├── .github
    ├── dependabot.yml
    └── workflows
    │   ├── build-wheel-rocm.yml
    │   ├── build-wheels-basic.yml
    │   ├── build-wheels-batch-basic.yml
    │   ├── build-wheels-batch-cpu.yml
    │   ├── build-wheels-batch-ggml.yml
    │   ├── build-wheels-batch-macos.yml
    │   ├── build-wheels-batch-oobabooga-basic.yml
    │   ├── build-wheels-batch-oobabooga.yml
    │   ├── build-wheels-batch-rocm.yml
    │   ├── build-wheels-batch.yml
    │   ├── build-wheels-cpu.yml
    │   ├── build-wheels-full-release.yml
    │   ├── build-wheels-ggml-cpu.yml
    │   ├── build-wheels-ggml-oobabooga.yml
    │   ├── build-wheels-ggml.yml
    │   ├── build-wheels-macos.yml
    │   ├── build-wheels-oobabooga-basic.yml
    │   ├── build-wheels-oobabooga-rocm.yml
    │   ├── build-wheels-oobabooga.yml
    │   ├── build-wheels-prioritized-release.yml
    │   ├── build-wheels-rocm-full.yml
    │   ├── build-wheels-test.yml
    │   ├── build-wheels.yml
    │   └── deploy-index.yml
├── LICENSE
├── README.md
├── generate-html.ps1
├── generate-textgen-html.ps1
├── host-index.bat
├── host-index.sh
├── index
    ├── AVX
    │   ├── cpu
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu116
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu117
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu118
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu120
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu121
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu122
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── index.html
    │   ├── rocm5.4.2
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── rocm5.5.1
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── rocm5.5
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   └── rocm5.6.1
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │       └── index.html
    ├── AVX2
    │   ├── cpu
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu116
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu117
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu118
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu120
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu121
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu122
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── index.html
    │   ├── rocm5.4.2
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── rocm5.5.1
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── rocm5.5
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   └── rocm5.6.1
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │       └── index.html
    ├── AVX512
    │   ├── cpu
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu116
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu117
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu118
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu120
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu121
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu122
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   └── index.html
    ├── basic
    │   ├── cpu
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu116
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu117
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu118
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu120
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu121
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── cu122
    │   │   ├── index.html
    │   │   ├── llama-cpp-python-ggml
    │   │   │   └── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── index.html
    │   ├── rocm5.4.2
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── rocm5.5.1
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   ├── rocm5.5
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │   │   └── index.html
    │   └── rocm5.6.1
    │   │   ├── index.html
    │   │   └── llama-cpp-python
    │   │       └── index.html
    ├── cpu.html
    ├── cu116.html
    ├── cu117.html
    ├── cu118.html
    ├── cu120.html
    ├── cu121.html
    ├── cu122.html
    ├── index.html
    ├── rocm5.4.2.html
    ├── rocm5.5.1.html
    ├── rocm5.5.html
    ├── rocm5.6.1.html
    └── textgen
    │   ├── AVX
    │       ├── cu117
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu118
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu120
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu121
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu122
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── index.html
    │       ├── rocm5.4.2
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       ├── rocm5.5.1
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       ├── rocm5.5
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       └── rocm5.6.1
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │       └── index.html
    │   ├── AVX2
    │       ├── cu117
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu118
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu120
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu121
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu122
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── index.html
    │       ├── rocm5.4.2
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       ├── rocm5.5.1
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       ├── rocm5.5
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       └── rocm5.6.1
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │       └── index.html
    │   ├── basic
    │       ├── cu117
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu118
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu120
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu121
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── cu122
    │       │   ├── index.html
    │       │   ├── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       │   └── llama-cpp-python-ggml-cuda
    │       │   │   └── index.html
    │       ├── index.html
    │       ├── rocm5.4.2
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       ├── rocm5.5.1
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       ├── rocm5.5
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │   │   └── index.html
    │       └── rocm5.6.1
    │       │   ├── index.html
