandit
@@ -1,3 +1,3 @@
 [bandit]
 exclude = tests
-skips = B101,B615,B102,B110
+skips = B101
--- a/.coderabbit.yaml
+++ b/.coderabbit.yaml
@@ -1,17 +0,0 @@
-# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json
-language: "en-US"
-early_access: false
-reviews:
-  profile: "chill"
-  request_changes_workflow: false
-  high_level_summary: true
-  review_status: true
-  collapse_walkthrough: true
-  poem: false
-  sequence_diagrams: false
-  auto_review:
-    enabled: true
-    drafts: false
-    auto_incremental_review: false
-chat:
-  auto_reply: true
--- a/.coveragerc
+++ b/.coveragerc
@@ -2,6 +2,7 @@
 source = axolotl
 omit =
    */tests/*
+    setup.py

 [report]
 exclude_lines =
--- a/.flake8
+++ b/.flake8
@@ -0,0 +1,5 @@
+[flake8]
+max-line-length = 88
+
+select = C,E,F,W,B,B950
+extend-ignore = E203, E501, W503
--- 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
@@ -62,13 +57,6 @@ We welcome ideas for improvements and new features. To suggest an enhancement, o
 5. Push your branch to your fork on GitHub.
 6. Open a new pull request against the `main` branch of the axolotl repository. Include a clear and concise description of your changes, referencing any related issues.

-#### Skipping CI Checks
-
-You can skip certain CI checks by including specific keywords in your commit messages:
-
- `[skip ci]` or `skip ci` - Skips all CI checks for that commit
- `[skip-e2e]` or `skip-e2e` - Skips only end-to-end tests while running other CI checks. You may also include this in the title of your PR to disable end-to-end tests for the entire PR.
-
 ## Style Guidelines

 ### Code Style
--- a/.github/workflows/base.yml
+++ b/.github/workflows/base.yml
@@ -5,19 +5,17 @@ on:
    branches:
      - "main"
    paths:
-      - 'docker/Dockerfile-base'
-      - 'docker/Dockerfile-uv-base'
+      - 'Dockerfile-base'
      - '.github/workflows/base.yml'
  pull_request:
    paths:
-      - 'docker/Dockerfile-base'
-      - 'docker/Dockerfile-uv-base'
+      - 'Dockerfile-base'
      - '.github/workflows/base.yml'
  workflow_dispatch:

 jobs:
  build-base:
-    if: ${{ github.repository_owner == 'axolotl-ai-cloud' && (github.event_name != 'pull_request' || !github.event.pull_request.draft) }}
+    if: github.repository_owner == 'axolotl-ai-cloud'
    timeout-minutes: 480
    # this job needs to be run on self-hosted GPU runners...
    runs-on: ubuntu-latest-m
@@ -25,6 +23,13 @@ jobs:
      fail-fast: false
      matrix:
        include:
+          - cuda: "124"
+            cuda_version: 12.4.1
+            cudnn_version: ""
+            python_version: "3.11"
+            pytorch: 2.5.1
+            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
+            dockerfile: "Dockerfile-base"
          - cuda: "124"
            cuda_version: 12.4.1
            cudnn_version: ""
@@ -47,7 +52,7 @@ jobs:
            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
            dockerfile: "Dockerfile-base"
          - cuda: "128"
-            cuda_version: 12.8.1
+            cuda_version: 12.6.3
            cudnn_version: ""
            python_version: "3.11"
            pytorch: 2.7.1
@@ -57,16 +62,9 @@ jobs:
            cuda_version: 12.8.1
            cudnn_version: ""
            python_version: "3.11"
-            pytorch: 2.8.0
+            pytorch: nightly
            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
-            dockerfile: "Dockerfile-base"
-#          - cuda: "128"
-#            cuda_version: 12.8.1
-#            cudnn_version: ""
-#            python_version: "3.11"
-#            pytorch: nightly
-#            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
-#            dockerfile: "Dockerfile-base-nightly"
+            dockerfile: "Dockerfile-base-nightly"
 #          # "next" is for release candidates of pytorch
 #          - cuda: "128"
 #            cuda_version: 12.8.1
@@ -98,9 +96,7 @@ jobs:
          context: .
          file: ./docker/${{ 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 }}
@@ -110,7 +106,7 @@ jobs:
            PYTORCH_VERSION=${{ matrix.pytorch }}
            TORCH_CUDA_ARCH_LIST=${{ matrix.torch_cuda_arch_list }}
  build-base-uv:
-    if: ${{ github.repository_owner == 'axolotl-ai-cloud' && (github.event_name != 'pull_request' || !github.event.pull_request.draft) }}
+    if: github.repository_owner == 'axolotl-ai-cloud'
    timeout-minutes: 480
    runs-on: ubuntu-latest-m
    strategy:
@@ -124,13 +120,6 @@ 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-uv-base"
-          - cuda: "126"
-            cuda_version: 12.6.3
-            cudnn_version: ""
-            python_version: "3.11"
-            pytorch: 2.7.1
-            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
-            dockerfile: "Dockerfile-uv-base"
          - cuda: "128"
            cuda_version: 12.8.1
            cudnn_version: ""
@@ -138,13 +127,6 @@ jobs:
            pytorch: 2.7.1
            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
            dockerfile: "Dockerfile-uv-base"
-          - cuda: "128"
-            cuda_version: 12.8.1
-            cudnn_version: ""
-            python_version: "3.11"
-            pytorch: 2.8.0
-            torch_cuda_arch_list: "7.0 7.5 8.0 8.6 8.7 8.9 9.0+PTX"
-            dockerfile: "Dockerfile-uv-base"
    steps:
      - name: Checkout
        uses: actions/checkout@v4
--- 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
@@ -3,10 +3,9 @@ on:
  # check on PRs, and manual triggers
  merge_group:
  pull_request:
-      types: [opened, synchronize, reopened, ready_for_review]
      paths:
       - '**.py'
-       - 'pyproject.toml'
+       - 'requirements.txt'
       - '.github/workflows/*.yml'
       - "*.[q]md"
       - "examples/**/*.y[a]?ml"
@@ -17,10 +16,10 @@ jobs:
  pre-commit:
    name: pre-commit
    runs-on: ubuntu-latest
-    if: ${{ !github.event.pull_request.draft }}
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"
+          cache: 'pip' # caching pip dependencies
      - uses: pre-commit/action@v3.0.1
--- a/.github/workflows/main.yml
+++ b/.github/workflows/main.yml
@@ -15,26 +15,25 @@ jobs:
      fail-fast: false
      matrix:
        include:
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            axolotl_extras:
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
            pytorch: 2.6.0
-            axolotl_extras:
+            axolotl_extras: vllm
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
            pytorch: 2.7.1
-            axolotl_extras: vllm
-            is_latest: true
-          - cuda: 128
-            cuda_version: 12.8.1
-            python_version: "3.11"
-            pytorch: 2.7.1
            axolotl_extras:
          - cuda: 128
            cuda_version: 12.8.1
            python_version: "3.11"
-            pytorch: 2.8.0
+            pytorch: 2.7.1
            axolotl_extras:
    runs-on: axolotl-gpu-runner
    steps:
@@ -68,8 +67,6 @@ jobs:
            PYTORCH_VERSION=${{ matrix.pytorch }}
            AXOLOTL_ARGS=${{ matrix.axolotl_args }}
            AXOLOTL_EXTRAS=${{ matrix.axolotl_extras}}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
@@ -85,32 +82,26 @@ jobs:
    strategy:
      matrix:
        include:
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            axolotl_extras:
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
            pytorch: 2.6.0
            axolotl_extras:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            axolotl_extras:
-            is_latest:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            axolotl_extras: vllm
            is_latest: true
-          - cuda: 128
-            cuda_version: 12.8.1
+          - cuda: 126
+            cuda_version: 12.6.3
            python_version: "3.11"
            pytorch: 2.7.1
            axolotl_extras:
          - cuda: 128
            cuda_version: 12.8.1
            python_version: "3.11"
-            pytorch: 2.8.0
+            pytorch: 2.7.1
            axolotl_extras:
    runs-on: axolotl-gpu-runner
    steps:
@@ -140,8 +131,6 @@ jobs:
          build-args: |
            BASE_TAG=${{ 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: ${{ github.event_name != 'pull_request' }}
          tags: |
@@ -161,24 +150,6 @@ jobs:
            python_version: "3.11"
            pytorch: 2.6.0
            axolotl_extras:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            axolotl_extras:
-            is_latest:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            axolotl_extras: vllm
-            is_latest: true
-          - cuda: 128
-            cuda_version: 12.8.1
-            python_version: "3.11"
-            pytorch: 2.8.0
-            axolotl_extras:
-            is_latest:
    runs-on: axolotl-gpu-runner
    steps:
      - name: Checkout
@@ -207,8 +178,6 @@ jobs:
          build-args: |
            BASE_TAG=${{ 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'
@@ -19,7 +21,7 @@ concurrency:

 jobs:
  test-axolotl-multigpu:
-    if: ${{ ! contains(github.event.commits[0].message, '[skip e2e]') && github.repository_owner == 'axolotl-ai-cloud' && (github.event_name != 'pull_request' || !github.event.pull_request.draft) }}
+    if: ${{ ! contains(github.event.commits[0].message, '[skip e2e]') && github.repository_owner == 'axolotl-ai-cloud' }}
    strategy:
      fail-fast: false
      matrix:
@@ -31,18 +33,18 @@ jobs:
            axolotl_extras:
            num_gpus: 2
            nightly_build: "true"
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            axolotl_extras:
+            num_gpus: 2
+            nightly_build: "true"
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
            pytorch: 2.7.1
-            axolotl_extras: vllm
-            num_gpus: 2
-            nightly_build: "true"
-          - cuda: 128
-            cuda_version: 12.8.1
-            python_version: "3.11"
-            pytorch: 2.8.0
-            axolotl_extras: fbgemm-gpu
+            axolotl_extras:
            num_gpus: 2
            nightly_build: "true"
    runs-on: [self-hosted, modal]
@@ -54,17 +56,13 @@ jobs:
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
      - name: Install Modal
        run: |
          python -m pip install --upgrade pip
-          pip install modal==1.0.2 jinja2 protobuf
+          pip install modal==1.0.2 jinja2
      - name: Update env vars
        run: |
-          echo "BASE_TAG=${{ 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 }}" >> $GITHUB_ENV
      - name: Run tests job on Modal
        run: |
-          modal run -m cicd.multigpu
+          modal run cicd.multigpu
--- a/.github/workflows/nightlies.yml
+++ b/.github/workflows/nightlies.yml
@@ -12,16 +12,16 @@ jobs:
      fail-fast: false
      matrix:
        include:
-          - cuda: 126
-            cuda_version: 12.6.3
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            axolotl_extras:
+          - cuda: 124
+            cuda_version: 12.4.1
            python_version: "3.11"
            pytorch: 2.6.0
            axolotl_extras:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            axolotl_extras:
    runs-on: axolotl-gpu-runner
    steps:
      - name: Checkout
@@ -52,8 +52,6 @@ jobs:
            CUDA=${{ matrix.cuda }}
            PYTORCH_VERSION=${{ matrix.pytorch }}
            AXOLOTL_ARGS=${{ matrix.axolotl_args }}
-            GIT_REF=${{ github.ref }}
-            GIT_SHA=${{ github.sha }}
          file: ./docker/Dockerfile
          push: ${{ github.event_name != 'pull_request' }}
          tags: |
@@ -67,16 +65,16 @@ jobs:
    strategy:
      matrix:
        include:
-          - cuda: 126
-            cuda_version: 12.6.3
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            axolotl_extras:
+          - cuda: 124
+            cuda_version: 12.4.1
            python_version: "3.11"
            pytorch: 2.6.0
            axolotl_extras:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            axolotl_extras:
    runs-on: axolotl-gpu-runner
    steps:
      - name: Checkout
@@ -104,8 +102,6 @@ jobs:
          build-args: |
            BASE_TAG=${{ 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
@@ -2,7 +2,7 @@ name: Preview
 on:
  workflow_dispatch:
  pull_request:
-    types: [opened, synchronize, reopened, ready_for_review]
+    types: [opened, synchronize, reopened]

    # Run the workflow only when one of these files changes
    paths:
@@ -25,12 +25,9 @@ permissions:
 jobs:
  preview:
    runs-on: ubuntu-latest
-    if: ${{ !github.event.pull_request.draft }}
    steps:
      - name: Check out repository
        uses: actions/checkout@v4
-        with:
-          ref: ${{ github.event.pull_request.head.sha }}

      - name: Set up Quarto
        uses: quarto-dev/quarto-actions/setup@v2
@@ -40,15 +37,10 @@ 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
@@ -58,12 +50,10 @@ jobs:

      - name: Netlify Publish
        uses: nwtgck/actions-netlify@v3.0
-        if: ${{ github.event.pull_request.head.repo.full_name == github.repository }}
-        id: netlify
        with:
          publish-dir: './_site'
-          enable-pull-request-comment: false
-          enable-github-deployment: false
+          enable-pull-request-comment: true
+          enable-github-deployment: true
          github-token: ${{ secrets.GITHUB_TOKEN }}
          deploy-message: "Deployed On Netlify"
          github-deployment-environment: 'preview'
@@ -71,13 +61,3 @@ jobs:
        env:
          NETLIFY_AUTH_TOKEN: ${{ secrets.NETLIFY_AUTH_TOKEN }}
          NETLIFY_SITE_ID: ${{ secrets.NETLIFY_SITE_ID }}
-
-      - name: Update PR with preview link
-        if: ${{ steps.netlify.outcome == 'success' }}
-        uses: marocchino/sticky-pull-request-comment@v2
-        with:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          message: |
-            📖 **Documentation Preview**: ${{ steps.netlify.outputs.deploy-url }}
-
-            Deployed on Netlify from commit ${{ github.event.pull_request.head.sha }}
--- 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
@@ -13,59 +13,59 @@ 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

-  pytest:
-    name: PyTest
+  preload-cache:
+    name: Preload HF cache
    runs-on: ubuntu-latest
    strategy:
      fail-fast: false
-      max-parallel: 2
      matrix:
        python_version: ["3.11"]
-        pytorch_version: ["2.6.0", "2.7.0"]
+        pytorch_version: ["2.6.0"]
    timeout-minutes: 20

+    env:
+      AXOLOTL_IS_CI_CACHE_PRELOAD: "1"
+
    steps:
      - name: Check out repository code
        uses: actions/checkout@v4

-      - name: Restore Cache from S3
-        id: hf-cache-restore-s3
-        run: |
-          mkdir -p /home/runner/.cache/huggingface/hub
-          curl -L https://d1dttdx32dkk5p.cloudfront.net/hf-cache.tar.zst | tar -xf - -C /home/runner/.cache/huggingface/hub/  --use-compress-program unzstd
+      - name: Restore HF cache
+        id: hf-cache-restore
+        uses: actions/cache/restore@v4
+        with:
+          path: |
+            /home/runner/.cache/huggingface/hub/datasets--*
+            /home/runner/.cache/huggingface/hub/models--*
+          key: ${{ runner.os }}-hf-hub-cache-v2

      - name: Setup Python
        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==${{ matrix.pytorch_version }} torchvision
-
-      - name: Update pyproject.toml for nightly builds
-        run: |
-          sed -i 's#"transformers==.*"#"transformers @ git+https://github.com/huggingface/transformers.git@main"#' pyproject.toml
-          sed -i 's#"peft==.*"#"peft @ git+https://github.com/huggingface/peft.git@main"#' pyproject.toml
-          sed -i 's#"accelerate==.*"#"accelerate @ git+https://github.com/huggingface/accelerate.git@main"#' pyproject.toml
-          sed -i 's#"trl==.*"#"trl @ git+https://github.com/huggingface/trl.git@main"#' pyproject.toml
-          sed -i 's#"datasets==.*"#"datasets @ git+https://github.com/huggingface/datasets.git@main"#' pyproject.toml
+          pip3 install torch==${{ matrix.pytorch_version }}

      - 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: |
@@ -75,38 +75,138 @@ jobs:
        run: |
          axolotl --help

+      - name: Pre-Download dataset fixture
+        run: |
+          huggingface-cli download --repo-type=dataset axolotl-ai-internal/axolotl-oss-dataset-fixtures
+
      - name: Run tests
        run: |
-          pytest -v --durations=10 -n8 --dist loadfile --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ tests/
-          pytest -v --durations=10 tests/patched/
-          pytest -v --durations=10 tests/cli/
+          pytest -v tests/conftest.py