    │       │   └── llama-cpp-python-cuda
    │       │       └── index.html
    │   └── index.html
├── old_workflows
    ├── build-all-wheels.yml
    ├── build-wheel-rocm-windows.yml
    ├── build-wheel-rocm.yml
    ├── build-wheels-0.1.62.yml
    ├── build-wheels-0.1.66.yml
    ├── build-wheels-0.1.67.yml
    ├── build-wheels-0.1.68.yml
    ├── build-wheels-0.1.69.yml
    ├── build-wheels-0.1.70.yml
    ├── build-wheels-0.1.71.yml
    ├── build-wheels-0.1.72.yml
    ├── build-wheels-0.1.73.yml
    ├── build-wheels-0.1.74.yml
    ├── build-wheels-0.1.76.yml
    ├── build-wheels-0.1.77.yml
    ├── build-wheels-avx.yml
    ├── build-wheels-oobabooga.yml
    ├── build-wheels-prioritized-release.yml
    ├── build-wheels-rocm-full.yml
    └── build-wheels.yml
└── workflows.md


/.github/dependabot.yml:
--------------------------------------------------------------------------------
1 | version: 2
2 | updates:
3 |   - package-ecosystem: "github-actions"
4 |     directory: "/"
5 |     schedule:
6 |       interval: "daily"
7 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheel-rocm.yml:
--------------------------------------------------------------------------------
 1 | name: Build ROCm Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       version:
 7 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
 8 |         default: 'v0.2.9'
 9 |         required: true
10 |         type: string
11 |   workflow_call:
12 |     inputs:
13 |       version:
14 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
15 |         default: 'v0.2.9'
16 |         required: true
17 |         type: string
18 | 
19 | permissions:
20 |   contents: write
21 | 
22 | jobs:
23 |   build_wheels_rocm:
24 |     name: ROCm Wheels
25 |     uses: ./.github/workflows/build-wheels-rocm-full.yml
26 |     with:
27 |       version: ${{ inputs.version }}
28 |       config: 'rename:0'
29 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-basic.yml:
--------------------------------------------------------------------------------
 1 | name: Build Basic Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       version:
 7 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
 8 |         default: 'v0.2.9'
 9 |         required: true
10 |         type: string
11 |   workflow_call:
12 |     inputs:
13 |       version:
14 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
15 |         default: 'v0.2.9'
16 |         required: true
17 |         type: string
18 | 
19 | permissions:
20 |   contents: write
21 | 
22 | jobs:
23 |   build_wheels_main:
24 |     name: Main Wheels
25 |     uses: ./.github/workflows/build-wheels.yml
26 |     with:
27 |       version: ${{ inputs.version }}
28 |       cpu: '0'
29 |       config: 'releasetag:basic'
30 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-basic.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build Basic Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.1.77,v0.1.76'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   define_matrix:
17 |     name: Define Workflow Matrix
18 |     runs-on: ubuntu-latest
19 |     outputs:
20 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
21 |     defaults:
22 |       run:
23 |         shell: pwsh
24 |     env:
25 |       PCKGVERS: ${{ inputs.versions }}
26 |         
27 |     steps:
28 |       - uses: actions/checkout@v4
29 |       
30 |       - name: Define Job Output
31 |         id: set-matrix
32 |         run: |
33 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
34 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
35 |           
36 |   run_workflows:
37 |     name: Build ${{ matrix.version }} Basic CUDA Wheels
38 |     needs: define_matrix
39 |     strategy:
40 |       max-parallel: 1
41 |       matrix:
42 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
43 |     uses: ./.github/workflows/build-wheels-basic.yml
44 |     with:
45 |       version: ${{ matrix.version }}
46 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-cpu.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build CPU-only Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.1.77,v0.1.76'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   define_matrix:
17 |     name: Define Workflow Matrix
18 |     runs-on: ubuntu-latest
19 |     outputs:
20 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
21 |     defaults:
22 |       run:
23 |         shell: pwsh
24 |     env:
25 |       PCKGVERS: ${{ inputs.versions }}
26 |         
27 |     steps:
28 |       - uses: actions/checkout@v4
29 |       
30 |       - name: Define Job Output
31 |         id: set-matrix
32 |         run: |
33 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
34 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
35 |           
36 |   run_workflows:
37 |     name: Build ${{ matrix.version }} CPU-only Wheels
38 |     needs: define_matrix
39 |     strategy:
40 |       max-parallel: 1
41 |       matrix:
42 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
43 |     uses: ./.github/workflows/build-wheels-cpu.yml
44 |     with:
45 |       version: ${{ matrix.version }}
46 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-ggml.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build GGML Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python-ggml to build'
 8 |         default: 'v0.1.78,v0.1.77'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   define_matrix:
17 |     name: Define Workflow Matrix
18 |     runs-on: ubuntu-latest
19 |     outputs:
20 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
21 |     defaults:
22 |       run:
23 |         shell: pwsh
24 |     env:
25 |       PCKGVERS: ${{ inputs.versions }}
26 |         
27 |     steps:
28 |       - uses: actions/checkout@v4
29 |       
30 |       - name: Define Job Output
31 |         id: set-matrix
32 |         run: |
33 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
34 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
35 |           
36 |   run_workflows:
37 |     name: Build ${{ matrix.version }} Wheels
38 |     needs: define_matrix
39 |     strategy:
40 |       max-parallel: 1
41 |       matrix:
42 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
43 |     uses: ./.github/workflows/build-wheels.yml
44 |     with:
45 |       version: ${{ matrix.version }}
46 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-macos.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build MacOS Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.2.6,v0.2.5,v0.2.4,v0.2.3,v0.2.2,v0.2.1,v0.2.0,v0.1.85'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   define_matrix:
17 |     name: Define Workflow Matrix
18 |     runs-on: ubuntu-latest
19 |     outputs:
20 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
21 |     defaults:
22 |       run:
23 |         shell: pwsh
24 |     env:
25 |       PCKGVERS: ${{ inputs.versions }}
26 |         
27 |     steps:
28 |       - uses: actions/checkout@v4
29 |       
30 |       - name: Define Job Output
31 |         id: set-matrix
32 |         run: |
33 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
34 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
35 |           
36 |   run_workflows:
37 |     name: Build ${{ matrix.version }} MacOS Metal Wheels
38 |     needs: define_matrix
39 |     strategy:
40 |       matrix:
41 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
42 |     uses: ./.github/workflows/build-wheels-macos.yml
43 |     with:
44 |       version: ${{ matrix.version }}
45 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-oobabooga-basic.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build Basic Wheels for Text Generation Webui
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.1.77,v0.1.76'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   define_matrix:
17 |     name: Define Workflow Matrix
18 |     runs-on: ubuntu-latest
19 |     outputs:
20 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
21 |     defaults:
22 |       run:
23 |         shell: pwsh
24 |     env:
25 |       PCKGVERS: ${{ inputs.versions }}
26 |         
27 |     steps:
28 |       - uses: actions/checkout@v4
29 |       
30 |       - name: Define Job Output
31 |         id: set-matrix
32 |         run: |
33 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
34 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
35 |           
36 |   run_workflows:
37 |     name: Build ${{ matrix.version }} Wheels
38 |     needs: define_matrix
39 |     strategy:
40 |       max-parallel: 1
41 |       matrix:
42 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
43 |     uses: ./.github/workflows/build-wheels-oobabooga-basic.yml
44 |     with:
45 |       version: ${{ matrix.version }}
46 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-oobabooga.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build Wheels for Text Generation Webui
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.2.18,v0.2.17,v0.2.16,v0.2.15,v0.2.14'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   define_matrix:
17 |     name: Define Workflow Matrix
18 |     runs-on: ubuntu-latest
19 |     outputs:
20 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
21 |     defaults:
22 |       run:
23 |         shell: pwsh
24 |     env:
25 |       PCKGVERS: ${{ inputs.versions }}
26 |         
27 |     steps:
28 |       - uses: actions/checkout@v4
29 |       
30 |       - name: Define Job Output
31 |         id: set-matrix
32 |         run: |
33 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
34 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
35 |           
36 |   run_workflows:
37 |     name: Build ${{ matrix.version }} Wheels
38 |     needs: define_matrix
39 |     strategy:
40 |       max-parallel: 1
41 |       matrix:
42 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
43 |     uses: ./.github/workflows/build-wheels-oobabooga.yml
44 |     with:
45 |       version: ${{ matrix.version }}
46 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch-rocm.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build ROCm Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.2.18,v0.2.17,v0.2.16,v0.2.15,v0.2.14'
 9 |         required: true
10 |         type: string
11 |       config:
12 |         description: 'Override configurations to build: key1:item1-1,item1-2;key2:item2-1,item2-2'
13 |         default: 'Default'
14 |         required: false
15 |         type: string
16 |       exclude:
17 |         description: 'Exclude build configurations: key1-1:item1-1,key1-2:item1-2;key2-1:item2-1,key2-2:item2-2'
18 |         default: 'Default'
19 |         required: false
20 |         type: string
21 | 
22 | permissions:
23 |   contents: write
24 |   
25 | jobs:
26 |   define_matrix:
27 |     name: Define Workflow Matrix
28 |     runs-on: ubuntu-latest
29 |     outputs:
30 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
31 |     defaults:
32 |       run:
33 |         shell: pwsh
34 |     env:
35 |       PCKGVERS: ${{ inputs.versions }}
36 |         
37 |     steps:
38 |       - uses: actions/checkout@v4
39 |       
40 |       - name: Define Job Output
41 |         id: set-matrix
42 |         run: |
43 |           $versions = $env:PCKGVERS.Split(',').Trim()
44 |           $versions.foreach({if ([version]$_.TrimStart('v') -lt [version]'0.1.80') {Throw "$_ does not support ROCm!"}})
45 |           $x = ConvertTo-Json $versions -Compress
46 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
47 |           
48 |   run_workflows:
49 |     name: Build ${{ matrix.version }} Wheels
50 |     needs: define_matrix
51 |     strategy:
52 |       max-parallel: 1
53 |       matrix:
54 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
55 |     uses: ./.github/workflows/build-wheels-rocm-full.yml
56 |     with:
57 |       version: ${{ matrix.version }}
58 |       config: ${{ inputs.config }}
59 |       exclude: ${{ inputs.exclude }}
60 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-batch.yml:
--------------------------------------------------------------------------------
 1 | name: Batch Build Wheels
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       versions:
 7 |         description: 'Comma-seperated version tags of llama-cpp-python to build'
 8 |         default: 'v0.1.77,v0.1.76'
 9 |         required: true
10 |         type: string
11 |       cpu:
12 |         description: 'Build CPU-only wheels as well? 0/1'
13 |         default: '1'
14 |         required: false
15 |         type: string
16 | 
17 | permissions:
18 |   contents: write
19 |   
20 | jobs:
21 |   define_matrix:
22 |     name: Define Workflow Matrix