+      - name: Upload coverage to Codecov
+        uses: codecov/codecov-action@v5
+        with:
+          token: ${{ secrets.CODECOV_TOKEN }}
+          files: ./coverage.xml
+          flags: unittests,pytorch-${{ 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 {} \;
+
+      - name: Save HF cache
+        id: hf-cache
+        uses: actions/cache/save@v4
+        with:
+          path: |
+            /home/runner/.cache/huggingface/hub/datasets--*
+            /home/runner/.cache/huggingface/hub/models--*
+          key: ${{ steps.hf-cache-restore.outputs.cache-primary-key }}
+
+  pytest:
+    name: PyTest
+    runs-on: ubuntu-latest
+    needs: [preload-cache]
+    strategy:
+      fail-fast: false
+      max-parallel: 2
+      matrix:
+        python_version: ["3.11"]
+        pytorch_version: ["2.5.1", "2.6.0", "2.7.0"]
+    timeout-minutes: 20
+
+    steps:
+      - name: Check out repository code
+        uses: actions/checkout@v4
+
+      - name: Restore HF cache
+        id: hf-cache-restore
+        uses: actions/cache/restore@v4
+        with:
+          path: |
+            /home/runner/.cache/huggingface/hub/datasets--*
+            /home/runner/.cache/huggingface/hub/models--*
+          key: ${{ runner.os }}-hf-hub-cache-v2
+
+      - name: Setup Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python_version }}
+          cache: 'pip' # caching pip dependencies
+
+      - name: upgrade pip
+        run: |
+          pip3 install --upgrade pip
+          pip3 install --upgrade packaging==23.2 setuptools==75.8.0 wheel
+
+      - name: Install PyTorch
+        run: |
+          pip3 install torch==${{ matrix.pytorch_version }}
+
+      - name: Update requirements.txt
+        run: |
+          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: |
+          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: |
+          python -c "import torch; assert '${{ matrix.pytorch_version }}' in torch.__version__"
+
+      - name: Ensure axolotl CLI was installed
+        run: |
+          axolotl --help
+
+      - name: Pre-Download dataset fixture
+        run: |
+          huggingface-cli download --repo-type=dataset axolotl-ai-internal/axolotl-oss-dataset-fixtures
+
+      - name: Run tests
+        run: |
+          pytest -v -n8 --dist loadfile --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ tests/
+          pytest -v tests/patched/
+          pytest -v tests/cli/
+
+      - name: cleanup pip cache
+        run: |
+          find "$(pip cache dir)/http-v2" -type f -mtime +14 -exec rm {} \;

  docker-e2e-tests:
    if: github.repository_owner == 'axolotl-ai-cloud'
    # this job needs to be run on self-hosted GPU runners...
    runs-on: [self-hosted, modal]
-    timeout-minutes: 120
+    timeout-minutes: 60
    needs: [pre-commit, pytest]

    strategy:
      fail-fast: false
      matrix:
        include:
-          - cuda: 126
-            cuda_version: 12.6.3
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            num_gpus: 1
+            axolotl_extras:
+            nightly_build: "true"
+          - cuda: 124
+            cuda_version: 12.4.1
            python_version: "3.11"
            pytorch: 2.6.0
            num_gpus: 1
            axolotl_extras:
            nightly_build: "true"
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            num_gpus: 1
-            axolotl_extras:
-            nightly_build: "true"
    steps:
      - name: Checkout
        uses: actions/checkout@v4
@@ -114,16 +214,13 @@ 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==0.71.8 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}}" >> $GITHUB_ENV
@@ -133,49 +230,4 @@ jobs:
          echo "CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }}" >> $GITHUB_ENV
      - name: Run tests job on Modal
        run: |
-          modal run -m cicd.e2e_tests
-  docker-e2e-multigpu-tests:
-    if: github.repository_owner == 'axolotl-ai-cloud'
-    # this job needs to be run on self-hosted GPU runners...
-    runs-on: [self-hosted, modal]
-    timeout-minutes: 120
-    needs: [pre-commit, pytest, docker-e2e-tests]
-
-    strategy:
-      fail-fast: false
-      matrix:
-        include:
-          - cuda: 126
-            cuda_version: 12.6.3
-            python_version: "3.11"
-            pytorch: 2.7.1
-            num_gpus: 2
-            axolotl_extras:
-            nightly_build: "true"
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - name: Install Python
-        uses: actions/setup-python@v5
-        with:
-          python-version: "3.11"
-      - name: Install uv
-        uses: astral-sh/setup-uv@v4
-        with:
-          version: "latest"
-      - name: Install Modal
-        run: |
-          uv pip install --system modal==1.0.2 jinja2
-      - name: Update env vars
-        run: |
-          echo "BASE_TAG=main-base-uv-py${{ 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}}" >> $GITHUB_ENV
-          echo "CUDA=${{ matrix.cuda }}" >> $GITHUB_ENV
-          echo "N_GPUS=${{ matrix.num_gpus }}" >> $GITHUB_ENV
-          echo "NIGHTLY_BUILD=${{ matrix.nightly_build }}" >> $GITHUB_ENV
-          echo "CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }}" >> $GITHUB_ENV
-      - name: Run tests job on Modal
-        run: |
-          modal run cicd.multigpu
+          modal run cicd.e2e_tests
--- a/.github/workflows/tests.yml
+++ b/.github/workflows/tests.yml
@@ -7,16 +7,17 @@ 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:
@@ -33,12 +34,12 @@ jobs:
  pre-commit:
    name: pre-commit
    runs-on: ubuntu-latest
-    if: ${{ !github.event.pull_request.draft }}
    steps:
      - uses: actions/checkout@v4
      - 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
@@ -46,13 +47,12 @@ jobs:
  pytest:
    name: PyTest
    runs-on: ubuntu-latest
-    if: ${{ !github.event.pull_request.draft }}
 #    needs: [preload-cache]
    strategy:
      fail-fast: false
      matrix:
        python_version: ["3.11"]
-        pytorch_version: ["2.6.0", "2.7.1", "2.8.0"]
+        pytorch_version: ["2.5.1", "2.6.0", "2.7.1"]
    timeout-minutes: 20

    steps:
@@ -69,25 +69,24 @@ 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 }}

      - 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 +102,9 @@ 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 -n8 --dist loadfile --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ tests/ --cov=axolotl --cov-report=xml
+          pytest -v tests/patched/ --cov=axolotl --cov-append --cov-report=xml
+          pytest -v tests/cli/ --cov=axolotl --cov-append --cov-report=xml

      - name: Upload coverage to Codecov
        uses: codecov/codecov-action@v5
@@ -116,16 +114,18 @@ 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
    runs-on: ubuntu-latest
-    if: ${{ !github.event.pull_request.draft }}
    strategy:
      fail-fast: false
      matrix:
        python_version: ["3.11"]
-        pytorch_version: ["2.6.0", "2.7.1", "2.8.0"]
+        pytorch_version: ["2.5.1", "2.6.0", "2.7.1"]
    timeout-minutes: 20

    steps:
@@ -142,26 +142,25 @@ 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 }}

      - 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: |
@@ -176,48 +175,21 @@ 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 -n8 --dist loadfile --ignore=tests/e2e/ --ignore=tests/patched/ --ignore=tests/cli/ tests/
+          pytest -v tests/patched/
+          pytest -v tests/cli/

-  gate-skip-e2e:
-    needs: [pre-commit, pytest, pytest-sdist]
-    runs-on: ubuntu-latest
-    outputs:
-      skip: ${{ steps.compute.outputs.skip }}
-    steps:
-      - uses: actions/github-script@v7
-        id: compute
-        with:
-          script: |
-            const token = /\[skip-e2e\]/i;
-            let msg = '';
-            if (context.eventName === 'push') {
-              msg = context.payload.head_commit?.message || '';
-            } else if (context.eventName === 'pull_request') {
-              const { owner, repo } = context.repo;
-              const prNumber = context.payload.pull_request.number;
-              const commits = await github.paginate(
-                github.rest.pulls.listCommits,
-                { owner, repo, pull_number: prNumber, per_page: 100 }
-              );
-              msg = commits.at(-1)?.commit?.message || '';
-            }
-            const title = context.payload.pull_request?.title || '';
-            const body  = context.payload.pull_request?.body  || '';
-            const skip = token.test(msg) || token.test(title) || token.test(body);
-            core.setOutput('skip', String(skip));
+      - name: cleanup pip cache
+        run: |
+          find "$(pip cache dir)/http-v2" -type f -mtime +14 -exec rm {} \;

  docker-e2e-tests-1st:
    # Run this job first as a gate for running the remainder of the test matrix
-    if: >
-      github.repository_owner == 'axolotl-ai-cloud' &&
-      (github.event_name != 'pull_request' || !github.event.pull_request.draft) &&
-      needs.gate-skip-e2e.outputs.skip != 'true'
+    if: ${{ ! contains(github.event.commits[0].message, '[skip e2e]') && github.repository_owner == 'axolotl-ai-cloud' }}
    # this job needs to be run on self-hosted GPU runners...
    runs-on: [self-hosted, modal]
    timeout-minutes: 120
-    needs: [pre-commit, pytest, pytest-sdist, gate-skip-e2e]
+    needs: [pre-commit, pytest, pytest-sdist]

    strategy:
      fail-fast: false
@@ -226,16 +198,16 @@ jobs:
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
-            pytorch: 2.7.1
+            pytorch: 2.6.0
            num_gpus: 1
            axolotl_extras:
          - cuda: 126
            cuda_version: 12.6.3
            python_version: "3.11"
-            pytorch: 2.7.1
+            pytorch: 2.6.0
            num_gpus: 1
            axolotl_extras:
-            dockerfile: "Dockerfile.jinja"
+            dockerfile: "Dockerfile-uv.jinja"
    steps:
      - name: Checkout
        uses: actions/checkout@v4
@@ -243,17 +215,13 @@ 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
@@ -267,16 +235,13 @@ jobs:
          modal run cicd.e2e_tests

  docker-e2e-tests:
-    if: >
-      github.repository_owner == 'axolotl-ai-cloud' &&
-      (github.event_name != 'pull_request' || !github.event.pull_request.draft) &&
-      needs.gate-skip-e2e.outputs.skip != 'true'
+    if: github.repository_owner == 'axolotl-ai-cloud'
    # this job needs to be run on self-hosted GPU runners...
    runs-on: [self-hosted, modal]
    timeout-minutes: 120
    # Only run the remainder of the matrix if the first e2e check passed;
    # this is to save on wasted compute costs for known failures that get caught in the first run
-    needs: [pre-commit, pytest, gate-skip-e2e, docker-e2e-tests-1st]
+    needs: [pre-commit, pytest, docker-e2e-tests-1st]

    strategy:
      fail-fast: false
@@ -287,9 +252,15 @@ jobs:
            python_version: "3.11"
            pytorch: 2.6.0
            num_gpus: 1
+            axolotl_extras: llmcompressor
+          - cuda: 124
+            cuda_version: 12.4.1
+            python_version: "3.11"
+            pytorch: 2.5.1
+            num_gpus: 1
            axolotl_extras:
-          - cuda: 128
-            cuda_version: 12.8.1
+          - cuda: 126
+            cuda_version: 12.6.3
            python_version: "3.11"
            pytorch: 2.7.1
            num_gpus: 1
@@ -297,10 +268,9 @@ jobs:
          - cuda: 128
            cuda_version: 12.8.1
            python_version: "3.11"
-            pytorch: 2.8.0
+            pytorch: 2.7.1
            num_gpus: 1
-            gpu_type: "B200"
-            axolotl_extras: fbgemm-gpu
+            axolotl_extras:
    steps:
      - name: Checkout
        uses: actions/checkout@v4
@@ -308,24 +278,19 @@ 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}}" >> $GITHUB_ENV
          echo "CUDA=${{ matrix.cuda }}" >> $GITHUB_ENV
          echo "MODAL_IMAGE_BUILDER_VERSION=2024.10" >> $GITHUB_ENV
          echo "N_GPUS=${{ matrix.num_gpus }}" >> $GITHUB_ENV
-          echo "GPU_TYPE=${{ matrix.gpu_type || 'L40S'}}" >> $GITHUB_ENV
          echo "CODECOV_TOKEN=${{ secrets.CODECOV_TOKEN }}" >> $GITHUB_ENV
          echo "E2E_DOCKERFILE=${{ matrix.dockerfile || 'Dockerfile.jinja'}}" >> $GITHUB_ENV
      - name: Run tests job on Modal
@@ -336,16 +301,15 @@ jobs:
    runs-on: [self-hosted, modal]
    timeout-minutes: 90
    needs: [docker-e2e-tests]
-    if: ${{ !github.event.pull_request.draft }}

    strategy:
      fail-fast: false
      matrix:
        include:
-          - cuda: 126
-            cuda_version: 12.6.3
+          - cuda: 124
+            cuda_version: 12.4.1
            python_version: "3.11"
-            pytorch: 2.7.1
+            pytorch: 2.6.0
            num_gpus: 1
            axolotl_extras:
    steps:
@@ -355,17 +319,13 @@ 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
@@ -190,6 +190,3 @@ out/

 # vim
 *.swp
-
-# setuptools-scm generated version file
-src/axolotl/_version.py
--- a/.isort.cfg
+++ b/.isort.cfg
@@ -0,0 +1,4 @@
+[settings]
+profile=black
+known_third_party=wandb,comet_ml
+known_local_folder=src,tests
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -3,21 +3,31 @@ default_language_version:

 repos:
 -   repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v6.0.0
+    rev: v5.0.0
    hooks:
    -   id: check-yaml
    -   id: end-of-file-fixer
    -   id: trailing-whitespace
    -   id: no-commit-to-branch
        args: ['--branch', 'main']
-   repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.12.12
+-   repo: https://github.com/psf/black
+    rev: 25.1.0
    hooks:
-    -   id: ruff
-        args: [--fix]
-    -   id: ruff-format
+    -   id: black
+-   repo: https://github.com/pycqa/isort
+    rev: 6.0.1
+    hooks:
+      - id: isort
+-   repo: https://github.com/PyCQA/flake8
+    rev: 7.3.0
+    hooks:
+    - id: flake8
+-   repo: https://github.com/pylint-dev/pylint
+    rev: v3.3.7
+    hooks:
+    - id: pylint
 -   repo: https://github.com/pre-commit/mirrors-mypy
-    rev: v1.17.1
+    rev: v1.16.1
    hooks:
    - id: mypy
      additional_dependencies:
@@ -26,7 +36,7 @@ repos:
            'pydantic>=2.5.3',
        ]
 -   repo: https://github.com/PyCQA/bandit
-    rev: 1.8.6
+    rev: 1.8.5
    hooks:
    -   id: bandit
        args: [
--- a/.pylintrc
+++ b/.pylintrc
@@ -0,0 +1,15 @@
+[MASTER]
+init-hook="from pylint.config import find_default_config_files; import sys; sys.path.append(next(find_default_config_files()).parent.as_posix())"
+
+[TYPECHECK]
+
+# List of members which are set dynamically and missed by Pylint inference
+# system, and so shouldn't trigger E1101 when accessed.
+generated-members=numpy.*, torch.*
+
+
+[pylint.messages_control]
+disable=missing-function-docstring, line-too-long, import-error,
+    too-many-arguments, too-many-locals, too-many-statements, too-many-branches, too-few-public-methods,
+    too-many-instance-attributes, fixme, import-outside-toplevel, logging-fstring-interpolation,
+    too-many-positional-arguments, possibly-used-before-assignment
--- 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/.runpod/README.md
+++ b/.runpod/README.md
@@ -119,15 +119,14 @@ datasets:

 ## Dataset Processing

-| Option                            | Default                    | Description                         |
-| --------------------------------- | -------------------------- | ----------------------------------- |
-| `dataset_prepared_path`           | `"data/last_run_prepared"` | Path for prepared dataset           |
-| `push_dataset_to_hub`             | `""`                       | Push dataset to HF hub              |
-| `dataset_processes`               | `4`                        | Number of preprocessing processes   |
-| `dataset_keep_in_memory`          | `false`                    | Keep dataset in memory              |
-| `shuffle_merged_datasets`         | `true`                     | Shuffle merged datasets             |
-| `shuffle_before_merging_datasets` | `false`                    | Shuffle each dataset before merging |
-| `dataset_exact_deduplication`     | `true`                     | Deduplicate datasets                |
+| Option                        | Default                    | Description                       |
+| ----------------------------- | -------------------------- | --------------------------------- |
+| `dataset_prepared_path`       | `"data/last_run_prepared"` | Path for prepared dataset         |
+| `push_dataset_to_hub`         | `""`                       | Push dataset to HF hub            |
+| `dataset_processes`           | `4`                        | Number of preprocessing processes |
+| `dataset_keep_in_memory`      | `false`                    | Keep dataset in memory            |
+| `shuffle_merged_datasets`     | `true`                     | Shuffle merged datasets           |
+| `dataset_exact_deduplication` | `true`                     | Deduplicate datasets              |

 ## LoRA Configuration

@@ -185,6 +184,7 @@ datasets:
 | `flash_attention`          | `false` | Use flash attention           |
 | `flash_attn_cross_entropy` | `false` | Flash attention cross entropy |
 | `flash_attn_rms_norm`      | `false` | Flash attention RMS norm      |
+| `flash_attn_fuse_qkv`      | `false` | Fuse QKV operations           |
 | `flash_attn_fuse_mlp`      | `false` | Fuse MLP operations           |
 | `sdp_attention`            | `false` | Use scaled dot product        |
 | `s2_attention`             | `false` | Use shifted sparse attention  |
--- a/.runpod/src/config/config.yaml
+++ b/.runpod/src/config/config.yaml
@@ -97,7 +97,7 @@
 #       # 'no_input_format' cannot include {input}
 #       no_input_format: "{instruction} "

-#       # For `completion` datasets only, uses the provided field instead of `text` column
+#       # For `completion` datsets only, uses the provided field instead of `text` column
 #       field:

 # # Axolotl attempts to save the dataset as an arrow after packing the data together so
@@ -296,6 +296,7 @@
 # flash_attention:
 # flash_attn_cross_entropy:  # Whether to use flash-attention cross entropy implementation - advanced use only
 # flash_attn_rms_norm:  # Whether to use flash-attention rms norm implementation - advanced use only
+# flash_attn_fuse_qkv: # Whether to fuse QKV into a single operation
 # flash_attn_fuse_mlp: # Whether to fuse part of the MLP into a single operation
 # # Whether to use scaled-dot-product attention
 # # https://pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html
@@ -540,6 +541,7 @@ xformers_attention: ${XFORMERS_ATTENTION}
 flash_attention: ${FLASH_ATTENTION}
 flash_attn_cross_entropy: ${FLASH_ATTN_CROSS_ENTROPY}
 flash_attn_rms_norm: ${FLASH_ATTN_RMS_NORM}
+flash_attn_fuse_qkv: ${FLASH_ATTN_FUSE_QKV}
 flash_attn_fuse_mlp: ${FLASH_ATTN_FUSE_MLP}
 sdp_attention: ${SDP_ATTENTION}
 s2_attention: ${S2_ATTENTION}
--- a/CITATION.cff
+++ b/CITATION.cff
@@ -1,10 +0,0 @@
-cff-version: 1.2.0
-type: software
-title: "Axolotl: Open Source LLM Post-Training"
-message: "If you use this software, please cite it as below."
-authors:
-  - name: "Axolotl maintainers and contributors"
-repository-code: "https://github.com/axolotl-ai-cloud/axolotl"
-url: "https://axolotl.ai/"
-license: Apache-2.0
-date-released: "2023-05-30"
--- a/MANIFEST.in
+++ b/MANIFEST.in
@@ -1,5 +1,5 @@
-include pyproject.toml
+include requirements.txt
 include README.md
 include LICENSE
-include src/axolotl/utils/chat_templates/templates/*.jinja
-recursive-include src/axolotl *.py
+include src/setuptools_axolotl_dynamic_dependencies.py
+recursive-include axolotl *.py
--- a/README.md
+++ b/README.md
@@ -5,9 +5,6 @@
        <img alt="Axolotl" src="https://raw.githubusercontent.com/axolotl-ai-cloud/axolotl/887513285d98132142bf5db2a74eb5e0928787f1/image/axolotl_logo_digital_black.svg" width="400" height="104" style="max-width: 100%;">
    </picture>
 </p>
-  <p align="center">
-      <strong>A Free and Open Source LLM Fine-tuning Framework</strong><br>
-  </p>

 <p align="center">
    <img src="https://img.shields.io/github/license/axolotl-ai-cloud/axolotl.svg?color=blue" alt="GitHub License">
@@ -20,7 +17,6 @@
    <br/>
    <a href="https://discord.com/invite/HhrNrHJPRb"><img src="https://img.shields.io/badge/discord-7289da.svg?style=flat-square&logo=discord" alt="discord" style="height: 20px;"></a>
    <a href="https://twitter.com/axolotl_ai"><img src="https://img.shields.io/twitter/follow/axolotl_ai?style=social" alt="twitter" style="height: 20px;"></a>
-    <a href="https://colab.research.google.com/github/axolotl-ai-cloud/axolotl/blob/main/examples/colab-notebooks/colab-axolotl-example.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="google-colab" style="height: 20px;"></a>
    <br/>
    <img src="https://github.com/axolotl-ai-cloud/axolotl/actions/workflows/tests-nightly.yml/badge.svg" alt="tests-nightly">
    <img src="https://github.com/axolotl-ai-cloud/axolotl/actions/workflows/multi-gpu-e2e.yml/badge.svg" alt="multigpu-semi-weekly tests">
@@ -29,81 +25,49 @@