23 |     runs-on: ubuntu-latest
24 |     outputs:
25 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
26 |     defaults:
27 |       run:
28 |         shell: pwsh
29 |     env:
30 |       PCKGVERS: ${{ inputs.versions }}
31 |         
32 |     steps:
33 |       - uses: actions/checkout@v4
34 |       
35 |       - name: Define Job Output
36 |         id: set-matrix
37 |         run: |
38 |           $x = ConvertTo-Json $env:PCKGVERS.Split(',').Trim() -Compress
39 |           Write-Output ('matrix=' + $x) >> $env:GITHUB_OUTPUT
40 |           
41 |   run_workflows:
42 |     name: Build ${{ matrix.version }} Wheels
43 |     needs: define_matrix
44 |     strategy:
45 |       max-parallel: 1
46 |       matrix:
47 |         version: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
48 |     uses: ./.github/workflows/build-wheels.yml
49 |     with:
50 |       version: ${{ matrix.version }}
51 |       cpu: ${{ inputs.cpu }}
52 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-cpu.yml:
--------------------------------------------------------------------------------
  1 | name: Build CPU-only Wheels
  2 | 
  3 | on:
  4 |   workflow_dispatch:
  5 |     inputs:
  6 |       version:
  7 |         description: 'Version tag of llama-cpp-python to build: v0.2.14'
  8 |         default: 'v0.2.14'
  9 |         required: true
 10 |         type: string
 11 |       config:
 12 |         description: 'Override configurations to build: key1:item1-1,item1-2;key2:item2-1,item2-2'
 13 |         default: 'Default'
 14 |         required: false
 15 |         type: string
 16 |       exclude:
 17 |         description: 'Exclude build configurations: key1-1:item1-1,key1-2:item1-2;key2-1:item2-1,key2-2:item2-2'
 18 |         default: 'None'
 19 |         required: false
 20 |         type: string
 21 |   workflow_call:
 22 |     inputs:
 23 |       version:
 24 |         description: 'Version tag of llama-cpp-python to build: v0.2.14'
 25 |         default: 'v0.2.14'
 26 |         required: true
 27 |         type: string
 28 |       config:
 29 |         description: 'Configurations to build: key1:item1-1,item1-2;key2:item2-1,item2-2'
 30 |         default: 'Default'
 31 |         required: false
 32 |         type: string
 33 |       exclude:
 34 |         description: 'Exclude build configurations: key1-1:item1-1,key1-2:item1-2;key2-1:item2-1,key2-2:item2-2'
 35 |         default: 'None'
 36 |         required: false
 37 |         type: string
 38 | 
 39 | permissions:
 40 |   contents: write
 41 | 
 42 | jobs:
 43 |   define_matrix:
 44 |     name: Define Build Matrix
 45 |     runs-on: ubuntu-latest
 46 |     outputs:
 47 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
 48 |     defaults:
 49 |       run:
 50 |         shell: pwsh
 51 |     env:
 52 |       CONFIGIN: ${{ inputs.config }}
 53 |       EXCLUDEIN: ${{ inputs.exclude }}
 54 |         
 55 |     steps:
 56 |       - name: Define Job Output
 57 |         id: set-matrix
 58 |         run: |
 59 |           $matrix = @{
 60 |               'os' = 'ubuntu-20.04', 'windows-latest'
 61 |               'pyver' = "3.10", "3.8", "3.9", "3.11"
 62 |               'avx' = "AVX", "AVX2", "AVX512", "basic"
 63 |           }
 64 |           
 65 |           if ($env:CONFIGIN -ne 'Default') {$env:CONFIGIN.split(';').foreach({$matrix[$_.split(':')[0]] = $_.split(':')[1].split(',')})}
 66 |           
 67 |           if ($env:EXCLUDEIN -ne 'None') {
 68 |               $exclusions = @()
 69 |               $exclusions += $env:EXCLUDEIN.split(';').replace(':','=').replace(',',"`n") | ConvertFrom-StringData
 70 |               $matrix['exclude'] = $exclusions
 71 |           }
 72 |           
 73 |           $matrixOut = ConvertTo-Json $matrix -Compress
 74 |           Write-Output ('matrix=' + $matrixOut) >> $env:GITHUB_OUTPUT
 75 |           
 76 |   build_wheels:
 77 |     name: ${{ matrix.os }} ${{ matrix.pyver }} CPU ${{ matrix.avx }}
 78 |     needs: define_matrix
 79 |     runs-on: ${{ matrix.os }}
 80 |     strategy:
 81 |       matrix: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
 82 |     defaults:
 83 |       run:
 84 |         shell: pwsh
 85 |     env:
 86 |       AVXVER: ${{ matrix.avx }}
 87 |       PCKGVER: ${{ inputs.version }}
 88 | 
 89 |     steps:
 90 |       - uses: actions/checkout@v4
 91 |         with:
 92 |           repository: 'abetlen/llama-cpp-python'
 93 |           ref: ${{ inputs.version }}
 94 |           submodules: 'recursive'
 95 |           
 96 |       - uses: actions/setup-python@v4
 97 |         with:
 98 |           python-version: ${{ matrix.pyver }}
 99 |           
100 |       - name: Install Dependencies
101 |         run: |
102 |           python -m pip install build wheel
103 |           
104 |       - name: Build Wheel
105 |         run: |
106 |           $packageVersion = [version]$env:PCKGVER.TrimStart('v')
107 |           $env:VERBOSE = '1'
108 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = '-DLLAMA_AVX2=off -DLLAMA_FMA=off -DLLAMA_F16C=off'}
109 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = '-DLLAMA_AVX512=on'}
110 |           if ($env:AVXVER -eq 'basic') {$env:CMAKE_ARGS = '-DLLAMA_AVX=off -DLLAMA_AVX2=off -DLLAMA_FMA=off -DLLAMA_F16C=off'}
111 |           if ($packageVersion -gt [version]'0.2.13') {$env:CMAKE_ARGS = "-DLLAMA_NATIVE=off $env:CMAKE_ARGS"}
112 |           $buildtag = "+cpu$env:AVXVER"
113 |           if ($packageVersion -lt [version]'0.2.0') {
114 |             $env:FORCE_CMAKE = '1'
115 |             python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=$buildtag"
116 |           } else {
117 |             $initpath = Join-Path '.' 'llama_cpp' '__init__.py' -resolve
118 |             $initcontent = Get-Content $initpath -raw
119 |             $regexstr = '(?s)(?<=__version__ \= ")\d+(?:\.\d+)*(?=")'
120 |             $regexmatch = [Regex]::Matches($initcontent,$regexstr)
121 |             if (!($regexmatch[0].Success)) {throw '__init__.py parsing failed'}
122 |             $newinit = $regexmatch[0].Result(('$`' + '$&' + $buildtag + '$'''))
123 |             New-Item $initpath -itemType File -value $newinit -force
124 |             python -m build --wheel
125 |           }
126 |           
127 |       - name: Upload files to a GitHub release