 ## 🎉 Latest Updates

- 2025/07:
-  - ND Parallelism support has been added into Axolotl. Compose Context Parallelism (CP), Tensor Parallelism (TP), and Fully Sharded Data Parallelism (FSDP) within a single node and across multiple nodes. Check out the [blog post](https://huggingface.co/blog/accelerate-nd-parallel) for more info.
-  - Axolotl adds more models: [GPT-OSS](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/gpt-oss), [Gemma 3n](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/gemma3n), [Liquid Foundation Model 2 (LFM2)](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/lfm2), and [Arcee Foundation Models (AFM)](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/afm).
-  - FP8 finetuning with fp8 gather op is now possible in Axolotl via `torchao`. Get started [here](https://docs.axolotl.ai/docs/mixed_precision.html#sec-fp8)!
-  - [Voxtral](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/voxtral), [Magistral 1.1](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/magistral), and [Devstral](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/devstral) with mistral-common tokenizer support has been integrated in Axolotl!
-  - TiledMLP support for single-GPU to multi-GPU training with DDP, DeepSpeed and FSDP support has been added to support Arctic Long Sequence Training. (ALST). See [examples](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/alst) for using ALST with Axolotl!
- 2025/05: Quantization Aware Training (QAT) support has been added to Axolotl. Explore the [docs](https://docs.axolotl.ai/docs/qat.html) to learn more!
- 2025/03: Axolotl has implemented Sequence Parallelism (SP) support. Read the [blog](https://huggingface.co/blog/axolotl-ai-co/long-context-with-sequence-parallelism-in-axolotl) and [docs](https://docs.axolotl.ai/docs/sequence_parallelism.html) to learn how to scale your context length when fine-tuning.
-
-<details>
-
-<summary>Expand older updates</summary>
-
 - 2025/06: Magistral with mistral-common tokenizer support has been added to Axolotl. See [examples](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/magistral) to start training your own Magistral models with Axolotl!
+- 2025/05: Quantization Aware Training (QAT) support has been added to Axolotl. Explore the [docs](https://docs.axolotl.ai/docs/qat.html) to learn more!
 - 2025/04: Llama 4 support has been added in Axolotl. See [examples](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/llama-4) to start training your own Llama 4 models with Axolotl's linearized version!
+- 2025/03: Axolotl has implemented Sequence Parallelism (SP) support. Read the [blog](https://huggingface.co/blog/axolotl-ai-co/long-context-with-sequence-parallelism-in-axolotl) and [docs](https://docs.axolotl.ai/docs/sequence_parallelism.html) to learn how to scale your context length when fine-tuning.
 - 2025/03: (Beta) Fine-tuning Multimodal models is now supported in Axolotl. Check out the [docs](https://docs.axolotl.ai/docs/multimodal.html) to fine-tune your own!
 - 2025/02: Axolotl has added LoRA optimizations to reduce memory usage and improve training speed for LoRA and QLoRA in single GPU and multi-GPU training (DDP and DeepSpeed). Jump into the [docs](https://docs.axolotl.ai/docs/lora_optims.html) to give it a try.
 - 2025/02: Axolotl has added GRPO support. Dive into our [blog](https://huggingface.co/blog/axolotl-ai-co/training-llms-w-interpreter-feedback-wasm) and [GRPO example](https://github.com/axolotl-ai-cloud/grpo_code) and have some fun!
 - 2025/01: Axolotl has added Reward Modelling / Process Reward Modelling fine-tuning support. See [docs](https://docs.axolotl.ai/docs/reward_modelling.html).

-</details>
-
 ## ✨ Overview

-Axolotl is a free and open-source tool designed to streamline post-training and fine-tuning for the latest large language models (LLMs).
+Axolotl is a tool designed to streamline post-training for various AI models.

 Features:

- **Multiple Model Support**: Train various models like GPT-OSS, LLaMA, Mistral, Mixtral, Pythia, and many more models available on the Hugging Face Hub.
- **Multimodal Training**: Fine-tune vision-language models (VLMs) including LLaMA-Vision, Qwen2-VL, Pixtral, LLaVA, SmolVLM2, and audio models like Voxtral with image, video, and audio support.
- **Training Methods**: Full fine-tuning, LoRA, QLoRA, GPTQ, QAT, Preference Tuning (DPO, IPO, KTO, ORPO), RL (GRPO), and Reward Modelling (RM) / Process Reward Modelling (PRM).
- **Easy Configuration**: Re-use a single YAML configuration file across the full fine-tuning pipeline: dataset preprocessing, training, evaluation, quantization, and inference.
+- **Multiple Model Support**: Train various models like LLaMA, Mistral, Mixtral, Pythia, and more. We are compatible with HuggingFace transformers causal language models.
+- **Training Methods**: Full fine-tuning, LoRA, QLoRA, GPTQ, QAT, Preference Tuning (DPO, IPO, KTO, ORPO), RL (GRPO), Multimodal, and Reward Modelling (RM) / Process Reward Modelling (PRM).
+- **Easy Configuration**: Re-use a single YAML file between dataset preprocess, training, evaluation, quantization, and inference.
 - **Performance Optimizations**: [Multipacking](https://docs.axolotl.ai/docs/multipack.html), [Flash Attention](https://github.com/Dao-AILab/flash-attention), [Xformers](https://github.com/facebookresearch/xformers), [Flex Attention](https://pytorch.org/blog/flexattention/), [Liger Kernel](https://github.com/linkedin/Liger-Kernel), [Cut Cross Entropy](https://github.com/apple/ml-cross-entropy/tree/main), [Sequence Parallelism (SP)](https://docs.axolotl.ai/docs/sequence_parallelism.html), [LoRA optimizations](https://docs.axolotl.ai/docs/lora_optims.html), [Multi-GPU training (FSDP1, FSDP2, DeepSpeed)](https://docs.axolotl.ai/docs/multi-gpu.html), [Multi-node training (Torchrun, Ray)](https://docs.axolotl.ai/docs/multi-node.html), and many more!
 - **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+

-### Google Colab
+## 🚀 Quick Start

-[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/axolotl-ai-cloud/axolotl/blob/main/examples/colab-notebooks/colab-axolotl-example.ipynb#scrollTo=msOCO4NRmRLa)
+**Requirements**:
+
+- NVIDIA GPU (Ampere or newer for `bf16` and Flash Attention) or AMD GPU
+- Python 3.11
+- PyTorch ≥2.5.1

 ### 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
@@ -115,20 +79,6 @@ docker run --gpus '"all"' --rm -it axolotlai/axolotl:main-latest

 Other installation approaches are described [here](https://docs.axolotl.ai/docs/installation.html).

-#### Cloud Providers
-
-<details>
-
- [RunPod](https://runpod.io/gsc?template=v2ickqhz9s&ref=6i7fkpdz)
- [Vast.ai](https://cloud.vast.ai?ref_id=62897&template_id=bdd4a49fa8bce926defc99471864cace&utm_source=github&utm_medium=developer_community&utm_campaign=template_launch_axolotl&utm_content=readme)
- [PRIME Intellect](https://app.primeintellect.ai/dashboard/create-cluster?image=axolotl&location=Cheapest&security=Cheapest&show_spot=true)
- [Modal](https://www.modal.com?utm_source=github&utm_medium=github&utm_campaign=axolotl)
- [Novita](https://novita.ai/gpus-console?templateId=311)
- [JarvisLabs.ai](https://jarvislabs.ai/templates/axolotl)
- [Latitude.sh](https://latitude.sh/blueprint/989e0e79-3bf6-41ea-a46b-1f246e309d5c)
-
-</details>
-
 ### Your First Fine-tune

 ```bash
@@ -170,22 +120,14 @@ Contributions are welcome! Please see our [Contributing Guide](https://github.co

 ## ❤️ Sponsors

+Thank you to our sponsors who help make Axolotl possible:
+
+- [Modal](https://www.modal.com?utm_source=github&utm_medium=github&utm_campaign=axolotl) - Modal lets you run
+jobs in the cloud, by just writing a few lines of Python. Customers use Modal to deploy Gen AI models at large scale,
+fine-tune large language models, run protein folding simulations, and much more.
+
 Interested in sponsoring? Contact us at [wing@axolotl.ai](mailto:wing@axolotl.ai)

-## 📝 Citing Axolotl
-
-If you use Axolotl in your research or projects, please cite it as follows:
-
-```bibtex
-@software{axolotl,
-  title = {Axolotl: Open Source LLM Post-Training},
-  author = {{Axolotl maintainers and contributors}},
-  url = {https://github.com/axolotl-ai-cloud/axolotl},
-  license = {Apache-2.0},
-  year = {2023}
-}
-```
-
 ## 📜 License

 This project is licensed under the Apache 2.0 License - see the [LICENSE](LICENSE) file for details.
--- a/TODO.md
+++ b/TODO.md
@@ -0,0 +1,10 @@
+# todo list
+
+- [] Validation of parameters for combinations that won't work
+
+
+
+## things that are known not to work
+
+- FSDP offload and gradient_checkpointing - https://github.com/pytorch/pytorch/issues/82203
+- adamw_bnb_8bit doesn't play well with FSDP offload
--- a/_quarto.yml
+++ b/_quarto.yml
@@ -35,30 +35,25 @@ quartodoc:
        - cli.train
        - cli.evaluate
        - cli.args
-        - cli.art
        - cli.checks
        - cli.config
-        - cli.delinearize_llama4
        - cli.inference
        - cli.merge_lora
        - cli.merge_sharded_fsdp_weights
        - cli.preprocess
-        - cli.quantize
+        - cli.sweeps
+        - cli.utils
        - cli.vllm_serve
        - cli.cloud.base
        - cli.cloud.modal_
-        - cli.utils
-        - cli.utils.args
-        - cli.utils.fetch
-        - cli.utils.load
-        - cli.utils.sweeps
-        - cli.utils.train
+        - cli.quantize
    - title: Trainers
      desc: Training implementations
      contents:
        - core.trainers.base
        - core.trainers.trl
        - core.trainers.mamba
+        - core.trainers.relora
        - core.trainers.dpo.trainer
        - core.trainers.grpo.trainer
        - core.trainers.grpo.sampler
@@ -153,7 +148,7 @@ quartodoc:
        - utils.distributed
        - utils.dict
        - utils.optimizers.adopt
-        - utils.data.streaming
+        - utils.data.pretraining
        - utils.data.sft
        - utils.quantization
    - title: Schemas
@@ -267,16 +262,12 @@ website:
            - docs/dataset_loading.qmd
            - docs/qat.qmd
            - docs/quantize.qmd
-            - docs/optimizations.qmd

        - section: "Core Concepts"
          contents:
            - docs/batch_vs_grad.qmd
            - docs/dataset_preprocessing.qmd
-            - docs/streaming.qmd
            - docs/multipack.qmd
-            - docs/mixed_precision.qmd
-            - docs/optimizers.qmd

        - section: "Advanced Features"
          contents:
@@ -285,8 +276,6 @@ website:
            - docs/torchao.qmd
            - docs/custom_integrations.qmd
            - docs/sequence_parallelism.qmd
-            - docs/gradient_checkpointing.qmd
-            - docs/nd_parallelism.qmd

        - section: "Troubleshooting"
          contents:
--- 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
+
+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 }}"
@@ -13,10 +9,9 @@ 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

 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
+    apt-get install -y --allow-change-held-packages vim curl nano libnccl2 libnccl-dev

 WORKDIR /workspace

@@ -29,27 +24,25 @@ 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
@@ -2,6 +2,8 @@
 modal application to run axolotl gpu tests in Modal
 """

+# pylint: disable=duplicate-code
+
 import os
 import pathlib
 import tempfile
@@ -22,9 +24,9 @@ df_template = template_env.get_template("Dockerfile.jinja")
 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"),
-    "CUDA": os.environ.get("CUDA", "126"),
+    "PYTORCH_VERSION": os.environ.get("PYTORCH_VERSION", "2.5.1"),
+    "BASE_TAG": os.environ.get("BASE_TAG", "main-base-py3.11-cu124-2.5.1"),
+    "CUDA": os.environ.get("CUDA", "124"),
    "GITHUB_REF": os.environ.get("GITHUB_REF", "refs/heads/main"),
    "GITHUB_SHA": os.environ.get("GITHUB_SHA", ""),
    "CODECOV_TOKEN": os.environ.get("CODECOV_TOKEN", ""),
@@ -61,7 +63,7 @@ def run_cmd(cmd: str, run_folder: str):

    # Propagate errors from subprocess.
    if exit_code := subprocess.call(cmd.split(), cwd=run_folder):  # nosec
-        exit(exit_code)
+        exit(exit_code)  # pylint: disable=consider-using-sys-exit


@app.function(
--- a/cicd/multigpu.sh
+++ b/cicd/multigpu.sh
@@ -2,7 +2,7 @@
 set -e

 # Only run two tests at a time to avoid OOM on GPU (with coverage collection)
-pytest -v --durations=10 -n2 \
+pytest -v -n2 \
  --ignore=/workspace/axolotl/tests/e2e/multigpu/solo/ \
  --ignore=/workspace/axolotl/tests/e2e/multigpu/patched/ \
  /workspace/axolotl/tests/e2e/multigpu/ \
@@ -19,7 +19,5 @@ pytest -v  --durations=10 -n1 /workspace/axolotl/tests/e2e/multigpu/patched/ \
  --cov-append \
  --cov-report=xml:multigpu-coverage.xml

-# Upload coverage to Codecov if CODECOV_TOKEN is available
-if [ -n "$CODECOV_TOKEN" ]; then
-  codecov upload-process -t "${CODECOV_TOKEN}" -f multigpu-coverage.xml -F multigpu,docker-tests,pytorch-${PYTORCH_VERSION} || true
-fi
+# Upload coverage to Codecov
+codecov upload-process -t "${CODECOV_TOKEN}" -f multigpu-coverage.xml -F multigpu,docker-tests,pytorch-${PYTORCH_VERSION} || true
--- a/cicd/single_gpu.py
+++ b/cicd/single_gpu.py
@@ -1,5 +1,7 @@
 """Modal app to run axolotl GPU tests"""

+# pylint: disable=duplicate-code
+
 import os
 import pathlib
 import tempfile
@@ -22,9 +24,9 @@ df_template = template_env.get_template(dockerfile)
 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"),
-    "CUDA": os.environ.get("CUDA", "126"),
+    "PYTORCH_VERSION": os.environ.get("PYTORCH_VERSION", "2.5.1"),
+    "BASE_TAG": os.environ.get("BASE_TAG", "main-base-py3.11-cu124-2.5.1"),
+    "CUDA": os.environ.get("CUDA", "124"),
    "GITHUB_REF": os.environ.get("GITHUB_REF", "refs/heads/main"),
    "GITHUB_SHA": os.environ.get("GITHUB_SHA", ""),
    "NIGHTLY_BUILD": os.environ.get("NIGHTLY_BUILD", ""),
@@ -57,16 +59,12 @@ VOLUME_CONFIG = {
 }

 N_GPUS = int(os.environ.get("N_GPUS", 1))
-GPU_TYPE = os.environ.get("GPU_TYPE", "L40S")
-GPU_CONFIG = f"{GPU_TYPE}:{N_GPUS}"
+GPU_CONFIG = f"L40S:{N_GPUS}"


 def run_cmd(cmd: str, run_folder: str):
    import subprocess  # nosec

-    sp_env = os.environ.copy()
-    sp_env["AXOLOTL_DATASET_PROCESSES"] = "8"
-
    # Propagate errors from subprocess.
-    if exit_code := subprocess.call(cmd.split(), cwd=run_folder, env=sp_env):  # nosec
-        exit(exit_code)
+    if exit_code := subprocess.call(cmd.split(), cwd=run_folder):  # nosec
+        exit(exit_code)  # pylint: disable=consider-using-sys-exit
--- a/codecov.yml
+++ b/codecov.yml
@@ -12,7 +12,7 @@ coverage:
      default:
        # basic
        target: auto
-        threshold: 1%
+        threshold: 0%
        base: auto
        # advanced
        branches: null
@@ -22,12 +22,11 @@ coverage:
        only_pulls: true
        flags: null
        paths: null
-        informational: true
    patch:
      default:
        # basic
        target: auto
-        threshold: 1%
+        threshold: 0%
        base: auto
        # advanced
        branches: null
--- a/deepspeed_configs/zero3.json
+++ b/deepspeed_configs/zero3.json
@@ -7,9 +7,9 @@
    "reduce_bucket_size": "auto",
    "stage3_prefetch_bucket_size": "auto",
    "stage3_param_persistence_threshold": "auto",
-    "max_live_parameters": 0,
-    "max_reuse_distance": 0,
-    "gather_16bit_weights_on_model_save": true
+    "stage3_max_live_parameters": 0,
+    "stage3_max_reuse_distance": 0,
+    "stage3_gather_16bit_weights_on_model_save": true
  },
  "bf16": {
    "enabled": "auto"
--- a/deepspeed_configs/zero3_bf16.json
+++ b/deepspeed_configs/zero3_bf16.json
@@ -7,9 +7,9 @@
    "reduce_bucket_size": "auto",
    "stage3_prefetch_bucket_size": "auto",
    "stage3_param_persistence_threshold": "auto",
-    "max_live_parameters": 0,
-    "max_reuse_distance": 0,
-    "gather_16bit_weights_on_model_save": true
+    "stage3_max_live_parameters": 0,
+    "stage3_max_reuse_distance": 0,
+    "stage3_gather_16bit_weights_on_model_save": true
  },
  "bf16": {
    "enabled": true
--- a/deepspeed_configs/zero3_bf16_cpuoffload_all.json
+++ b/deepspeed_configs/zero3_bf16_cpuoffload_all.json
@@ -17,9 +17,9 @@
    "reduce_bucket_size": "auto",
    "stage3_prefetch_bucket_size": "auto",
    "stage3_param_persistence_threshold": "auto",
-    "max_live_parameters": 0,
-    "max_reuse_distance": 0,
-    "gather_16bit_weights_on_model_save": true
+    "stage3_max_live_parameters": 0,
+    "stage3_max_reuse_distance": 0,
+    "stage3_gather_16bit_weights_on_model_save": true
  },
  "bf16": {
    "enabled": true
--- a/deepspeed_configs/zero3_bf16_cpuoffload_params.json
+++ b/deepspeed_configs/zero3_bf16_cpuoffload_params.json
@@ -13,9 +13,9 @@
    "reduce_bucket_size": "auto",
    "stage3_prefetch_bucket_size": "auto",
    "stage3_param_persistence_threshold": "auto",
-    "max_live_parameters": 0,
-    "max_reuse_distance": 0,
-    "gather_16bit_weights_on_model_save": true
+    "stage3_max_live_parameters": 0,
+    "stage3_max_reuse_distance": 0,
+    "stage3_gather_16bit_weights_on_model_save": true
  },
  "bf16": {
    "enabled": true
--- a/docker/Dockerfile
+++ b/docker/Dockerfile
@@ -1,24 +1,16 @@
-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 && \
-    rm -rf /var/cache/apt/archives && \
-    rm -rf /var/lib/apt/lists/*
+    apt-get install -y --allow-change-held-packages vim curl nano libnccl2 libnccl-dev rsync s3fs

 WORKDIR /workspace

@@ -26,25 +18,22 @@ 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] $AXOLOTL_ARGS; \
-    fi && \
-    uv pip install --python "$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
+        pip install --no-build-isolation -e .[deepspeed,flash-attn,ring-flash-attn,optimizers,ray] $AXOLOTL_ARGS; \
+    fi

-# fix so that git fetch/pull from remote works with shallow clone
+RUN python scripts/unsloth_install.py | sh
+RUN python scripts/cutcrossentropy_install.py | sh
+
+# So we can test the Docker image
+RUN pip install pytest
+
+# 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 && \
-    git config --global credential.helper store
+    git config --get remote.origin.fetch

-COPY .axolotl-complete.bash /root/.axolotl-complete.bash
-RUN chmod +x /root/.axolotl-complete.bash && \
-    echo 'source /root/.axolotl-complete.bash' >> ~/.bashrc
+# helper for huggingface-login cli
+RUN git config --global credential.helper store
--- a/docker/Dockerfile-base
+++ b/docker/Dockerfile-base
@@ -16,19 +16,12 @@ ENV PYTHON_VERSION=$PYTHON_VERSION
 ENV TORCH_CUDA_ARCH_LIST=$TORCH_CUDA_ARCH_LIST

 RUN apt-get update \
-    && apt-get install -y --no-install-recommends \
-        wget git build-essential ninja-build git-lfs libaio-dev pkg-config \
-        ibverbs-providers ibverbs-utils infiniband-diags  \
-        librdmacm-dev librdmacm1 rdmacm-utils slurm-wlm \
-    && rm -rf /var/cache/apt/archives \
-    && rm -rf /var/lib/apt/lists/* \
+    && apt-get install -y wget git build-essential ninja-build git-lfs libaio-dev pkg-config && rm -rf /var/lib/apt/lists/* \
    && wget \
    https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh \
    && mkdir /root/.conda \
    && bash Miniconda3-latest-Linux-x86_64.sh -b \
    && rm -f Miniconda3-latest-Linux-x86_64.sh \
-    && conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/main \
-    && conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/r \
    && conda create -n "py${PYTHON_VERSION}" python="${PYTHON_VERSION}"

 ENV PATH="/root/miniconda3/envs/py${PYTHON_VERSION}/bin:${PATH}"
@@ -37,16 +30,10 @@ WORKDIR /workspace

 RUN python3 -m pip install --upgrade pip && pip3 install -U packaging==23.2 setuptools==75.8.0 wheel && \
    python3 -m pip install --no-cache-dir -U torch==${PYTORCH_VERSION}+cu${CUDA} torchvision --extra-index-url https://download.pytorch.org/whl/cu$CUDA && \
-    CAUSAL_CONV1D_FORCE_CXX11_ABI=TRUE CAUSAL_CONV1D_FORCE_BUILD=TRUE python3 -m pip install --no-cache-dir causal_conv1d==1.5.2 && \
-    python3 -m pip install --no-cache-dir "mamba_ssm @ git+https://github.com/state-spaces/mamba.git@main" && \
-    python3 -m pip cache purge
+    python3 -m pip install --no-cache-dir "causal_conv1d @ git+https://github.com/Dao-AILab/causal-conv1d.git@main" && \
+    python3 -m pip install --no-cache-dir "mamba_ssm @ git+https://github.com/state-spaces/mamba.git@main"

 RUN git lfs install --skip-repo && \
    pip3 install awscli && \
    # The base image ships with `pydantic==1.8.2` which is not working
-    pip3 install -U --no-cache-dir pydantic==1.10.10 && \
-    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; \
-    fi
+    pip3 install -U --no-cache-dir pydantic==1.10.10
--- a/docker/Dockerfile-base-nightly
+++ b/docker/Dockerfile-base-nightly
@@ -22,22 +22,18 @@ RUN apt-get update \
    && mkdir /root/.conda \
    && bash Miniconda3-latest-Linux-x86_64.sh -b \
    && rm -f Miniconda3-latest-Linux-x86_64.sh \
-    && conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/main \
-    && conda tos accept --override-channels --channel https://repo.anaconda.com/pkgs/r \
    && conda create -n "py${PYTHON_VERSION}" python="${PYTHON_VERSION}"

 ENV PATH="/root/miniconda3/envs/py${PYTHON_VERSION}/bin:${PATH}"

 WORKDIR /workspace

-RUN python3 -m pip install --upgrade pip && pip3 install -U packaging==23.2 setuptools==75.8.0 wheel && \
+RUN python3 -m pip install --upgrade pip && pip3 install packaging && \
    python3 -m pip install --no-cache-dir -U torch --extra-index-url https://download.pytorch.org/whl/nightly/cu$CUDA && \
    python3 -m pip install --no-cache-dir "causal_conv1d @ git+https://github.com/Dao-AILab/causal-conv1d.git@main" && \
-    python3 -m pip install --no-cache-dir "mamba_ssm @ git+https://github.com/state-spaces/mamba.git@main" && \
-    python3 -m pip cache purge
+    python3 -m pip install --no-cache-dir "mamba_ssm @ git+https://github.com/state-spaces/mamba.git@main"

 RUN git lfs install --skip-repo && \
    pip3 install awscli && \
    # The base image ships with `pydantic==1.8.2` which is not working
-    pip3 install -U --no-cache-dir pydantic==1.10.10 && \
-    pip3 cache purge
+    pip3 install -U --no-cache-dir pydantic==1.10.10
--- a/docker/Dockerfile-cloud
+++ b/docker/Dockerfile-cloud
@@ -12,12 +12,9 @@ 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 apt update && \
-    apt install --yes --no-install-recommends openssh-server tmux iproute2 nvtop && \
-    rm -rf /var/cache/apt/archives && \
-    rm -rf /var/lib/apt/lists/* && \
+RUN pip install jupyterlab notebook ipywidgets && \
+    jupyter lab clean
+RUN apt install --yes --no-install-recommends openssh-server tmux iproute2 nvtop && \
    mkdir -p ~/.ssh && \
    chmod 700 ~/.ssh && \
    printf "\n[[ -z \"\$TMUX\"  ]] && { tmux attach-session -t ssh_tmux || tmux new-session -s ssh_tmux; exit; }\n" >> ~/.bashrc && \
--- a/docker/Dockerfile-cloud-no-tmux
+++ b/docker/Dockerfile-cloud-no-tmux
@@ -9,15 +9,13 @@ ENV HF_HUB_ENABLE_HF_TRANSFER="1"
 EXPOSE 8888
 EXPOSE 22

-COPY scripts/cloud-entrypoint.sh /root/cloud-entrypoint.sh
+COPY scripts/cloud-entrypoint-term.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 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 && \
-    rm -rf /var/lib/apt/lists/* && \
+RUN pip install jupyterlab notebook ipywidgets && \
+    jupyter lab clean
+RUN apt install --yes --no-install-recommends openssh-server tmux sudo && \
+    pip3 install -U --no-cache-dir grpcio ray[default]==2.9.3 && \
    mkdir -p ~/.ssh && \
    chmod 700 ~/.ssh && \
    printf "[ ! -z \"\$TERM\" -a -r /etc/motd ] && cat /etc/motd\n" >> ~/.bashrc && \
--- 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/cli.qmd
+++ b/docs/cli.qmd
@@ -23,20 +23,6 @@ axolotl <command> [config.yml] [options]

 The config file can be local or a URL to a raw YAML file.

-### Launcher Arguments
-
-For commands that support multi-GPU (`train`, `evaluate`, ...), you can pass launcher-specific arguments using the `--` separator:
-
-```bash
-# Pass torchrun arguments
-axolotl train config.yml --launcher torchrun -- --nproc_per_node=2 --nnodes=1
-
-# Pass accelerate arguments
-axolotl train config.yml --launcher accelerate -- --config_file=accelerate_config.yml --num_processes=4
-```
-
-Arguments after `--` are passed directly to the launcher (torchrun, accelerate launch, etc.).
-
 ## Command Reference

 ### fetch
@@ -94,11 +80,7 @@ axolotl train config.yml \
    --num-epochs 3

 # Training without accelerate
-axolotl train config.yml --launcher python
-
-# Pass launcher-specific arguments using -- separator
-axolotl train config.yml --launcher torchrun -- --nproc_per_node=2 --nnodes=1
-axolotl train config.yml --launcher accelerate -- --config_file=accelerate_config.yml
+axolotl train config.yml --no-accelerate

 # Resume training from checkpoint
 axolotl train config.yml --resume-from-checkpoint path/to/checkpoint
@@ -193,9 +175,6 @@ Evaluates a model's performance (loss etc) on the train and eval datasets.
 ```bash
 # Basic evaluation
 axolotl evaluate config.yml
-
-# Evaluation with launcher arguments
-axolotl evaluate config.yml --launcher torchrun -- --nproc_per_node=2
 ```

 ### lm-eval
@@ -308,6 +287,9 @@ axolotl preprocess config.yml --cloud cloud_config.yml
 # Train on cloud
 axolotl train config.yml --cloud cloud_config.yml

+# Train without accelerate on cloud
+axolotl train config.yml --cloud cloud_config.yml --no-accelerate
+
 # Run lm-eval on cloud
 axolotl lm-eval config.yml --cloud cloud_config.yml
 ```
--- a/docs/custom_integrations.qmd
+++ b/docs/custom_integrations.qmd
@@ -7,7 +7,6 @@ toc-depth: 3
 ```{python}
 #| echo: false

-import os
 import re

 def process_readme(integration_name):
@@ -54,24 +53,6 @@ sections = [
    ("LLMCompressor", "llm_compressor")
 ]

-for folder_name in os.listdir("../src/axolotl/integrations/"):
-    if folder_name in [path for name, path in sections]:
-        # skip if already in sections
-        continue
-    if os.path.exists(f"../src/axolotl/integrations/{folder_name}/README.md"):
-        # grab the first heading in README.md as the section name
-        with open(f"../src/axolotl/integrations/{folder_name}/README.md", "r") as f:
-            txt = f.read()
-            matches = re.search(r'^# (.*)\n?', txt, flags=re.MULTILINE)
-            if matches:
-                name = matches.group(1)
-            else:
-                continue
-            sections.append((name, folder_name))
-
-# sort sections by name
-sections = sorted(sections, key=lambda x: x[0])
-
 for section_name, folder_name in sections:
    print(print_section(section_name, folder_name))
 ```
--- a/docs/dataset-formats/conversation.qmd
+++ b/docs/dataset-formats/conversation.qmd
@@ -187,7 +187,6 @@ Instead of passing `tools` via the system prompt, an alternative method would be
            "role": "assistant", // call the function via assistant
            "tool_calls": [
                {
-                    "id": "...",  // required only for mistral
                    "type": "function",
                    "function": {
                        "name": "...",
@@ -200,7 +199,6 @@ Instead of passing `tools` via the system prompt, an alternative method would be
        },
        {
            "role": "tool",
-            "tool_call_id": "...",  // required only for mistral
            "name": "...",
            "content": "..."
        },
@@ -212,19 +210,10 @@ Instead of passing `tools` via the system prompt, an alternative method would be
 Tools need to follow [JSON schema](https://json-schema.org/learn/getting-started-step-by-step).
 :::

-::: {.callout-warning}
-If you have tool arguments with same name but different dtypes (like `"time": string` and `"time": number`), please save `arguments: ` as JSON string to prevent `datasets` from having casting issues.
-
-```
-"arguments": "{\"...\": \"...\"}"
-```
-:::
-
-Example config for Llama4:
 ```yaml
 chat_template: llama4
 datasets:
-  - path: Nanobit/text-tools-2k-test
+  - path: ...
    type: chat_template
    # field_tools: tools # default is `tools`
 ```
--- a/docs/dataset-formats/index.qmd
+++ b/docs/dataset-formats/index.qmd
@@ -61,7 +61,7 @@ While we recommend `.jsonl`, you can also use the other formats (`csv`, `parquet

 ### Pre-training without streaming

-In the case that the dataset is small and can be loaded entirely into memory, another approach to running pre-training is to use the `completion` format. This would mean that the entire dataset is pre-tokenized instead of on-demand in streaming.
+On the rare case that the dataset is small and can be loaded entirely into memory, another approach to running pre-training is to use the `completion` format. This would mean that the entire dataset is pre-tokenized instead of on-demand in streaming.

 One benefit of this is that the tokenization can be performed separately on a CPU-only machine, and then transferred to a GPU machine for training to save costs.

--- 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/docker.qmd
+++ b/docs/docker.qmd
@@ -34,9 +34,9 @@ Tags examples:

 - `main-base-py3.11-cu128-2.7.1`
 - `main-base-py3.11-cu126-2.7.1`
- `main-base-py3.11-cu126-2.7.0`
 - `main-base-py3.11-cu126-2.6.0`
 - `main-base-py3.11-cu124-2.6.0`
+- `main-base-py3.11-cu124-2.5.1`

 ## Main

@@ -76,12 +76,12 @@ Tags examples:

 - `main-py3.11-cu128-2.7.1`
 - `main-py3.11-cu126-2.7.1`
- `main-py3.11-cu126-2.7.0`
 - `main-py3.11-cu126-2.6.0`
 - `main-py3.11-cu124-2.6.0`
+- `main-py3.11-cu124-2.5.1`
 - `main-latest`
 - `main-20250303-py3.11-cu124-2.6.0`
- `main-20250303-py3.11-cu126-2.6.0`
+- `main-20250303-py3.11-cu124-2.5.1`
 - `0.10.1`

 ## Cloud
--- a/docs/faq.qmd
+++ b/docs/faq.qmd
@@ -51,18 +51,6 @@ description: Frequently asked questions
 >   pad_token: "..."
 > ```

-**Q: `IterableDataset error` or `KeyError: 'input_ids'` when using `preprocess` CLI**
-
-> A: This is because you may be using `preprocess` CLI with `pretraining_dataset:` or `skip_prepare_dataset: true` respectively. Please use `axolotl train` CLI directly instead as these datasets are prepared on demand.
-
-**Q: vLLM is not working with Axolotl**
-
-> A: We currently recommend torch 2.6.0 for use with `vllm`. Please ensure you use the right version. For Docker, please use the `main-py3.11-cu124-2.6.0` tag.
-
-**Q: FA2 2.8.0 `undefined symbol` runtime error on CUDA 12.4**
-
-> A: There seems to be a wheel issue with FA2 2.8.0 on CUDA 12.4. Try CUDA 12.6 instead or downgrade to FA2 2.7.4. Please refer to the upstream issue: https://github.com/Dao-AILab/flash-attention/issues/1717.
-
 ### Chat templates

 **Q: `jinja2.exceptions.UndefinedError: 'dict object' has no attribute 'content' / 'role' / ____`**
@@ -136,11 +124,3 @@ description: Frequently asked questions
 >   dynamic: false
 >   mode: max-autotune-no-cudagraphs
 > ```
-
-**Q: `ValueError("Backward pass should have cleared tracker of all tensors")`
-
-> A: This may happen due to edge cases in using the modern OffloadActivations context manager for CUDA streams. If you encounter this error, you may have success using the naive implementation with `offload_activations: legacy` in your YAML.
-
-**Q: `Error parsing tool_calls arguments as JSON.`
-
-> A: There is an error parsing string arguments to a dict. Please check your dataset and the error message for more details.
--- 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:
@@ -20,15 +20,9 @@ To enable `QLoRA` with `FSDP`, you need to perform the following steps:
 > See the [example config](#example-config) file in addition to reading these instructions.

 1. Set `adapter: qlora` in your axolotl config file.
-2. Enable FSDP in your axolotl config, as [described here](multi-gpu.qmd#sec-fsdp).
+2. Enable FSDP in your axolotl config, as [described here](https://github.com/axolotl-ai-cloud/axolotl?tab=readme-ov-file#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/gradient_checkpointing.qmd
+++ b/docs/gradient_checkpointing.qmd
@@ -1,29 +0,0 @@
---
-title: Gradient Checkpointing and Activation Offloading
---
-
-Gradient checkpointing and activation offloading are techniques used to optimize the performance of deep learning
-models by reducing the memory footprint and improving computational efficiency.
-
-### Enabling Gradient Checkpointing
-
-```yaml
-gradient_checkpointing: true
-```
-
-### Enabling Activation Offloading
-
-```yaml
-gradient_checkpointing: true  # required for activation offloading
-activation_offloading: true
-```
-
-Activation offloading variants:
-
-The default `activation_offloading: true` offloads activations to CPU and uses CUDA streams
-to overlap the communications and computations when offloading.
-
-The `activation_offloading: legacy` naively offloads activations to CPU and without additional optimizations.
-
-For resource constrained environments with limited CPU memory, `activation_offloading: disk` offloads
-activations to disk instead of CPU RAM so that much larger context lengths can be trained with minimal memory.
--- a/docs/installation.qmd
+++ b/docs/installation.qmd
@@ -15,7 +15,7 @@ This guide covers all the ways you can install and set up Axolotl for your envir

 - NVIDIA GPU (Ampere architecture or newer for `bf16` and Flash Attention) or AMD GPU
 - Python ≥3.11
- PyTorch ≥2.6.0
+- PyTorch ≥2.5.1

 ## Installation Methods {#sec-installation-methods}

@@ -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}
@@ -155,24 +124,21 @@ For providers supporting Docker:

 - Use `axolotlai/axolotl-cloud:main-latest`
 - Available on:
-    - [RunPod](https://runpod.io/gsc?template=v2ickqhz9s&ref=6i7fkpdz)
-    - [Vast.ai](https://cloud.vast.ai?ref_id=62897&template_id=bdd4a49fa8bce926defc99471864cace&utm_source=axolotl&utm_medium=partner&utm_campaign=template_launch_july2025&utm_content=docs_link)
-    - [PRIME Intellect](https://app.primeintellect.ai/dashboard/create-cluster?image=axolotl&location=Cheapest&security=Cheapest&show_spot=true)
-    - [Modal](https://www.modal.com?utm_source=github&utm_medium=github&utm_campaign=axolotl)
-    - [Novita](https://novita.ai/gpus-console?templateId=311)
-    - [JarvisLabs.ai](https://jarvislabs.ai/templates/axolotl)
-    - [Latitude.sh](https://latitude.sh/blueprint/989e0e79-3bf6-41ea-a46b-1f246e309d5c)
+  - [Latitude.sh](https://latitude.sh/blueprint/989e0e79-3bf6-41ea-a46b-1f246e309d5c)
+  - [JarvisLabs.ai](https://jarvislabs.ai/templates/axolotl)
+  - [RunPod](https://runpod.io/gsc?template=v2ickqhz9s&ref=6i7fkpdz)
+  - [Novita](https://novita.ai/gpus-console?templateId=311)

 ### Google Colab {#sec-colab}

-[![](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/axolotl-ai-cloud/axolotl/blob/main/examples/colab-notebooks/colab-axolotl-example.ipynb#scrollTo=msOCO4NRmRLa)
+Use our [example notebook](../examples/colab-notebooks/colab-axolotl-example.ipynb).