128 |         id: upload-release
129 |         uses: svenstaro/upload-release-action@2.7.0
130 |         continue-on-error: true
131 |         with:
132 |           file: ./dist/*.whl
133 |           tag: 'cpu'
134 |           file_glob: true
135 |           make_latest: false
136 |           overwrite: true
137 |         
138 |       - uses: actions/upload-artifact@v3
139 |         if: steps.upload-release.outcome == 'failure'
140 |         with:
141 |           name: cpu
142 |           path: ./dist/*.whl
143 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-full-release.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels for New Release
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       version:
 7 |         description: 'Version tag of llama-cpp-python to build: v0.2.11'
 8 |         default: 'v0.2.11'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 |   
15 | jobs:
16 |   run_main:
17 |     name: Build ${{ inputs.version }} CUDA Wheels
18 |     uses: ./.github/workflows/build-wheels.yml
19 |     with:
20 |       version: ${{ inputs.version }}
21 |       cpu: '0'
22 |       
23 |   run_ooba:
24 |     name: Build ${{ inputs.version }} CUDA Wheels for Text Generation Webui
25 |     needs: run_main
26 |     uses: ./.github/workflows/build-wheels-oobabooga.yml
27 |     with:
28 |       version: ${{ inputs.version }}
29 |       
30 |   run_cpu:
31 |     name: Build CPU-only Wheels
32 |     needs: run_ooba
33 |     uses: ./.github/workflows/build-wheels-cpu.yml
34 |     with:
35 |       version: ${{ inputs.version }}
36 |       
37 |   run_macos:
38 |     name: Build MacOS Metal Wheels
39 |     needs: run_cpu
40 |     uses: ./.github/workflows/build-wheels-macos.yml
41 |     with:
42 |       version: ${{ inputs.version }}
43 |       
44 |   run_rocm:
45 |     name: Build ROCm Wheels
46 |     needs: run_macos
47 |     uses: ./.github/workflows/build-wheels-rocm-full.yml
48 |     with:
49 |       version: ${{ inputs.version }}
50 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-ggml-cpu.yml:
--------------------------------------------------------------------------------
  1 | name: Build CPU-only GGML Wheels
  2 | 
  3 | on:
  4 |   workflow_dispatch:
  5 |     inputs:
  6 |       version:
  7 |         description: 'Version tag of llama-cpp-python to build: v0.1.78'
  8 |         default: 'v0.1.78'
  9 |         required: false
 10 |         type: string
 11 |   workflow_call:
 12 |     inputs:
 13 |       version:
 14 |         description: 'Version tag of llama-cpp-python to build: v0.1.78'
 15 |         default: 'v0.1.78'
 16 |         required: false
 17 |         type: string
 18 | 
 19 | permissions:
 20 |   contents: write
 21 | 
 22 | jobs:
 23 |   build_wheels:
 24 |     name: ${{ matrix.os }} ${{ matrix.pyver }} CPU ${{ matrix.avx }}
 25 |     runs-on: ${{ matrix.os }}
 26 |     strategy:
 27 |       matrix:
 28 |         os: [ubuntu-20.04, windows-latest]
 29 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
 30 |         avx: ["AVX","AVX2","AVX512","basic"]
 31 |     defaults:
 32 |       run:
 33 |         shell: pwsh
 34 |     env:
 35 |       AVXVER: ${{ matrix.avx }}
 36 |       PCKGVER: ${{ inputs.version }}
 37 | 
 38 |     steps:
 39 |       - uses: actions/checkout@v4
 40 |         with:
 41 |           repository: 'abetlen/llama-cpp-python'
 42 |           ref: ${{ inputs.version }}
 43 |           submodules: 'recursive'
 44 |           
 45 |       - uses: actions/setup-python@v4
 46 |         with:
 47 |           python-version: ${{ matrix.pyver }}
 48 |           
 49 |       - name: Install Dependencies
 50 |         run: |
 51 |           python -m pip install build wheel
 52 |           
 53 |       - name: Change Package Name
 54 |         run: |
 55 |           $packageVersion = [version]$env:PCKGVER.TrimStart('v')
 56 |           $setup = Get-Content 'setup.py' -raw
 57 |           $pyproject = Get-Content 'pyproject.toml' -raw
 58 |           $cmakelists = Get-Content 'CMakeLists.txt' -raw
 59 |           $regexstr = '(?s)name="llama_cpp_python",(.+)(package_dir={"llama_cpp": "llama_cpp", "llama_cpp.server": "llama_cpp/server"},.+?packages=\["llama_cpp", "llama_cpp.server"],)'
 60 |           if ($packageVersion -gt [version]'0.1.77') {$regexstr = '(?s)name="llama_cpp_python",(.+)(package_dir={"llama_cpp": "llama_cpp", "llama_cpp.server": "llama_cpp/server"},.+?package_data={"llama_cpp": \["py.typed"]},.+?packages=\["llama_cpp", "llama_cpp.server"],)'}
 61 |           $regexmatch = [Regex]::Matches($setup,$regexstr)
 62 |           if (!($regexmatch[0].Success)) {throw 'setup.py parsing failed'}
 63 |           $newstr = 'name="llama_cpp_python_ggml",' + $regexmatch[0].Groups[1].Value + $regexmatch[0].Groups[2].Value.Replace('llama_cpp','llama_cpp_ggml')
 64 |           $newsetup = $regexmatch[0].Result(('$`'+$newstr+'$'''))
 65 |           $regexstr = '(?s)(?<=name = ")llama_cpp_python(".+?packages = \[{include = ")llama_cpp(".+)'
 66 |           $regexmatch = [Regex]::Matches($pyproject,$regexstr)
 67 |           if (!($regexmatch[0].Success)) {throw 'pyproject.toml parsing failed'}
 68 |           $newpyproject = $regexmatch[0].Result(('$`'+'llama_cpp_python_ggml'+'$1llama_cpp_ggml$2'))
 69 |           Copy-Item 'llama_cpp' 'llama_cpp_ggml' -recurse
 70 |           New-Item 'setup.py' -itemType File -value $newsetup -force
 71 |           New-Item 'pyproject.toml' -itemType File -value $newpyproject -force
 72 |           New-Item 'CMakeLists.txt' -itemType File -value $cmakelists.Replace('llama_cpp','llama_cpp_ggml') -force
 73 |           
 74 |       - name: Build Wheel
 75 |         run: |
 76 |           $env:VERBOSE = '1'
 77 |           $env:FORCE_CMAKE = '1'
 78 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = '-DLLAMA_AVX2=off -DLLAMA_FMA=off -DLLAMA_F16C=off'}
 79 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = '-DLLAMA_AVX512=on'}
 80 |           if ($env:AVXVER -eq 'basic') {$env:CMAKE_ARGS = '-DLLAMA_AVX=off -DLLAMA_AVX2=off -DLLAMA_FMA=off -DLLAMA_F16C=off'}
 81 |           if ($packageVersion -gt [version]'0.2.13') {$env:CMAKE_ARGS = "-DLLAMA_NATIVE=off $env:CMAKE_ARGS"}
 82 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cpu$env:AVXVER"
 83 |           