 ## Platform-Specific Instructions {#sec-platform-specific}

 ### 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 +156,10 @@ 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/mixed_precision.qmd
+++ b/docs/mixed_precision.qmd
@@ -1,149 +0,0 @@
---
-title: "Mixed Precision Training"
-format:
-  html:
-    toc: true
-    toc-depth: 3
-    number-sections: true
-    code-tools: true
-execute:
-  enabled: false
---
-
-Mixed precision training uses lower precision data types to reduce memory usage and increase training speed while maintaining model quality. Axolotl supports several mixed precision formats:
-
- **FP16** - Half precision 16-bit (Pascal generation+)
- **BF16** - Brain Float 16-bit (Ampere generation+)
- **FP8** - 8-bit floating point (Hopper generation+)
-
-## FP16 Mixed Precision {#sec-fp16}
-
-### Overview {#sec-fp16-overview}
-
-FP16 is the traditional half-precision format, supported on older GPUs but can be less numerically stable than BF16.
-
-### Configuration {#sec-fp16-config}
-
-```{.yaml}
-fp16: true
-```
-
-### FP16 Considerations {#sec-fp16-considerations}
-
- May require gradient scaling to prevent underflow
- Less numerically stable than BF16
- Can cause training instability with some model architectures
- Consider using BF16 if your hardware supports it
-
-## BF16 Mixed Precision {#sec-bf16}
-
-### Overview {#sec-bf16-overview}
-
-BF16 (Brain Float 16) offers better numerical stability than FP16 and is the recommended mixed precision format for modern GPUs. It provides the same dynamic range as FP32 while using half the memory.
-
-### Configuration {#sec-bf16-config}
-
-```{.yaml}
-# Automatic BF16 detection (recommended)
-bf16: auto
-
-# Or explicitly enable
-bf16: true
-
-# For evaluation with BF16
-bf16: full  # Equivalent to bf16_full_eval in the HF trainer
-```
-
-## FP8 Mixed Precision {#sec-fp8}
-
-::: {.callout-note}
-FP8 support is experimental and requires compatible hardware (H100, H200) and recent PyTorch versions with TorchAO.
-:::
-
-### What is FP8? {#sec-fp8-overview}
-
-FP8 (8-bit floating point) can provide significant time savings compared to FP16/BF16 while maintaining training stability. Axolotl's implementation uses PyTorch's TorchAO library with "tensorwise" scaling strategy.
-
-### Requirements {#sec-fp8-software}
-
- Hopper+ GPUs (H100/H200)
- PyTorch 2.7+ (+ compatible TorchAO version)
- CUDA 12.4+
-
-### Configuration {#sec-fp8-config}
-
-Add to your YAML config:
-
-```{.yaml}
-# Enable FP8 mixed precision
-fp8: true
-
-# Optional: Enable FP8 for FSDP all-gather operations
-fp8_enable_fsdp_float8_all_gather: true
-
-# Enable torch.compile (almost always necessary for FP8 speedups)
-torch_compile: true
-```
-
-::: {.callout-important}
-**torch.compile is critical for FP8 performance**
-
-FP8 training requires `torch_compile: true` to see meaningful speedups. Without compilation, FP8 may actually be slower and use more memory than FP16/BF16.
-:::
-
-### Advanced FP8 Configs {#sec-fp8-advanced}
-
-For [FSDP](multi-gpu.qmd#sec-fsdp) (Fully Sharded Data Parallel) training:
-
-```{.yaml}
-fp8: true
-fp8_enable_fsdp_float8_all_gather: true
-
-torch_compile: true
-
-# FSDP configuration
-fsdp_version: 2
-fsdp_config:
-  offload_params: false
-  cpu_ram_efficient_loading: true
-  auto_wrap_policy: TRANSFORMER_BASED_WRAP
-  transformer_layer_cls_to_wrap: LlamaDecoderLayer
-  state_dict_type: FULL_STATE_DICT
-  reshard_after_forward: true
-```
-
-## Best Practices {#sec-best-practices}
-
-### Choosing Precision Format {#sec-choosing-format}
-
- **Start with automatic detection**: `bf16: auto`
- **For Hopper+ (H100/H200)**: Try FP8 + torch.compile for maximum speed
- **For Ampere (A100/RTX 30/40)**: Use BF16
- **For older Pascal/Turing GPUs**: Use FP16 with caution
- **For very old or unsupported GPUs**: Use FP32
-
-### Validation and Testing {#sec-validation}
-
-Always validate your mixed precision setup:
-
- **Start with a small dataset** to verify stability
- **Monitor loss curves** for irregularities
- **Compare with FP32 baseline** when possible
- **Test evaluation metrics** match expectations
-
-### FP8 Particulars {#sec-fp8-details}
-
- Use cases
-  - Single GPU training
-  - Multi GPU training with FSDP2 or Deepspeed
- Speedups
-  - Please refer to the [TorchAO FP8 training benchmarks](https://github.com/pytorch/ao/tree/main/torchao/float8#rowwise-scaling) for expected matmul speedups for different (M, K, N) settings
-  - Concrete number for LLaMA 3 8B training can be found [here](https://github.com/pytorch/ao/tree/main/torchao/float8#training-benchmarks)
- Known issues:
-  - FP8 + DDP + `torch.compile` (causes [error](https://gist.github.com/djsaunde/0c1664c32e44a64d31b5e01b4aafe5c4))
-  - FP8 + FSDP2 + `torch.compile` + FSDP2 activation checkpointing tends to be _slower_ than the BF16 equivalent training
-  - Flash Attention 2 does not play nicely with `torch.compile`
-
-See `examples/llama-3/3b-fp8-fsdp2.yaml` for an optimized example config. Enabling FP8 mixed precision + FP8 all-gather training results in ~10% faster iterations per second vs. BF16 for a relatively small (3B param) model
-
-For more information on multi-GPU training, see our [Multi-GPU guide](multi-gpu.qmd).
--- a/docs/multi-gpu.qmd
+++ b/docs/multi-gpu.qmd
@@ -23,6 +23,8 @@ Axolotl supports several methods for multi-GPU training:

 ## DeepSpeed {#sec-deepspeed}

+DeepSpeed is the recommended approach for multi-GPU training due to its stability and performance. It provides various optimization levels through ZeRO stages.
+
 ### Configuration {#sec-deepspeed-config}

 Add to your YAML config:
@@ -30,6 +32,7 @@ Add to your YAML config:
 ```{.yaml}
 deepspeed: deepspeed_configs/zero1.json
 ```
+
 ### Usage {#sec-deepspeed-usage}

 ```{.bash}
@@ -63,66 +66,9 @@ Start from Stage 1 -> Stage 2 -> Stage 3.

 :::

-## Fully Sharded Data Parallel (FSDP) {#sec-fsdp}
+## FSDP {#sec-fsdp}

-::: {.callout-note}
-
-FSDP2 is recommended for new users. FSDP1 is deprecated and will be removed in an upcoming release of Axolotl.
-
-:::
-
-### Migrating from FSDP1 to FSDP2 {#sec-migrate-fsdp1-fsdp2}
-
-To migrate your config from FSDP1 to FSDP2, you must use the `fsdp_version` top-level config field to specify the FSDP version, and
-also follow the config field mapping below to update field names.
-
-#### Config mapping
-
-FSDP1 | FSDP2
-------- | --------
-fsdp_sharding_strategy | reshard_after_forward
-fsdp_backward_prefetch_policy | **REMOVED**
-fsdp_backward_prefetch | **REMOVED**
-fsdp_forward_prefetch | **REMOVED**
-fsdp_sync_module_states | **REMOVED**
-fsdp_cpu_ram_efficient_loading | cpu_ram_efficient_loading
-fsdp_state_dict_type | state_dict_type
-fsdp_use_orig_params | **REMOVED**
-
-For more details, please see the migration guide in the [torchtitan repo](https://github.com/pytorch/torchtitan/blob/main/docs/fsdp.md). In Axolotl,
-if you were using the following FSDP1 config:
-
-```{.yaml}
-fsdp_version: 1
-fsdp_config:
-  fsdp_offload_params: false
-  fsdp_cpu_ram_efficient_loading: true
-  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
-  fsdp_transformer_layer_cls_to_wrap: Qwen3DecoderLayer
-  fsdp_state_dict_type: FULL_STATE_DICT
-  fsdp_sharding_strategy: FULL_SHARD
-```
-
-You can migrate to the following FSDP2 config:
-
-```{.yaml}
-fsdp_version: 2
-fsdp_config:
-  offload_params: false
-  cpu_ram_efficient_loading: true
-  auto_wrap_policy: TRANSFORMER_BASED_WRAP
-  transformer_layer_cls_to_wrap: Qwen3DecoderLayer
-  state_dict_type: FULL_STATE_DICT
-  reshard_after_forward: true
-```
-
-### FSDP1 (deprecated) {#sec-fsdp-config}
-
-::: {.callout-note}
-
-Using `fsdp` to configure FSDP is deprecated and will be removed in an upcoming release of Axolotl. Please use `fsdp_config` as above instead.
-
-:::
+### Basic FSDP Configuration {#sec-fsdp-config}

 ```{.yaml}
 fsdp:
@@ -134,7 +80,6 @@ fsdp_config:
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
 ```

-
 ## Sequence parallelism {#sec-sequence-parallelism}

 We support sequence parallelism (SP) via the
--- a/docs/multi-node.qmd
+++ b/docs/multi-node.qmd
@@ -40,13 +40,13 @@ use_cpu: false

 Configure your model to use FSDP in the Axolotl yaml. For example:
 ```yaml
-fsdp_version: 2
+fsdp:
+  - full_shard
+  - auto_wrap
 fsdp_config:
-  offload_params: true
-  state_dict_type: FULL_STATE_DICT
-  auto_wrap_policy: TRANSFORMER_BASED_WRAP
-  transformer_layer_cls_to_wrap: LlamaDecoderLayer
-  reshard_after_forward: true
+  fsdp_offload_params: true
+  fsdp_state_dict_type: FULL_STATE_DICT
+  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
 ```

 All you have to do now is launch using accelerate as you would usually do on each machine and voila, the processes will start once you have launched accelerate on every machine.
@@ -69,19 +69,11 @@ export NCCL_BUFFSIZE=2097152

 Run the following on each node:

-### Option 1: New Axolotl CLI with launcher args (Recommended)
-
-```bash
-axolotl train config.yaml --launcher torchrun -- --nnodes $num_nodes --nproc_per_node $gpu_per_node --rdzv_id $rdzv_id --rdzv_backend c10d --rdzv_endpoint "$head_node_ip:$head_node_port"
-```
-
-### Option 2: Direct torchrun (Legacy)
-
 ```bash
 torchrun --nnodes $num_nodes --nproc_per_node $gpu_per_node --rdzv_id $rdzv_id --rdzv_backend c10d --rdzv_endpoint "$head_node_ip:$head_node_port" -m axolotl.cli.train config.yaml
 ```

-Please make sure to substitute the placeholder variables:
+Please make sure to substitute the placeholder variables.

 - `num_nodes`: Number of nodes (containing GPUs)
 - `gpu_per_node`: Number of gpus per node
@@ -89,6 +81,8 @@ Please make sure to substitute the placeholder variables:
 - `head_node_port`: Port of the head node (make sure other machines can connect to this. Default 29400)
 - `rdzv_id`: A unique job ID that is used by the job across nodes.

-The new CLI approach (Option 1) is recommended as it provides consistent argument handling and works seamlessly with other Axolotl CLI features.
+::: {.callout-note}
+You need to call `axolotl.cli.train` instead of `axolotl train` as the latter calls accelerate under the hood
+:::

 More info on the available configs can be found on the Pytorch docs [here](https://pytorch.org/docs/stable/elastic/run.html)
--- a/docs/multimodal.qmd
+++ b/docs/multimodal.qmd
@@ -13,14 +13,10 @@ format:
 - [Pixtral](#sec-pixtral)
 - [Llava-1.5](#sec-llava-15)
 - [Mistral-Small-3.1](#sec-mistral-small-31)
- [Magistral-Small-2509](#sec-magistral-small-2509)
- [Voxtral](#sec-voxtral)
 - [Gemma-3](#sec-gemma-3)
- [Gemma-3n](#sec-gemma-3n)
 - [Qwen2-VL](#sec-qwen2-vl)
 - [Qwen2.5-VL](#sec-qwen25-vl)
- [SmolVLM2](#sec-smolvlm2)
- [LFM2-VL](#sec-lfm2-vl)
+- [Phi3-V](#sec-phi3-v)

 ## Usage

@@ -35,13 +31,14 @@ skip_prepare_dataset: true
 remove_unused_columns: false  # leave columns in place as they are needed to handle image embeddings during training
 sample_packing: false  # not yet supported with multimodal

-chat_template:  # see in next section if specified
+chat_template:  # see in next section

 # example dataset
 datasets:
  - path: HuggingFaceH4/llava-instruct-mix-vsft
    type: chat_template
    split: train[:1%]
+    field_messages: messages

 # (optional) if doing lora, only finetune the Language model,
 # leave the vision model and vision tower frozen
@@ -94,32 +91,10 @@ 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'`
-:::
-
 ```yaml
 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'`
-:::
-
-```yaml
-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'`
-:::
-
-```yaml
-base_model: mistralai/Voxtral-Mini-3B-2507
+chat_template: mistral_v7_tekken
 ```

 ### Gemma-3 {#sec-gemma-3}
@@ -136,22 +111,6 @@ base_model: google/gemma-3-4b-it
 chat_template: gemma3
 ```

-### Gemma-3n {#sec-gemma-3n}
-
-::: {.callout-warning}
-The model's initial loss and grad norm will be very high. We suspect this to be due to the Conv in the vision layers.
-:::
-
-::: {.callout-tip}
-Please make sure to install `timm` via `uv pip install timm==1.0.17`
-:::
-
-```yaml
-base_model: google/gemma-3n-E2B-it
-
-chat_template: gemma3n
-```
-
 ### Qwen2-VL {#sec-qwen2-vl}

 ```yaml
@@ -168,24 +127,13 @@ base_model: Qwen/Qwen2.5-VL-7B-Instruct
 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`
-:::
+### Phi3-V {#sec-phi3-v}

 ```yaml
-base_model: HuggingFaceTB/SmolVLM2-500M-Video-Instruct
-```
+base_model: microsoft/Phi-3.5-vision-instruct

-### LFM2-VL {#sec-lfm2-vl}
-
-::: {.callout-warning}
-Please uninstall `causal-conv1d` via `uv pip uninstall -y causal-conv1d`
-:::
-
-```yaml
-base_model: LiquidAI/LFM2-VL-450M
+trust_remote_code: true
+chat_template: phi_35_vl
 ```

 ## Dataset Format
@@ -194,9 +142,7 @@ For multi-modal datasets, we adopt an extended `chat_template` format similar to

 - A message is a list of `role` and `content`.
 - `role` can be `system`, `user`, `assistant`, etc.
- `content` is a list of `type` and (`text`, `image`, `path`, `url`, `base64`, or `audio`).
-
-### Image
+- `content` is a list of `type` and (`text` or `image` or `path` or `url` or `base64`).

 ::: {.callout-note}
 For backwards compatibility:
@@ -205,43 +151,15 @@ For backwards compatibility:
 - If `content` is a string, it will be converted to a list with `type` as `text`.
 :::

+::: {.callout-tip}
 For image loading, you can use the following keys within `content` alongside `"type": "image"`:

 - `"path": "/path/to/image.jpg"`
 - `"url": "https://example.com/image.jpg"`
 - `"base64": "..."`
 - `"image": PIL.Image`
-
-### Audio
-
-For audio loading, you can use the following keys within `content` alongside `"type": "audio"`:
-
- `"path": "/path/to/audio.mp3"`
- `"url": "https://example.com/audio.mp3"`
- `"audio": np.ndarray`
-
-::: {.callout-tip}
-
-You may need to install `librosa` via `uv pip install librosa==0.11.0`.
-
 :::

-### Video
-
-::: {.callout-warning}
-
-This is not well tested at the moment. We welcome contributors!
-
-:::
-
-For video loading, you can use the following keys within `content` alongside `"type": "video"`:
-
- `"path": "/path/to/video.mp4"`
- `"url": "https://example.com/video.mp4"`
- `"video": np.ndarray | list[PIL.Image.Image] | torch.Tensor` (or list of the aforementioned)
-
-### Example
-
 Here is an example of a multi-modal dataset:
 ```json
 [
@@ -270,9 +188,3 @@ Here is an example of a multi-modal dataset:
  }
 ]
 ```
-
-## FAQ
-
-1. `PIL.UnidentifiedImageError: cannot identify image file ...`
-
-`PIL` could not retrieve the file at `url` using `requests`. Please check for typo. One alternative reason is that the request is blocked by the server.
--- a/docs/nd_parallelism.qmd
+++ b/docs/nd_parallelism.qmd
@@ -1,108 +0,0 @@
---
-title: "N-D Parallelism (Beta)"
---
-
-Axolotl enables training models at scale by composing different parallelism techniques. This is essential when:
-
- A model's weights are too large to fit on a single GPU's memory.
- A model's activations, especially with very long contexts, are too large for a single GPU.
- You want to accelerate training by using multiple GPUs or nodes.
-
-or combinations of the above!
-
-## Core Concepts
-
-Parallelism strategies can be combined. The key is understanding how each one divides the workload. PyTorch's `DeviceMesh` is the modern way to manage these combinations, creating a logical grid of your GPUs and assigning different parallel strategies to different dimensions of the grid.
-
-### Data Parallelism {#sec-dp}
-
-Data Parallelism focuses on splitting the global data batch across GPUs.
-
- Distributed Data Parallel (DDP): The classic approach. The full model is replicated on every GPU. Each GPU processes a different slice of the data batch. Gradients are then averaged across all GPUs after the backward pass to keep the models synchronized. This can substantially improve data throughput compared to single-device training, but requires that each GPU is able to hold the entire model, its gradients, and optimizer states.
-
- [Fully Sharded Data Parallel (FSDP)](multi-gpu.qmd#fully-sharded-data-parallel-(fsdp)): A highly memory-efficient form of data parallelism (inspired by DeepSpeed's ZeRO). Instead of replicating the model, FSDP shards the model's *parameters, gradients, and optimizer states* across the GPUs in the data-parallel group. During computation, each GPU receives the specific parameters it needs via an `all_gather` operation just before they are used, and they can be discarded immediately after (`reshard-after-forward`).
-    - FSDP maps to ZeRO stages:
-        - ZeRO-2 (`reshard_after_forward=False`): Shards gradients and optimizer states. Model weights are replicated on each GPU.
-        - ZeRO-3 (`reshard_after_forward=True`): Shards gradients, optimizer states, AND model parameters. This provides the most memory savings at the cost of more communication (re-gathering parameters for both forward and backward passes).
-
-### [Experimental] Tensor Parallelism (TP) {#sec-tp}
-
-Also known as "horizontal model parallelism," as described in the [Megatron-LM paper](https://arxiv.org/pdf/1909.08053.pdf). Instead of splitting the batch, TP splits the model's layers themselves across GPUs.
-
- How it works: For a linear layer `Y = XA`, the weight matrix `A` is split column-wise (`A = [A_1, A_2]`). The computation becomes `Y_1 = XA_1` and `Y_2 = XA_2`, which can happen in parallel on different GPUs. The final output `Y` is simply the concatenation of `Y_1` and `Y_2`. Check [this comment](https://github.com/huggingface/transformers/issues/10321#issuecomment-783543530) for more detailed info.
- Requirement: TP involves frequent, small communications within a forward/backward pass. It requires a very fast interconnect between GPUs (e.g., NVLink) and is typically not recommended across different nodes.
-
-### Context Parallelism (CP) {#sec-cp}
-
-Context Parallelism, also called [Sequence Parallelism](sequence_parallelism.qmd), addresses the memory bottleneck from long sequences. The input sequence itself is split along the sequence length dimension and distributed across GPUs.
-
- How it works: If you have a sequence of 8192 tokens and a `context_parallel_size` of 4, each GPU will only handle a chunk of 2048 tokens.
- The Challenge: Attention is not local; every token needs to "attend to" every other token. Splitting the sequence breaks this.
- The Solution (`ring-flash-attention`): An efficient communication protocol is used. To compute attention for its local sequence chunk, each GPU passes its Key-Value (KV) cache to its neighbor in a "ring." After `N-1` steps, every GPU has seen the KV-cache from all other GPUs, allowing it to compute the correct attention values for its chunk. This is implemented using the highly optimized `flash-attention` kernel at each step.
-
-### Hybrid Sharding Data Parallel (HSDP) {#sec-hsdp}
-
-HSDP is a 2D strategy that intelligently combines FSDP and DDP, typically for multi-node training.
-
- Intra-Node (within a machine): Use FSDP. This is efficient because GPUs on the same node have fast interconnects (NVLink), making the `all_gather` operations for sharded parameters fast.
- Inter-Node (across machines): Use DDP. The gradient synchronization between nodes is less frequent than FSDP's parameter gathering, making it a better fit for the slower node-to-node network (e.g., Ethernet/Infiniband).
- Example: With 2 nodes of 8 GPUs each (16 total), you could have `dp_shard_size=8` (FSDP within each node) and `dp_replicate_size=2` (DDP across the two nodes).
-
-## Usage
-
-```yaml
-# FSDP config. See https://docs.axolotl.ai/docs/multi-gpu.html#sec-fsdp
-fsdp_version: 2
-fsdp_config:
-  # ...
-
-# The number of GPUs to shard the model parameters across (FSDP dimension).
-dp_shard_size: 4
-
-# The number of times to replicate the sharded model (DDP dimension).
-dp_replicate_size: 2
-
-# Number of GPUs for Tensor Parallelism.
-tensor_parallel_size: 1  # (default is 1, no TP)
-
-# Number of GPUs for Context/Sequence Parallelism.
-context_parallel_size: 1 # (default is 1, no CP)
-```
-
-Note: We recommend FSDP. DeepSpeed is only compatible with `tensor_parallel_size`.
-
-## Examples
-
-::: {.callout-tip}
-See our example configs [here](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/distributed-parallel).
-:::
-
-1.  HSDP on 2 nodes with 4 GPUs each (8 GPUs total):
-    - You want FSDP within each node and DDP across nodes.
-    - Set `dp_shard_size: 4` and `dp_replicate_size: 2`.
-
-2.  FSDP + TP on a single 8-GPU node:
-    - You want to split the model across 4 GPUs using FSDP, and further split each layer across 2 GPUs with TP.
-    - Set `dp_shard_size: 4` and `tensor_parallel_size: 2`.
-
-3.  FSDP + CP on a single 8-GPU node for long context:
-    - You want to shard the model across all 8 GPUs and also split the sequence length across all 8 GPUs.
-    - Set `dp_shard_size: 8` and `context_parallel_size: 8`. Note: this means the data parallel group and context parallel group are the same. A more common setup might be to shard across a smaller group.
-
-## Support Matrix
-
-This matrix describes how different parallelism methods can be combined in Axolotl.
-
-| Combination | `dp_replicate_size` | `dp_shard_size` | `tp_size` | `cp_size` | Status & Notes |
-| --- | :---: | :---: |:---:|:---:|---|
-| **FSDP** (ZeRO-3) | 1 | >1 | 1 | 1 | ✅ Fully supported. Shards model across all GPUs. |
-| **HSDP** | >1 | >1 | 1 | 1 | ✅ Fully supported. FSDP intra-node, DDP inter-node. |
-| **FSDP + TP** | 1 | >1 | >1 | 1 | ✅ **2D Parallelism**. Shards the model across a `dp_shard` group, and TP-splits layers within the `tp` group. |
-| **HSDP + TP** | >1 | >1 | >1 | 1 | ✅ **3D Parallelism**. A powerful but complex combination. |
-| **FSDP + CP** | 1 | >1 | 1 | >1 | ✅ **2D Parallelism**. Combines FSDP with context parallelism. |
-| **FSDP + TP + CP**| 1 | >1 | >1| >1| ✅ **3D Parallelism**. Another advanced combination. |
-| DDP + TP/CP | >1 | 1 | >1 | >1 | ❌ **Not Supported**. The `ParallelismConfig` explicitly prevents this, as composing pure DDP with TP or CP is currently not supported. You should use FSDP + TP/CP instead (`dp_shard_size > 1`). |
-| Just TP / CP | 1 | 1 | >1 | >1 | ✅ Supported. Useful for inference or when the model fits on one GPU but context is too long. |
-
- `tp_size` refers to `tensor_parallel_size`
- `cp_size` refers to `context_parallel_size`
--- a/docs/optimizations.qmd
+++ b/docs/optimizations.qmd
@@ -1,133 +0,0 @@
---
-title: Optimizations Guide
-description: A guide to the performance and memory optimizations available in Axolotl.
---
-
-Axolotl includes numerous optimizations to speed up training, reduce memory usage, and handle large models.
-
-This guide provides a high-level overview and directs you to the detailed documentation for each feature.
-
-## Speed Optimizations
-
-These optimizations focus on increasing training throughput and reducing total training time.
-
-### Sample Packing
-
-Improves GPU utilization by combining multiple short sequences into a single packed sequence for training. This requires enabling one of the [attention](#attention-implementations) implementations below.
-
- **Config:** `sample_packing: true`
- **Learn more:** [Sample Packing](multipack.qmd)
-
-### Attention Implementations
-
-Using an optimized attention implementation is critical for training speed.
-
- **[Flash Attention 2](https://github.com/Dao-AILab/flash-attention)**: `flash_attention: true`. **(Recommended)** The industry standard for fast attention on modern GPUs. Requires Ampere or higher. For AMD, check [AMD Support](https://github.com/Dao-AILab/flash-attention?tab=readme-ov-file#amd-rocm-support).
- **[Flex Attention](https://pytorch.org/blog/flexattention/)**: `flex_attention: true`.
- **[SDP Attention](https://docs.pytorch.org/docs/stable/generated/torch.nn.functional.scaled_dot_product_attention.html)**: `sdp_attention: true`. PyTorch's native implementation.
- **[Xformers](https://github.com/facebookresearch/xformers)**: `xformers_attention: true`. Works with FP16.
-
-*Note: You should only enable one attention backend.*
-
-### LoRA Optimizations
-
-Leverages optimized kernels to accelerate LoRA training and reduce memory usage.
-
- **Learn more:** [LoRA Optimizations Documentation](lora_optims.qmd)
-
-## Memory Optimizations
-
-These techniques help you fit larger models or use bigger batch sizes on your existing hardware.
-
-### Parameter Efficient Finetuning (LoRA & QLoRA)
-
-Drastically reduces memory by training a small set of "adapter" parameters instead of the full model. This is the most common and effective memory-saving technique.
-
- Examples: Find configs with `lora` or `qlora` in the [examples directory](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/llama-3).
- Config Reference: See `adapter`, `load_in_4bit`, and `load_in_8bit` in the [Configuration Reference](config-reference.qmd).
-
-### Gradient Checkpointing & Activation Offloading
-
-These techniques save VRAM by changing how activations are handled.
-
- Gradient Checkpointing: re-computes activations during the backward pass, trading compute time for VRAM.
- Activation Offloading: moves activations to CPU RAM or disk, trading I/O overhead for VRAM.
- Learn more: [Gradient Checkpointing and Offloading Docs](gradient_checkpointing.qmd)
-
-### Cut Cross Entropy (CCE)
-
-Reduces VRAM usage by using an optimized cross-entropy loss calculation.
-
- **Learn more:** [Custom Integrations - CCE](custom_integrations.qmd#cut-cross-entropy)
-
-### Liger Kernels
-
-Provides efficient Triton kernels to improve training speed and reduce memory usage.
-
- **Learn more:** [Custom Integrations - Liger Kernels](custom_integrations.qmd#liger-kernels)
-
-## Long Context Models
-
-Techniques to train models on sequences longer than their original context window.
-
-### RoPE Scaling
-
-Extends a model's context window by interpolating its Rotary Position Embeddings.
-
- **Config:** Pass the `rope_scaling` config under the `overrides_of_model_config: `. To learn how to set RoPE, check the respective model config.
-
-### Sequence Parallelism
-
-Splits long sequences across multiple GPUs, enabling training with sequence lengths that would not fit on a single device.
-
- **Learn more:** [Sequence Parallelism Documentation](sequence_parallelism.qmd)
-
-### Artic Long Sequence Training (ALST)
-
-ALST is a recipe that combines several techniques to train long-context models efficiently. It typically involves:
-
- TiledMLP to reduce memory usage in MLP layers.
- Tiled Loss functions (like [CCE](#cut-cross-entropy-(cce) or [Liger](#liger-kernels)).
- Activation Offloading to CPU.
-
- Example: [ALST Example Configuration](https://github.com/axolotl-ai-cloud/axolotl/tree/main/examples/alst)
-
-## Large Models (Distributed Training)
-
-To train models that don't fit on a single GPU, you'll need to use a distributed training strategy like FSDP or DeepSpeed. These frameworks shard the model weights, gradients, and optimizer states across multiple GPUs and nodes.
-
- **Learn more:** [Multi-GPU Guide](multi-gpu.qmd)
- **Learn more:** [Multi-Node Guide](multi-node.qmd)
-
-### N-D Parallelism (Beta)
-
-For advanced scaling, Axolotl allows you to compose different parallelism techniques (e.g., Data, Tensor, Sequence Parallelism). This is a powerful approach to train an extremely large model by overcoming multiple bottlenecks at once.
-
- **Learn more:** [N-D Parallelism Guide](nd_parallelism.qmd)
-
-
-## Quantization
-
-Techniques to reduce the precision of model weights for memory savings.
-
-### 4-bit Training (QLoRA)
-
-The recommended approach for quantization-based training. It loads the base model in 4-bit using `bitsandbytes` and then trains QLoRA adapters. See [Adapter Finetuning](#adapter-finetuning-lora-qlora) for details.
-
-### FP8 Training
-
-Enables training with 8-bit floating point precision on supported hardware (e.g., NVIDIA Hopper series GPUs) for significant speed and memory gains.
-
- **Example:** [Llama 3 FP8 FSDP Example](https://github.com/axolotl-ai-cloud/axolotl/blob/main/examples/llama-3/3b-fp8-fsdp2.yaml)
-
-### Quantization Aware Training (QAT)
-
-Simulates quantization effects during training, helping the model adapt and potentially improving the final accuracy of the quantized model.
-
- **Learn more:** [QAT Documentation](qat.qmd)
-
-### GPTQ
-
-Allows you to finetune LoRA adapters on top of a model that has already been quantized using the GPTQ method.
-
- **Example:** [GPTQ LoRA Example](https://github.com/axolotl-ai-cloud/axolotl/blob/main/examples/llama-2/gptq-lora.yml)
--- a/docs/optimizers.qmd
+++ b/docs/optimizers.qmd
@@ -1,129 +0,0 @@
---
-title: Optimizers
-description: Configuring optimizers
---
-
-## Overview
-
-Axolotl supports all optimizers supported by [transformers OptimizerNames](https://github.com/huggingface/transformers/blob/51f94ea06d19a6308c61bbb4dc97c40aabd12bad/src/transformers/training_args.py#L142-L187)
-
-Here is a list of optimizers supported by transformers as of `v4.54.0`:
-
- `adamw_torch`
- `adamw_torch_fused`
- `adamw_torch_xla`
- `adamw_torch_npu_fused`
- `adamw_apex_fused`
- `adafactor`
- `adamw_anyprecision`
- `adamw_torch_4bit`
- `adamw_torch_8bit`
- `ademamix`
- `sgd`
- `adagrad`
- `adamw_bnb_8bit`
- `adamw_8bit`  # alias for adamw_bnb_8bit
- `ademamix_8bit`
- `lion_8bit`
- `lion_32bit`
- `paged_adamw_32bit`
- `paged_adamw_8bit`
- `paged_ademamix_32bit`
- `paged_ademamix_8bit`
- `paged_lion_32bit`
- `paged_lion_8bit`
- `rmsprop`
- `rmsprop_bnb`
- `rmsprop_bnb_8bit`
- `rmsprop_bnb_32bit`
- `galore_adamw`
- `galore_adamw_8bit`
- `galore_adafactor`
- `galore_adamw_layerwise`
- `galore_adamw_8bit_layerwise`
- `galore_adafactor_layerwise`
- `lomo`
- `adalomo`
- `grokadamw`
- `schedule_free_radam`
- `schedule_free_adamw`
- `schedule_free_sgd`
- `apollo_adamw`
- `apollo_adamw_layerwise`
- `stable_adamw`
-
-
-## Custom Optimizers
-
-Enable custom optimizers by passing a string to the `optimizer` argument. Each optimizer will receive beta and epsilon args, however, some may accept additional args which are detailed below.
-
-### optimi_adamw
-
-```yaml
-optimizer: optimi_adamw
-```
-
-### ao_adamw_4bit
-
-Deprecated: Please use `adamw_torch_4bit`.
-
-### ao_adamw_8bit
-
-Deprecated: Please use `adamw_torch_8bit`.
-
-### ao_adamw_fp8
-
-
-```yaml
-optimizer: ao_adamw_fp8
-```
-
-### adopt_adamw
-
-GitHub: [https://github.com/iShohei220/adopt](https://github.com/iShohei220/adopt)
-Paper: [https://arxiv.org/abs/2411.02853](https://arxiv.org/abs/2411.02853)
-
-```yaml
-optimizer: adopt_adamw
-```
-
-### came_pytorch
-
-GitHub: [https://github.com/yangluo7/CAME/tree/master](https://github.com/yangluo7/CAME/tree/master)
-Paper: [https://arxiv.org/abs/2307.02047](https://arxiv.org/abs/2307.02047)
-
-```yaml
-optimizer: came_pytorch
-
-# optional args (defaults below)
-adam_beta1: 0.9
-adam_beta2: 0.999
-adam_beta3: 0.9999
-adam_epsilon: 1e-30
-adam_epsilon2: 1e-16
-```
-
-### muon
-
-Blog: [https://kellerjordan.github.io/posts/muon/](https://kellerjordan.github.io/posts/muon/)
-Paper: [https://arxiv.org/abs/2502.16982v1](https://arxiv.org/abs/2502.16982v1)
-
-```yaml
-optimizer: muon
-```
-
-### dion
-
-Microsoft's Dion (DIstributed OrthoNormalization) optimizer is a scalable and communication-efficient
-orthonormalizing optimizer that uses low-rank approximations to reduce gradient communication.
-
-GitHub: [https://github.com/microsoft/dion](https://github.com/microsoft/dion)
-Paper: [https://arxiv.org/pdf/2504.05295](https://arxiv.org/pdf/2504.05295)
-Note: Implementation written for PyTorch 2.7+ for DTensor
-
-```yaml
-optimizer: dion
-dion_lr: 0.01
-dion_momentum: 0.95
-lr: 0.00001  # learning rate for embeddings and parameters that fallback to AdamW
-```
--- a/docs/qat.qmd
+++ b/docs/qat.qmd
@@ -23,18 +23,10 @@ To enable QAT in axolotl, add the following to your configuration file:

 ```yaml
 qat:
-  activation_dtype: # Optional[str] = "int8". Fake quantization layout to use for activation quantization. Valid options are "int4", "int8", "float8"
-  weight_dtype: # Optional[str] = "int8". Fake quantization layout to use for weight quantization. Valid options are "int4", "fp8", and "nvfp4".
+  activation_dtype: # Optional[str] = "int8". Fake quantization layout to use for activation quantization. Valid options are "int4" and "int8"
+  weight_dtype: # Optional[str] = "int8". Fake quantization layout to use for weight quantization. Valid options are "int4" and "int8"
  group_size: # Optional[int] = 32. The number of elements in each group for per-group fake quantization
  fake_quant_after_n_steps: # Optional[int] = None. The number of steps to apply fake quantization after
 ```

-We support the following quantization schemas:
-
- `Int4WeightOnly` (requires the `fbgemm-gpu` extra when installing Axolotl)
- `Int8DynamicActivationInt4Weight`
- `Float8DynamicActivationFloat8Weight`
- `Float8DynamicActivationInt4Weight`
- `NVFP4`
-
 Once you have finished training, you must quantize your model by using the same quantization configuration which you used to train the model with. You can use the [`quantize`](./quantize.qmd) command to do this.
--- a/docs/quantize.qmd
+++ b/docs/quantize.qmd
@@ -22,8 +22,8 @@ Quantization is configured using the `quantization` key in your configuration fi
 ```yaml
 base_model: # The path to the model to quantize.
 quantization:
-  activation_dtype: # Optional[str] = "int8". Fake quantization layout to use for activation quantization. Valid options are "int4", "int8", "float8"
-  weight_dtype: # Optional[str] = "int8". Fake quantization layout to use for weight quantization. Valid options are "int4", "fp8", and "nvfp4".
+  weight_dtype: # Optional[str] = "int8". Fake quantization layout to use for weight quantization. Valid options are uintX for X in [1, 2, 3, 4, 5, 6, 7], or int4, or int8
+  activation_dtype: # Optional[str] = "int8". Fake quantization layout to use for activation quantization. Valid options are "int4" and "int8"
  group_size: # Optional[int] = 32. The number of elements in each group for per-group fake quantization
  quantize_embedding: # Optional[bool] = False. Whether to quantize the embedding layer.