 84 |       - name: Upload files to a GitHub release
 85 |         id: upload-release
 86 |         uses: svenstaro/upload-release-action@2.7.0
 87 |         continue-on-error: true
 88 |         with:
 89 |           file: ./dist/*.whl
 90 |           tag: 'cpu'
 91 |           file_glob: true
 92 |           make_latest: false
 93 |           overwrite: true
 94 |         
 95 |       - uses: actions/upload-artifact@v3
 96 |         if: steps.upload-release.outcome == 'failure'
 97 |         with:
 98 |           name: cpu
 99 |           path: ./dist/*.whl
100 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-macos.yml:
--------------------------------------------------------------------------------
  1 | name: Build Wheels MacOS
  2 | 
  3 | on:
  4 |   workflow_dispatch:
  5 |     inputs:
  6 |       version:
  7 |         description: 'Version tag of llama-cpp-python to build: v0.2.20'
  8 |         default: 'v0.2.20'
  9 |         required: true
 10 |         type: string
 11 |       config:
 12 |         description: 'Override configurations to build: key1:item1-1,item1-2;key2:item2-1,item2-2'
 13 |         default: 'Default'
 14 |         required: false
 15 |         type: string
 16 |       exclude:
 17 |         description: 'Exclude build configurations: key1-1:item1-1,key1-2:item1-2;key2-1:item2-1,key2-2:item2-2'
 18 |         default: 'None'
 19 |         required: false
 20 |         type: string
 21 |   workflow_call:
 22 |     inputs:
 23 |       version:
 24 |         description: 'Version tag of llama-cpp-python to build: v0.2.20'
 25 |         default: 'v0.2.20'
 26 |         required: true
 27 |         type: string
 28 |       config:
 29 |         description: 'Configurations to build: key1:item1-1,item1-2;key2:item2-1,item2-2'
 30 |         default: 'Default'
 31 |         required: false
 32 |         type: string
 33 |       exclude:
 34 |         description: 'Exclude build configurations: key1-1:item1-1,key1-2:item1-2;key2-1:item2-1,key2-2:item2-2'
 35 |         default: 'None'
 36 |         required: false
 37 |         type: string
 38 | 
 39 | permissions:
 40 |   contents: write
 41 | 
 42 | jobs:
 43 |   define_matrix:
 44 |     name: Define Build Matrix
 45 |     runs-on: ubuntu-latest
 46 |     outputs:
 47 |       matrix: ${{ steps.set-matrix.outputs.matrix }}
 48 |     defaults:
 49 |       run:
 50 |         shell: pwsh
 51 |     env:
 52 |       CONFIGIN: ${{ inputs.config }}
 53 |       EXCLUDEIN: ${{ inputs.exclude }}
 54 |         
 55 |     steps:
 56 |       - name: Define Job Output
 57 |         id: set-matrix
 58 |         run: |
 59 |           $matrix = @{
 60 |               'os' = "macos-13", "macos-12"
 61 |               'pyver' = "3.10", "3.8", "3.9", "3.11"
 62 |           }
 63 |           
 64 |           if ($env:CONFIGIN -ne 'Default') {$env:CONFIGIN.split(';').foreach({$matrix[$_.split(':')[0]] = $_.split(':')[1].split(',')})}
 65 |           
 66 |           if ($env:EXCLUDEIN -ne 'None') {
 67 |               $exclusions = @()
 68 |               $exclusions += $env:EXCLUDEIN.split(';').replace(':','=').replace(',',"`n") | ConvertFrom-StringData
 69 |               $matrix['exclude'] = $exclusions
 70 |           }
 71 |           
 72 |           $matrixOut = ConvertTo-Json $matrix -Compress
 73 |           Write-Output ('matrix=' + $matrixOut) >> $env:GITHUB_OUTPUT
 74 |           
 75 |   build_wheels:
 76 |     name: ${{ matrix.os }} Python ${{ matrix.pyver }}
 77 |     needs: define_matrix
 78 |     runs-on: ${{ matrix.os }}
 79 |     strategy:
 80 |       matrix: ${{ fromJSON(needs.define_matrix.outputs.matrix) }}
 81 |     env:
 82 |       OSVER: ${{ matrix.os }}
 83 | 
 84 |     steps:
 85 |       - uses: actions/checkout@v4
 86 |         with:
 87 |           repository: 'abetlen/llama-cpp-python'
 88 |           ref: ${{ inputs.version }}
 89 |           submodules: 'recursive'
 90 |           
 91 |       - uses: actions/setup-python@v4
 92 |         with:
 93 |           python-version: ${{ matrix.pyver }}
 94 |           
 95 |       - name: Install Dependencies
 96 |         run: |
 97 |           python -m pip install build wheel cmake
 98 |           
 99 |       - name: Build Wheel
100 |         run: |
101 |           XCODE15PATH="/Applications/Xcode_15.0.app/Contents/Developer"
102 |           XCODE15BINPATH="${XCODE15PATH}/Toolchains/XcodeDefault.xctoolchain/usr/bin"
103 |           export CMAKE_ARGS="-DLLAMA_NATIVE=off -DLLAMA_METAL=on"
104 |           [[ "$OSVER" == "macos-13" ]] && export CC="${XCODE15BINPATH}/cc" && export CXX="${XCODE15BINPATH}/c++" && export MACOSX_DEPLOYMENT_TARGET="13.0"
105 |           [[ "$OSVER" == "macos-12" ]] && export MACOSX_DEPLOYMENT_TARGET="12.0"
106 |           [[ "$OSVER" == "macos-11" ]] && export MACOSX_DEPLOYMENT_TARGET="11.0"
107 |           
108 |           export CMAKE_OSX_ARCHITECTURES="arm64" && export ARCHFLAGS="-arch arm64"
109 |           VERBOSE=1 python -m build --wheel
110 |           
111 |           if [[ "$OSVER" == "macos-13" ]]; then
112 |             export SDKROOT="${XCODE15PATH}/Platforms/MacOSX.platform/Developer/SDKs/MacOSX14.0.sdk"
113 |             export MACOSX_DEPLOYMENT_TARGET="14.0"
114 |             VERBOSE=1 python -m build --wheel
115 |           fi
116 |           
117 |           for file in ./dist/*.whl; do cp "$file" "${file/arm64.whl/aarch64.whl}"; done
118 |           
119 |           export CMAKE_OSX_ARCHITECTURES="x86_64" && export CMAKE_ARGS="-DLLAMA_NATIVE=off -DLLAMA_AVX=off -DLLAMA_AVX2=off -DLLAMA_FMA=off -DLLAMA_F16C=off -DLLAMA_METAL=on" && export ARCHFLAGS="-arch x86_64"
120 |           VERBOSE=1 python -m build --wheel
121 |           
122 |           if [[ "$OSVER" == "macos-13" ]]; then
123 |             export SDKROOT="${XCODE15PATH}/Platforms/MacOSX.platform/Developer/SDKs/MacOSX14.0.sdk"
124 |             export MACOSX_DEPLOYMENT_TARGET="14.0"
125 |             VERBOSE=1 python -m build --wheel
126 |           fi
127 |           
128 |       - name: Upload files to a GitHub release
129 |         id: upload-release
130 |         uses: svenstaro/upload-release-action@2.7.0
131 |         continue-on-error: true
132 |         with:
133 |           file: ./dist/*.whl
134 |           tag: 'metal'
135 |           file_glob: true
136 |           make_latest: false
137 |           overwrite: true
138 |         
139 |       - uses: actions/upload-artifact@v3
140 |         if: steps.upload-release.outcome == 'failure'
141 |         with:
142 |           name: macos-wheels
143 |           path: ./dist/*.whl
144 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-oobabooga-basic.yml:
--------------------------------------------------------------------------------
 1 | name: Build Basic Wheels for Text Generation Webui
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       version:
 7 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
 8 |         default: 'v0.2.9'
 9 |         required: true
10 |         type: string
11 |   workflow_call:
12 |     inputs:
13 |       version:
14 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
15 |         default: 'v0.2.9'
16 |         required: true
17 |         type: string
18 | 
19 | permissions:
20 |   contents: write
21 | 
22 | jobs:
23 |   build_textgen_wheels:
24 |     name: Textgen Wheels
25 |     uses: ./.github/workflows/build-wheels-oobabooga.yml
26 |     with:
27 |       version: ${{ inputs.version }}
28 |       config: 'avxver:basic'


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-oobabooga-rocm.yml:
--------------------------------------------------------------------------------
 1 | name: Build ROCm Wheels for Text Generation Webui
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       version:
 7 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
 8 |         default: 'v0.2.9'
 9 |         required: true
10 |         type: string
11 |   workflow_call:
12 |     inputs:
13 |       version:
14 |         description: 'Version tag of llama-cpp-python to build: v0.2.9'
15 |         default: 'v0.2.9'
16 |         required: true
17 |         type: string
18 | 
19 | permissions:
20 |   contents: write
21 | 
22 | jobs:
23 |   build_wheels_rocm:
24 |     name: ROCm Wheels
25 |     uses: ./.github/workflows/build-wheels-rocm-full.yml
26 |     with:
27 |       version: ${{ inputs.version }}
28 |       config: 'rename:1'
29 | 


--------------------------------------------------------------------------------
/.github/workflows/build-wheels-prioritized-release.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels for New Release with Prioritization
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |     inputs:
 6 |       version:
 7 |         description: 'Version tag of llama-cpp-python to build: v0.2.14'
 8 |         default: 'v0.2.14'
 9 |         required: true
10 |         type: string
11 | 
12 | permissions:
13 |   contents: write
14 | 
15 | jobs:          
16 |   build_textgen_wheels_prio:
17 |     name: Textgen Prioritized
18 |     uses: ./.github/workflows/build-wheels-oobabooga.yml
19 |     with:
20 |       version: ${{ inputs.version }}
21 |       config: 'pyver:3.10;cuda:12.1.1'
22 |       
23 |   build_wheels_main_prio:
24 |     name: Main Prioritized
25 |     uses: ./.github/workflows/build-wheels.yml
26 |     with:
27 |       version: ${{ inputs.version }}
28 |       cpu: '0'
29 |       config: 'pyver:3.10;cuda:12.1.1'
30 |       
31 |   build_wheels_cpu_prio:
32 |     name: CPU-only Prioritized
33 |     uses: ./.github/workflows/build-wheels-cpu.yml
34 |     with:
35 |       version: ${{ inputs.version }}
36 |       config: 'pyver:3.10;avx:basic'
37 |       
38 |   build_wheels_rocm_prio:
39 |     name: ROCm Prioritized
40 |     uses: ./.github/workflows/build-wheels-rocm-full.yml
41 |     with:
42 |       version: ${{ inputs.version }}
43 |       config: 'os:ubuntu-20.04;pyver:3.10;rocm:5.6.1;rename:1'
44 |       exclude: 'None'
45 |       
46 |   build_wheels_macos_prio:
47 |     name: MacOS Metal Prioritized
48 |     uses: ./.github/workflows/build-wheels-macos.yml
49 |     with:
50 |       version: ${{ inputs.version }}
51 |       config: 'pyver:3.10'
52 |       
53 |   build_wheels_main:
54 |     name: Main Wheels
55 |     needs: ['build_wheels_main_prio', 'build_textgen_wheels_prio', 'build_wheels_cpu_prio', 'build_wheels_rocm_prio', 'build_wheels_macos_prio']
56 |     uses: ./.github/workflows/build-wheels.yml
57 |     with:
58 |       version: ${{ inputs.version }}
59 |       cpu: '0'
60 |       exclude: 'pyver:3.10,cuda:12.1.1'
61 |       
62 |   build_textgen_wheels:
63 |     name: Textgen Wheels
64 |     needs: build_wheels_main
65 |     uses: ./.github/workflows/build-wheels-oobabooga.yml
66 |     with:
67 |       version: ${{ inputs.version }}
68 |       exclude: 'pyver:3.10,cuda:12.1.1'
69 |       
70 |   build_wheels_cpu:
71 |     name: CPU-only Wheels
72 |     needs: build_textgen_wheels
73 |     uses: ./.github/workflows/build-wheels-cpu.yml
74 |     with:
75 |       version: ${{ inputs.version }}
76 |       exclude: 'pyver:3.10,avx:basic'
77 |       
78 |   build_wheels_macos:
79 |     name: MacOS Metal Wheels
80 |     needs: build_wheels_cpu
81 |     uses: ./.github/workflows/build-wheels-macos.yml
82 |     with:
83 |       version: ${{ inputs.version }}
84 |       exclude: 'pyver:3.10'
85 |       
86 |   build_wheels_rocm:
87 |     name: ROCm Wheels
88 |     needs: build_wheels_macos
89 |     uses: ./.github/workflows/build-wheels-rocm-full.yml
90 |     with:
91 |       version: ${{ inputs.version }}
92 |       exclude: 'os:ubuntu-20.04,pyver:3.10,rocm:5.6.1,rename:1'
93 | 


--------------------------------------------------------------------------------
/.github/workflows/deploy-index.yml:
--------------------------------------------------------------------------------
 1 | name: Deploy package index to Pages
 2 | 
 3 | on:
 4 |   # Runs on pushes targeting the default branch when html files are modified under docs directory
 5 |   push:
 6 |     branches: ["main"]
 7 |     paths:
 8 |       - 'index/**.html'
 9 | 
10 |   # Allows you to run this workflow manually from the Actions tab
11 |   workflow_dispatch:
12 | 
13 | # Sets permissions of the GITHUB_TOKEN to allow deployment to GitHub Pages
14 | permissions:
15 |   contents: read
16 |   pages: write
17 |   id-token: write
18 | 
19 | # Allow only one concurrent deployment, skipping runs queued between the run in-progress and latest queued.
20 | # However, do NOT cancel in-progress runs as we want to allow these production deployments to complete.