@@ -39,8 +39,9 @@ you used to train the model:
 # qat.yml
 qat:
  activation_dtype: int8
-  weight_dtype: int4
+  weight_dtype: int8
  group_size: 256
+  quantize_embedding: true

 output_dir: # The path to the output directory used during training where the final checkpoint has been saved.
 ```
@@ -50,11 +51,3 @@ axolotl quantize qat.yml
 ```

 This ensures that an identical quantization configuration is used to quantize the model as was used to train it.
-
-
-::: {.callout-note}
-
-If you have configured pushing to hub with `hub_model_id`, your model hub name will have the quantization schema appended to it,
-e.g. `axolotl-ai-cloud/qat-nvfp4-llama3B` will become `axolotl-ai-cloud/qat-nvfp4-llama3B-nvfp4w`
-
-:::
--- a/docs/reward_modelling.qmd
+++ b/docs/reward_modelling.qmd
@@ -11,7 +11,6 @@ We support the reward modelling techniques supported by `trl`.
 ### (Outcome) Reward Models

 Outcome reward models are trained using data which contains preference annotations for an entire interaction between the user and model (e.g. rather than per-turn or per-step).
-For improved training stability, you can use the `center_rewards_coefficient` parameter to encourage mean-zero reward outputs ([see TRL docs](https://huggingface.co/docs/trl/v0.10.1/en/reward_trainer#centering-rewards)).

 ```yaml
 base_model: google/gemma-2-2b
--- a/docs/rlhf.qmd
+++ b/docs/rlhf.qmd
@@ -17,6 +17,7 @@ feedback. Various methods include, but not limited to:
 - [Kahneman-Tversky Optimization (KTO)](#kto)
 - [Odds Ratio Preference Optimization (ORPO)](#orpo)
 - [Group Relative Policy Optimization (GRPO)](#grpo)
+- Proximal Policy Optimization (PPO) (not yet supported in axolotl, if you're interested in contributing, please reach out!)


 ## RLHF using Axolotl
@@ -274,14 +275,15 @@ rl: dpo
 datasets:
  - path: ...
    split: train
-    type:
-      field_prompt: "prompt"
-      field_system: "system"
-      field_chosen: "chosen"
-      field_rejected: "rejected"
-      prompt_format: "{prompt}"
-      chosen_format: "{chosen}"
-      rejected_format: "{rejected}"
+    type: user_defined.default
+
+    field_prompt: "prompt"
+    field_system: "system"
+    field_chosen: "chosen"
+    field_rejected: "rejected"
+    prompt_format: "{prompt}"
+    chosen_format: "{chosen}"
+    rejected_format: "{rejected}"
 ```

 The input format is a simple JSON input with customizable fields based on the above config.
@@ -474,13 +476,14 @@ rl: kto
 datasets:
  - path: ...
    split: train
-    type:
-      field_prompt: "prompt"
-      field_system: "system"
-      field_completion: "completion"
-      field_label: "label"
-      prompt_format: "{prompt}"
-      completion_format: "{completion}"
+    type: user_defined.default
+
+    field_prompt: "prompt"
+    field_system: "system"
+    field_completion: "completion"
+    field_label: "label"
+    prompt_format: "{prompt}"
+    completion_format: "{completion}"
 ```

 The input format is a simple JSON input with customizable fields based on the above config.
--- a/docs/scripts/generate_config_docs.py
+++ b/docs/scripts/generate_config_docs.py
@@ -47,6 +47,7 @@ class QuartoGenerator:
        """Check if a type is a Pydantic BaseModel."""
        return inspect.isclass(type_obj) and issubclass(type_obj, BaseModel)

+    # pylint: disable=too-many-return-statements
    def _extract_nested_type(self, field_type) -> Any:
        """Extract the actual type from complex type annotations."""
        # Handle Annotated types (Python 3.9+)
@@ -123,6 +124,7 @@ class QuartoGenerator:

        return field_type

+    # pylint: disable=too-many-return-statements
    def _extract_all_pydantic_models_from_type(
        self, field_type
    ) -> list[type[BaseModel]]:
@@ -316,6 +318,7 @@ class QuartoGenerator:

        return all_groups

+    # pylint: disable=too-many-return-statements
    def _extract_field_groups_from_source(
        self, model_class: type[BaseModel]
    ) -> list[dict]:
@@ -500,7 +503,7 @@ class QuartoGenerator:
                    nested_schema = nested_model.model_json_schema()
                    nested_properties = nested_schema.get("properties", {})
                    nested_required = nested_schema.get("required", [])
-                except Exception:
+                except Exception:  # pylint: disable=broad-exception-caught
                    # Fallback: use model fields directly
                    nested_properties = {}
                    nested_required = []
@@ -604,7 +607,7 @@ class QuartoGenerator:
            schema = model_class.model_json_schema()
            properties = schema.get("properties", {})
            required = schema.get("required", [])
-        except Exception as e:
+        except Exception as e:  # pylint: disable=broad-exception-caught
            print(
                f"Warning: Could not generate JSON schema ({e}). Using model fields instead."
            )
--- a/docs/sequence_parallelism.qmd
+++ b/docs/sequence_parallelism.qmd
@@ -22,7 +22,7 @@ To enable sequence parallelism, add the following to your configuration file:

 ```yaml
 # Set to a divisor (> 1) of the number of GPUs available
-context_parallel_size: 4  # Split sequences across 4 GPUs
+sequence_parallel_degree: 4  # Split sequences across 4 GPUs
 # Optional; strides across the key dimension. Larger values use more memory but should make training faster.
 heads_k_stride: 1
 # Optional; one of "varlen_llama3" or "batch_ring". Defaults to
@@ -30,7 +30,7 @@ heads_k_stride: 1
 ring_attn_func:
 ```

-The `context_parallel_size` should be a divisor of the total number of GPUs. For example:
+The `sequence_parallel_degree` should be a divisor of the total number of GPUs. For example:

 - With 8 GPUs, valid values would be 2, 4, or 8
 - With 4 GPUs, valid values would be 2 or 4
@@ -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

@@ -66,7 +66,7 @@ sequence_len: 8192

 ...

-context_parallel_size: 4  # Split each sequence into 4 parts, one per GPU
+sequence_parallel_degree: 4  # Split each sequence into 4 parts, one per GPU
 # Optional; strides across the key dimension. Larger values use more memory but should make training faster.
 heads_k_stride: 1
 # Optional; one of "varlen_llama3" or "batch_ring". Defaults to
@@ -89,12 +89,12 @@ Sequence parallelism is compatible with Axolotl's sample packing functionality.

 ## Effect on Batch Size

-When using sequence parallelism, your effective global batch size is **divided** by the `context_parallel_size`. This happens because:
+When using sequence parallelism, your effective global batch size is **divided** by the `sequence_parallel_degree`. This happens because:

- Each group of `context_parallel_size` GPUs works on the same batch (just different parts of each sequence)
+- Each group of `sequence_parallel_degree` GPUs works on the same batch (just different parts of each sequence)
 - The number of batches processed per step decreases

 For example:
 - With 8 GPUs and no sequence parallelism: 8 different batches processed per step
- With 8 GPUs and `context_parallel_size=4`: Only 2 different batches processed per step (each split across 4 GPUs)
+- With 8 GPUs and `sequence_parallel_degree=4`: Only 2 different batches processed per step (each split across 4 GPUs)
 - If your per-GPU `micro_batch_size` is 2, the global batch size decreases from 16 to 4
--- a/docs/streaming.qmd
+++ b/docs/streaming.qmd
@@ -1,120 +0,0 @@
---
-title: Streaming Datasets
-description: How to use streaming mode for large-scale datasets and memory-efficient training
-order: 10
---
-
-Streaming enables memory-efficient training with large datasets by loading data
-incrementally rather than loading the entire dataset into memory at once.
-
-Use streaming when:
-
- Your dataset is too large to fit in memory (e.g. when you're doing pretraining with massive text corpora)
- You want to start training immediately without preprocessing the entire dataset
-
-Streaming works with both remote and locally stored datasets!
-
-::: {.callout-note}
-Streaming currently only supports a single dataset. Multi-dataset support will be added soon.
-:::
-
-
-## Configuration
-
-### Basic Streaming
-
-Enable streaming mode by setting the `streaming` flag:
-
-```yaml
-streaming: true
-```
-
-### Pretraining with Streaming
-
-For pretraining tasks, streaming is automatically enabled when using `pretraining_dataset`:
-
-```yaml
-pretraining_dataset:
-  - path: HuggingFaceFW/fineweb-edu
-    type: pretrain
-    text_column: text
-    split: train
-
-# Optionally, enable sample packing
-streaming_multipack_buffer_size: 10000
-sample_packing: true
-```
-
-### SFT with Streaming
-
-For supervised fine-tuning with streaming:
-
-```yaml
-streaming: true
-datasets:
-  - path: tatsu-lab/alpaca
-    type: alpaca
-    split: train
-
-# Optionally, enable sample packing
-streaming_multipack_buffer_size: 10000
-sample_packing: true
-```
-
-## Configuration Options
-
-### `streaming_multipack_buffer_size`
-
-Controls the buffer size for multipack streaming (default: 10,000). This determines how
-many samples are buffered before packing. Larger buffers can improve packing efficiency
-but use more memory.
-
-### `shuffle_merged_datasets`
-
-When enabled, shuffles the streaming dataset using the buffer. This requires additional
-memory for the shuffle buffer.
-
-## Sample Packing with Streaming
-
-Sample packing is supported for streaming datasets. When enabled, multiple samples are
-packed into a single sequence to maximize GPU utilization:
-
-```yaml
-sample_packing: true
-streaming_multipack_buffer_size: 10000
-
-# For SFT: attention is automatically isolated between packed samples
-# For pretraining: control with pretrain_multipack_attn
-pretrain_multipack_attn: true  # prevent cross-attention between packed samples
-```
-
-For more information, see our [documentation](multipack.qmd) on multipacking.
-
-## Important Considerations
-
-### Memory Usage
-
-While streaming reduces memory usage compared to loading entire datasets, you still need
-to consider:
-
- You can control the memory usage by adjusting `streaming_multipack_buffer_size`
- Sample packing requires buffering multiple samples
- Shuffling requires additional memory for the shuffle buffer
-
-### Performance
-
- Streaming may have slightly higher latency compared to preprocessed datasets, as samples are processed on-the-fly
- Network speed and disk read speed are important when streaming from remote sources or a local dataset, respectively
- Consider using `axolotl preprocess` for smaller or more frequently used datasets
-
-### Evaluation Datasets
-
-Evaluation datasets are not streamed to ensure consistent evaluation metrics. They're
-loaded normally even when training uses streaming.
-
-## Examples
-
-See the `examples/streaming/` directory for complete configuration examples:
-
- `pretrain.yaml`: Pretraining with streaming dataset
- `sft.yaml`: Supervised fine-tuning with streaming
--- a/examples/LiquidAI/README.md
+++ b/examples/LiquidAI/README.md
@@ -1,63 +0,0 @@
-# Finetune Liquid Foundation Models 2 (LFM2) with Axolotl
-
-[Liquid Foundation Models 2 (LFM2)](https://huggingface.co/collections/LiquidAI/lfm2-686d721927015b2ad73eaa38) are a family of small, open-weight models from [Liquid AI](https://www.liquid.ai/) focused on quality, speed, and memory efficiency. Liquid AI released text-only [LFM2](https://huggingface.co/collections/LiquidAI/lfm2-686d721927015b2ad73eaa38) and text+vision [LFM2-VL](https://huggingface.co/collections/LiquidAI/lfm2-vl-68963bbc84a610f7638d5ffa) models.
-
-LFM2 features a new hybrid Liquid architecture with multiplicative gates, short-range convolutions, and grouped query attention, enabling fast training and inference.
-
-This guide shows how to fine-tune both the LFM2 and LFM2-VL models with Axolotl.
-
-## Getting Started
-
-1.  Install Axolotl following the [installation guide](https://docs.axolotl.ai/docs/installation.html).
-
-    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
-    ```
-
-2.  Run one of the finetuning examples below.
-
-    **LFM2**
-    ```bash
-    # FFT SFT (1x48GB @ 25GiB)
-    axolotl train examples/LiquidAI/lfm2-350m-fft.yaml
-    ```
-
-    **LFM2-VL**
-    ```bash
-    # LoRA SFT (1x48GB @ 2.7GiB)
-    axolotl train examples/LiquidAI/lfm2-vl-lora.yaml
-    ```
-
-### TIPS
-
- **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
-  ```
-
- **Dataset Loading**: Read more on how to load your own dataset in our [documentation](https://docs.axolotl.ai/docs/dataset_loading.html).
- **Dataset Formats**:
-  - For LFM2 models, the dataset format follows the OpenAI Messages format as seen [here](https://docs.axolotl.ai/docs/dataset-formats/conversation.html#chat_template).
-  - For LFM2-VL models, Axolotl follows the multi-content Messages format. See our [Multimodal docs](https://docs.axolotl.ai/docs/multimodal.html#dataset-format) for details.
-
-## Optimization Guides
-
- [Multi-GPU Training](https://docs.axolotl.ai/docs/multi-gpu.html)
- [LoRA Optimizations](https://docs.axolotl.ai/docs/lora_optims.html)
- [Multi-Node Training](https://docs.axolotl.ai/docs/multi-node.html)
-
-## Related Resources
-
- [LFM2 Blog](https://www.liquid.ai/blog/liquid-foundation-models-v2-our-second-series-of-generative-ai-models)
- [LFM2-VL Blog](https://www.liquid.ai/blog/lfm2-vl-efficient-vision-language-models)
- [Axolotl Docs](https://docs.axolotl.ai)
- [Axolotl GitHub](https://github.com/axolotl-ai-cloud/axolotl)
- [Axolotl Discord](https://discord.gg/7m9sfhzaf3)
--- a/examples/LiquidAI/lfm2-350m-fft.yaml
+++ b/examples/LiquidAI/lfm2-350m-fft.yaml
@@ -1,49 +0,0 @@
-base_model: LiquidAI/LFM2-350M
-
-chunked_cross_entropy: true
-
-eot_tokens:
-  - "<|im_end|>"
-datasets:
-  - path: mlabonne/FineTome-100k
-    type: chat_template
-    split: train[:20%]
-    field_messages: conversations
-    message_field_role: from
-    message_field_content: value
-dataset_prepared_path: last_run_prepared
-val_set_size: 0.05
-output_dir: ./outputs/out
-
-sequence_len: 4096
-sample_packing: true
-
-
-wandb_project:
-wandb_entity:
-wandb_watch:
-wandb_name:
-wandb_log_model:
-
-gradient_accumulation_steps: 2
-micro_batch_size: 4
-num_epochs: 1
-optimizer: adamw_torch_fused
-lr_scheduler: cosine
-learning_rate: 5e-5
-
-bf16: true
-tf32: true
-
-gradient_checkpointing: false
-resume_from_checkpoint:
-logging_steps: 1
-flash_attention: true
-
-warmup_ratio: 0.1
-evals_per_epoch: 2
-saves_per_epoch: 1
-
-weight_decay: 0.0
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/alst/README.md
+++ b/examples/alst/README.md
@@ -1,30 +0,0 @@
-# Arctic Long Sequence Training (ALST)
-
-Artic Long Sequence Training (ALST) is a technique for training long context models using a variety of optimization
-techniques. It is a combination of:
- TiledMLP: Leverage tiling over the sequence dimension on MLP layers to reduce memory usage
- Tiled Loss: Using optimized loss functions like Liger-Kernel or Cut Cross Entropy to reduce memory usage
- Activation Offloading: Offload activations to CPU RAM to reduce memory usage
-
-For more information, you can check out the ALST paper [here](https://www.arxiv.org/abs/2506.13996).
-
-## Usage
-
-```yaml
-tiled_mlp: true
-
-# See Sequence Parallelism docs
-# https://docs.axolotl.ai/docs/sequence_parallelism.html
-context_parallel_size: int
-
-plugins:
-# See Cut Cross Entropy docs
-# https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
-  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
-
-# or Liger Kernel docs
-# https://docs.axolotl.ai/docs/custom_integrations.html#liger-kernels
-  - axolotl.integrations.liger.LigerPlugin
-# ...
-
-```
--- a/examples/alst/llama3-8b-deepspeed-alst.yaml
+++ b/examples/alst/llama3-8b-deepspeed-alst.yaml
@@ -1,53 +0,0 @@
-base_model: meta-llama/Llama-3.1-8B
-# Automatically upload checkpoint and final model to HF
-# hub_model_id: username/custom_model_name
-
-datasets:
-  - path: togethercomputer/Long-Data-Collections
-    type: completion
-    field: text
-    data_files:
-      - pretrain/rp_sub.jsonl.zst
-  - path: princeton-nlp/TextbookChapters
-    type: completion
-    field: chapter
-dataset_prepared_path: last_run_prepared
-val_set_size: 0.0
-output_dir: ./outputs/out
-
-sequence_len: 500_000
-min_sample_len: 200_000
-sample_packing: true
-
-tiled_mlp: true
-context_parallel_size: 8
-plugins:
-  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
-
-gradient_accumulation_steps: 1
-micro_batch_size: 1
-num_epochs: 1
-optimizer: adamw_torch_8bit
-lr_scheduler: cosine
-learning_rate: 2e-5
-
-bf16: auto
-tf32: true
-
-gradient_checkpointing: true
-activation_offloading: legacy
-
-resume_from_checkpoint:
-logging_steps: 1
-flash_attention: true
-
-warmup_steps: 100
-saves_per_epoch: 1
-evals_per_epoch: 2
-weight_decay: 0.0
-special_tokens:
-  pad_token: <|end_of_text|>
-
-deepspeed: deepspeed_configs/zero3_bf16_cpuoffload_all.json
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/alst/llama3-8b-fsdp2-alst.yaml
+++ b/examples/alst/llama3-8b-fsdp2-alst.yaml
@@ -1,59 +0,0 @@
-base_model: meta-llama/Llama-3.1-8B
-# Automatically upload checkpoint and final model to HF
-# hub_model_id: username/custom_model_name
-
-datasets:
-  - path: togethercomputer/Long-Data-Collections
-    type: completion
-    field: text
-    data_files:
-      - pretrain/rp_sub.jsonl.zst
-  - path: princeton-nlp/TextbookChapters
-    type: completion
-    field: chapter
-dataset_prepared_path: last_run_prepared
-val_set_size: 0.0
-output_dir: ./outputs/out
-
-sequence_len: 500_000
-min_sample_len: 200_000
-sample_packing: true
-
-tiled_mlp: true
-context_parallel_size: 8
-plugins:
-  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
-
-gradient_accumulation_steps: 1
-micro_batch_size: 1
-num_epochs: 1
-optimizer: adamw_torch_8bit
-lr_scheduler: cosine
-learning_rate: 2e-5
-
-bf16: auto
-tf32: true
-
-gradient_checkpointing: true
-activation_offloading: legacy
-
-resume_from_checkpoint:
-logging_steps: 1
-flash_attention: true
-
-warmup_steps: 100
-saves_per_epoch: 1
-evals_per_epoch: 2
-weight_decay: 0.0
-special_tokens:
-  pad_token: <|end_of_text|>
-
-fsdp_version: 2
-fsdp_config:
-  offload_params: false  # offloading is currently not compatible with SP + torchao optimizer
-  state_dict_type: SHARDED_STATE_DICT
-  auto_wrap_policy: TRANSFORMER_BASED_WRAP
-  transformer_layer_cls_to_wrap: LlamaDecoderLayer
-  reshard_after_forward: true
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/apertus/README.md
+++ b/examples/apertus/README.md
@@ -1,110 +0,0 @@
-# Finetune Swiss-AI's Apertus with Axolotl
-
-[Apertus](https://huggingface.co/collections/swiss-ai/apertus-llm-68b699e65415c231ace3b059) is a family of opensource models trained by Swiss-ai.
-
-This guide shows how to fine-tune it with Axolotl with multi-turn conversations and proper masking.
-
-## Getting started
-
-1. Install Axolotl following the [installation guide](https://docs.axolotl.ai/docs/installation.html). You need to install from main as Apertus is only on nightly or use our latest [Docker images](https://docs.axolotl.ai/docs/docker.html).
-
-    Here is an example of how to install from main for pip:
-
-```bash
-# Ensure you have Pytorch installed (Pytorch 2.6.0 min)
-git clone https://github.com/axolotl-ai-cloud/axolotl.git
-cd axolotl
-
-uv sync
-uv pip install flash-attn --no-build-isolation
-
-# Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
-python scripts/cutcrossentropy_install.py | sh
-```
-
-2. (Optional, highly recommended) Install XIELU CUDA
-
-```bash
-## Recommended for reduced VRAM and faster speeds
-
-# Point to CUDA toolkit directory
-# 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
-```
-
-For any installation errors, see [XIELU Installation Issues](#xielu-installation-issues)
-
-3. Run the finetuning example:
-
-```bash
-axolotl train examples/apertus/apertus-8b-qlora.yaml
-```
-
-This config uses about 8.7 GiB VRAM.
-
-Let us know how it goes. Happy finetuning! 🚀
-
-### Tips
-
- For inference, the official Apertus team recommends `top_p=0.9` and `temperature=0.8`.
- You can instead use full paremter fine-tuning by removing the `adapter: qlora` and `load_in_4bit: true` from the config.
- Read more on how to load your own dataset at [docs](https://docs.axolotl.ai/docs/dataset_loading.html).
- The dataset format follows the OpenAI Messages format as seen [here](https://docs.axolotl.ai/docs/dataset-formats/conversation.html#chat_template).
-
-### XIELU Installation Issues
-
-#### `ModuleNotFoundError: No module named 'torch'`
-
-Please check these one by one:
- Running in correct environment
- Env has PyTorch installed
- CUDA toolkit is at `CUDA_HOME`
-
-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
-    ```
-
-2. Git clone the repo and manually hardcode python path:
-
-    ```bash
-    git clone https://github.com/nickjbrowning/XIELU
-    cd xielu
-    git checkout 59d6031
-
-    cd xielu
-    nano CMakeLists.txt  # or vi depending on your preference
-    ```
-
-    ```diff
-    execute_process(
-    -    COMMAND ${Python_EXECUTABLE} -c "import torch.utils; print(torch.utils.cmake_prefix_path)"
-    +    COMMAND /root/miniconda3/envs/py3.11/bin/python -c "import torch.utils; print(torch.utils.cmake_prefix_path)"
-        RESULT_VARIABLE TORCH_CMAKE_PATH_RESULT
-        OUTPUT_VARIABLE TORCH_CMAKE_PATH_OUTPUT
-        ERROR_VARIABLE TORCH_CMAKE_PATH_ERROR
-    )
-    ```
-
-    ```bash
-    uv pip install . --no-build-isolation --no-deps
-    ```
-
-## Optimization Guides
-
- [Multi-GPU Training](https://docs.axolotl.ai/docs/multi-gpu.html)
- [Multi-Node Training](https://docs.axolotl.ai/docs/multi-node.html)
- [LoRA Optimizations](https://docs.axolotl.ai/docs/lora_optims.html)
-
-## Related Resources
-
- [Apertus Tech Report](https://github.com/swiss-ai/apertus-tech-report/blob/main/Apertus_Tech_Report.pdf)
- [Axolotl Docs](https://docs.axolotl.ai)
- [Axolotl Website](https://axolotl.ai)
- [Axolotl GitHub](https://github.com/axolotl-ai-cloud/axolotl)
- [Axolotl Discord](https://discord.gg/7m9sfhzaf3)
--- a/examples/apertus/apertus-8b-qlora.yaml
+++ b/examples/apertus/apertus-8b-qlora.yaml
@@ -1,64 +0,0 @@
-base_model: swiss-ai/Apertus-8B-Instruct-2509
-
-# Automatically upload checkpoint and final model to HF
-# hub_model_id: username/custom_model_name
-
-plugins:
-  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
-
-load_in_8bit: false
-load_in_4bit: true
-
-datasets:
-  - path: fozziethebeat/alpaca_messages_2k_test
-    type: chat_template
-
-dataset_prepared_path: last_run_prepared
-val_set_size: 0.1
-output_dir: ./outputs/lora-out
-
-adapter: qlora
-lora_model_dir:
-
-sequence_len: 2048
-sample_packing: true
-
-lora_r: 32
-lora_alpha: 16
-lora_dropout: 0.05
-lora_target_linear: true
-lora_target_modules:
-  - gate_proj
-  - down_proj
-  - up_proj
-  - q_proj
-  - v_proj
-  - k_proj
-  - o_proj
-
-wandb_project:
-wandb_entity:
-wandb_watch:
-wandb_name:
-wandb_log_model:
-
-gradient_accumulation_steps: 4
-micro_batch_size: 2
-num_epochs: 1
-optimizer: adamw_bnb_8bit
-lr_scheduler: cosine
-learning_rate: 0.0002
-
-bf16: auto
-tf32: false
-
-gradient_checkpointing: true
-resume_from_checkpoint:
-logging_steps: 1
-flash_attention: true
-
-warmup_ratio: 0.1
-evals_per_epoch: 1
-saves_per_epoch: 1
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/arcee/README.md
+++ b/examples/arcee/README.md
@@ -1,56 +0,0 @@
-# Finetune ArceeAI's AFM with Axolotl
-
-[Arcee Foundation Models (AFM)](https://huggingface.co/collections/arcee-ai/afm-45b-68823397c351603014963473) are a family of 4.5B parameter open weight models trained by Arcee.ai.
-
-This guide shows how to fine-tune it with Axolotl with multi-turn conversations and proper masking.
-
-Thanks to the team at Arcee.ai for using Axolotl in supervised fine-tuning the AFM model.
-
-## Getting started
-
-1. Install Axolotl following the [installation guide](https://docs.axolotl.ai/docs/installation.html). You need to install from main as AFM is only on nightly or use our latest [Docker images](https://docs.axolotl.ai/docs/docker.html).
-
-    Here is an example of how to install from main for pip:
-
-```bash
-# Ensure you have Pytorch installed (Pytorch 2.6.0 min)
-git clone https://github.com/axolotl-ai-cloud/axolotl.git
-cd axolotl
-
-uv sync
-uv pip install flash-attn --no-build-isolation
-
-# Install CCE https://docs.axolotl.ai/docs/custom_integrations.html#cut-cross-entropy
-python scripts/cutcrossentropy_install.py | sh
-```
-
-2. Run the finetuning example:
-
-```bash
-axolotl train examples/arcee/afm-4.5b-qlora.yaml
-```
-
-This config uses about 7.8GiB VRAM.
-
-Let us know how it goes. Happy finetuning! 🚀
-
-### TIPS
-
- For inference, the official Arcee.ai team recommends `top_p: 0.95`, `temperature: 0.5`, `top_k: 50`, and `repeat_penalty: 1.1`.
- You can run a full finetuning by removing the `adapter: qlora` and `load_in_4bit: true` from the config.
- Read more on how to load your own dataset at [docs](https://docs.axolotl.ai/docs/dataset_loading.html).
- The dataset format follows the OpenAI Messages format as seen [here](https://docs.axolotl.ai/docs/dataset-formats/conversation.html#chat_template).
-
-## Optimization Guides
-
- [Multi-GPU Training](https://docs.axolotl.ai/docs/multi-gpu.html)
- [Multi-Node Training](https://docs.axolotl.ai/docs/multi-node.html)
- [LoRA Optimizations](https://docs.axolotl.ai/docs/lora_optims.html)
-
-## Related Resources
-
- [AFM Blog](https://docs.arcee.ai/arcee-foundation-models/introduction-to-arcee-foundation-models)
- [Axolotl Docs](https://docs.axolotl.ai)
- [Axolotl Website](https://axolotl.ai)
- [Axolotl GitHub](https://github.com/axolotl-ai-cloud/axolotl)
- [Axolotl Discord](https://discord.gg/7m9sfhzaf3)
--- a/examples/arcee/afm-4.5b-qlora.yaml
+++ b/examples/arcee/afm-4.5b-qlora.yaml
@@ -1,64 +0,0 @@
-base_model: arcee-ai/AFM-4.5B
-
-# Automatically upload checkpoint and final model to HF
-# hub_model_id: username/custom_model_name
-
-plugins:
-  - axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
-
-load_in_8bit: false
-load_in_4bit: true
-
-datasets:
-  - path: fozziethebeat/alpaca_messages_2k_test
-    type: chat_template
-
-dataset_prepared_path: last_run_prepared
-val_set_size: 0.1
-output_dir: ./outputs/lora-out
-
-adapter: qlora
-lora_model_dir:
-
-sequence_len: 2048
-sample_packing: true
-
-lora_r: 32
-lora_alpha: 16
-lora_dropout: 0.05
-lora_target_linear: true
-lora_target_modules:
-  - gate_proj
-  - down_proj
-  - up_proj
-  - q_proj
-  - v_proj
-  - k_proj
-  - o_proj
-
-wandb_project:
-wandb_entity:
-wandb_watch:
-wandb_name:
-wandb_log_model:
-
-gradient_accumulation_steps: 4
-micro_batch_size: 2
-num_epochs: 1
-optimizer: adamw_bnb_8bit
-lr_scheduler: cosine
-learning_rate: 0.0002
-
-bf16: auto
-tf32: false
-
-gradient_checkpointing: true
-resume_from_checkpoint:
-logging_steps: 1
-flash_attention: true
-
-warmup_ratio: 0.1
-evals_per_epoch: 1
-saves_per_epoch: 1
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/archived/README.md
+++ b/examples/archived/README.md
@@ -1,5 +0,0 @@
-# Archived Examples
-
-This directory contains examples that are no longer maintained and may no longer be functional.
-
-We keep them around for archival purposes in case they are useful to others.
--- a/examples/archived/cerebras/btlm-ft.yml
+++ b/examples/archived/cerebras/btlm-ft.yml
@@ -66,7 +66,7 @@ flash_optimum:
 gptq_groupsize:
 gptq_model_v1:

-warmup_ratio: 0.1
+warmup_steps: 32
 evals_per_epoch: 4
 saves_per_epoch: 1
 save_total_limit:
--- a/examples/archived/cerebras/qlora.yml
+++ b/examples/archived/cerebras/qlora.yml
@@ -43,7 +43,7 @@ xformers_attention: true
 flash_attention:
 gptq_groupsize:
 gptq_model_v1:
-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.1
--- a/examples/cloud/baseten.yaml
+++ b/examples/cloud/baseten.yaml
@@ -1,10 +0,0 @@
-provider: baseten
-project_name:
-
-secrets:
-  - HF_TOKEN
-  - WANDB_API_KEY
-
-gpu: h100
-gpu_count: 8
-node_count: 1
--- a/examples/archived/code-llama/13b/lora.yml
+++ b/examples/archived/code-llama/13b/lora.yml
@@ -17,7 +17,7 @@ output_dir: ./outputs/lora-out

 sequence_len: 4096
 sample_packing: true
-
+pad_to_sequence_len: true

 adapter: lora
 lora_model_dir:
@@ -47,7 +47,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/archived/code-llama/13b/qlora.yml
+++ b/examples/archived/code-llama/13b/qlora.yml
@@ -20,7 +20,7 @@ lora_model_dir:

 sequence_len: 4096
 sample_packing: true
-
+pad_to_sequence_len: true

 lora_r: 32
 lora_alpha: 16
@@ -48,7 +48,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/archived/code-llama/34b/lora.yml
+++ b/examples/archived/code-llama/34b/lora.yml
@@ -17,7 +17,7 @@ output_dir: ./outputs/lora-out

 sequence_len: 4096
 sample_packing: true
-
+pad_to_sequence_len: true

 adapter: lora
 lora_model_dir:
@@ -47,7 +47,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/archived/code-llama/34b/qlora.yml
+++ b/examples/archived/code-llama/34b/qlora.yml
@@ -20,7 +20,7 @@ lora_model_dir:

 sequence_len: 4096
 sample_packing: true
-
+pad_to_sequence_len: true

 lora_r: 32
 lora_alpha: 16
@@ -48,7 +48,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/archived/code-llama/7b/lora.yml
+++ b/examples/archived/code-llama/7b/lora.yml
@@ -17,7 +17,7 @@ output_dir: ./outputs/lora-out

 sequence_len: 4096
 sample_packing: true
-
+pad_to_sequence_len: true

 adapter: lora
 lora_model_dir:
@@ -47,7 +47,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/archived/code-llama/7b/qlora.yml
+++ b/examples/archived/code-llama/7b/qlora.yml
@@ -20,7 +20,7 @@ lora_model_dir:

 sequence_len: 4096
 sample_packing: true
-
+pad_to_sequence_len: true

 lora_r: 32
 lora_alpha: 16
@@ -48,7 +48,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 4
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/archived/code-llama/README.md
+++ b/examples/archived/code-llama/README.md
--- a/examples/cohere/command-r-7b-qlora.yml
+++ b/examples/cohere/command-r-7b-qlora.yml
@@ -27,7 +27,7 @@ lora_target_linear: true
 sequence_len: 2048
 sample_packing: true
 eval_sample_packing: false
-
+pad_to_sequence_len: true

 wandb_project:
 wandb_entity:
@@ -35,6 +35,7 @@ wandb_watch:
 wandb_name:
 wandb_log_model:

+
 gradient_accumulation_steps: 4
 micro_batch_size: 1
 num_epochs: 4
@@ -55,5 +56,3 @@ evals_per_epoch:
 saves_per_epoch: 1
 weight_decay: 0.0
 special_tokens:
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/examples/colab-notebooks/colab-axolotl-example.ipynb
+++ b/examples/colab-notebooks/colab-axolotl-example.ipynb
--- a/examples/archived/dbrx/16bit-lora.yaml
+++ b/examples/archived/dbrx/16bit-lora.yaml
@@ -54,7 +54,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch:
 saves_per_epoch: 1

--- a/examples/archived/dbrx/8bit-lora.yaml
+++ b/examples/archived/dbrx/8bit-lora.yaml
@@ -57,7 +57,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch:
 saves_per_epoch: 1

--- a/examples/archived/dbrx/README.md
+++ b/examples/archived/dbrx/README.md
--- a/examples/archived/dbrx/fft-ds-zero3.yaml
+++ b/examples/archived/dbrx/fft-ds-zero3.yaml
@@ -41,7 +41,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch:
 saves_per_epoch: 1

--- a/examples/archived/deepcoder/deepcoder-14B-preview-lora.yml
+++ b/examples/archived/deepcoder/deepcoder-14B-preview-lora.yml
@@ -9,6 +9,10 @@ strict: false
 datasets:
  - path: fozziethebeat/alpaca_messages_2k_test
    type: chat_template
+    field_messages: messages
+    message_property_mappings:
+      role: role
+      content: content

 dataset_prepared_path:
 val_set_size: 0.05
@@ -17,7 +21,7 @@ output_dir: ./outputs/lora-out
 sequence_len: 4096
 sample_packing: true
 eval_sample_packing: false
-
+pad_to_sequence_len: true

 adapter: lora
 lora_model_dir:
@@ -47,7 +51,7 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 1
 saves_per_epoch: 1
 weight_decay: 0.0
--- a/examples/deepcogito/cogito-v1-preview-llama-3B-lora.yml
+++ b/examples/deepcogito/cogito-v1-preview-llama-3B-lora.yml
@@ -9,6 +9,10 @@ strict: false
 datasets:
  - path: fozziethebeat/alpaca_messages_2k_test
    type: chat_template
+    field_messages: messages
+    message_property_mappings:
+      role: role
+      content: content

 dataset_prepared_path:
 val_set_size: 0.05
@@ -17,7 +21,7 @@ output_dir: ./outputs/lora-out
 sequence_len: 4096
 sample_packing: true
 eval_sample_packing: false
-
+pad_to_sequence_len: true

 adapter: lora
 lora_model_dir:
@@ -47,10 +51,8 @@ resume_from_checkpoint:
 logging_steps: 1
 flash_attention: true

-warmup_ratio: 0.1
+warmup_steps: 10
 evals_per_epoch: 1
 saves_per_epoch: 1
 weight_decay: 0.0
 special_tokens:
-
-# save_first_step: true  # uncomment this to validate checkpoint saving works with your config
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
NanoCode012	1f2f285173	fix: missing key in enum	2025-07-03 13:46:16 +08:00
NanoCode012	98e912e416	feat: add custom processing strategy for phi35 vl	2025-07-03 13:46:16 +08:00
NanoCode012	e1528fb381	feat: add phi_35_vl support	2025-07-03 13:46:16 +08:00