21 | concurrency:
22 |   group: "pages"
23 |   cancel-in-progress: false
24 | 
25 | jobs:
26 |   # Single deploy job since we're just deploying
27 |   deploy:
28 |     environment:
29 |       name: github-pages
30 |       url: ${{ steps.deployment.outputs.page_url }}
31 |     runs-on: ubuntu-latest
32 |     steps:
33 |       - name: Checkout
34 |         uses: actions/checkout@v4
35 |         with:
36 |           sparse-checkout: 'index/'
37 |       - name: Setup Pages
38 |         uses: actions/configure-pages@v4
39 |       - name: Upload artifact
40 |         uses: actions/upload-pages-artifact@v2
41 |         with:
42 |           # Upload docs directory
43 |           path: 'index/'
44 |       - name: Deploy to GitHub Pages
45 |         id: deployment
46 |         uses: actions/deploy-pages@v3
47 | 


--------------------------------------------------------------------------------
/LICENSE:
--------------------------------------------------------------------------------
 1 | This is free and unencumbered software released into the public domain.
 2 | 
 3 | Anyone is free to copy, modify, publish, use, compile, sell, or
 4 | distribute this software, either in source code form or as a compiled
 5 | binary, for any purpose, commercial or non-commercial, and by any
 6 | means.
 7 | 
 8 | In jurisdictions that recognize copyright laws, the author or authors
 9 | of this software dedicate any and all copyright interest in the
10 | software to the public domain. We make this dedication for the benefit
11 | of the public at large and to the detriment of our heirs and
12 | successors. We intend this dedication to be an overt act of
13 | relinquishment in perpetuity of all present and future rights to this
14 | software under copyright law.
15 | 
16 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
17 | EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
18 | MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
19 | IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
20 | OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
21 | ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
22 | OTHER DEALINGS IN THE SOFTWARE.
23 | 
24 | For more information, please refer to <https://unlicense.org>
25 | 


--------------------------------------------------------------------------------
/README.md:
--------------------------------------------------------------------------------
 1 | # llama-cpp-python cuBLAS wheels
 2 | Wheels for [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) compiled with cuBLAS support.
 3 | 
 4 | Requirements:
 5 | - Windows x64, Linux x64, or MacOS 11.0+
 6 | - CUDA 11.6 - 12.2
 7 | - CPython 3.8 - 3.11
 8 | 
 9 | > [!WARNING]
10 | > MacOS 11 and Windows ROCm wheels are unavailable for 0.2.21+.  
11 | > This is due to build issues with llama.cpp that are not yet resolved.
12 | 
13 | ROCm builds for AMD GPUs: https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/tag/rocm  
14 | Metal builds for MacOS 11.0+: https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/tag/metal
15 | 
16 | Installation instructions:
17 | ---
18 | To install, you can use this command:
19 | ```
20 | python -m pip install llama-cpp-python --prefer-binary --extra-index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
21 | ```
22 | This will install the latest llama-cpp-python version available from here for CUDA 11.7. You can change `cu117` to change the CUDA version.  
23 | You can also change `AVX2` to `AVX`, `AVX512` or `basic` based on what your CPU supports.  
24 | `basic` is a build without `AVX`, `FMA` and `F16C` instructions for old or basic CPUs.  
25 | CPU-only builds are also available by changing `cu117` to `cpu`.
26 | 
27 | You can install a specific version with:
28 | ```
29 | python -m pip install llama-cpp-python==<version> --prefer-binary --extra-index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
30 | ```
31 | An example for installing 0.1.62 for CUDA 12.1 on a CPU without AVX2 support:
32 | ```
33 | python -m pip install llama-cpp-python==0.1.62 --prefer-binary --extra-index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX/cu121
34 | ```
35 | List of available versions:
36 | ```
37 | python -m pip index versions llama-cpp-python --index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
38 | ```
39 | 
40 | If you are replacing an already existing installation, you may need to uninstall that version before running the command above.  
41 | You can also replace the existing version in one command like so:
42 | ```
43 | python -m pip install llama-cpp-python --force-reinstall --no-deps --index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
44 | -OR-
45 | python -m pip install llama-cpp-python==0.1.66 --force-reinstall --no-deps --index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
46 | -OR-
47 | python -m pip install llama-cpp-python --prefer-binary --upgrade --extra-index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
48 | ```
49 | 
50 | Wheels can be manually downloaded from: https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels
51 | 
52 | ---
53 | I have renamed llama-cpp-python packages available to ease the transition to GGUF.  
54 | This is accomplished by installing the renamed package alongside the main llama-cpp-python package.  
55 | This should allow applications to maintain GGML support while still supporting GGUF.
56 | ```
57 | python -m pip install llama-cpp-python-ggml --prefer-binary --extra-index-url=https://jllllll.github.io/llama-cpp-python-cuBLAS-wheels/AVX2/cu117
58 | ```
59 | 
60 | ---
61 | ### All wheels are compiled using GitHub Actions
62 | 


--------------------------------------------------------------------------------
/host-index.bat:
--------------------------------------------------------------------------------
1 | @echo off
2 | 
3 | cd index
4 | python -m http.server 7860
5 | 


--------------------------------------------------------------------------------
/host-index.sh:
--------------------------------------------------------------------------------
1 | #!/bin/bash
2 | 
3 | cd index
4 | python -m http.server 7860
5 | 


--------------------------------------------------------------------------------
/index/AVX/cpu/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cpu/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/cu116/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cu116/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cu117/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cu118/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cu120/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cu121/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX/cu122/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu116/">CUDA 11.6</a><br/>
 5 |     <a href="cu117/">CUDA 11.7</a><br/>
 6 |     <a href="cu118/">CUDA 11.8</a><br/>
 7 |     <a href="cu120/">CUDA 12.0</a><br/>
 8 |     <a href="cu121/">CUDA 12.1</a><br/>
 9 |     <a href="cu122/">CUDA 12.2</a><br/>
10 |     <a href="rocm5.4.2/">ROCm 5.4.2</a><br/>
11 |     <a href="rocm5.5/">ROCm 5.5</a><br/>
12 |     <a href="rocm5.5.1/">ROCm 5.5.1</a><br/>
13 |     <a href="rocm5.6.1/">ROCm 5.6.1</a><br/>
14 |     <a href="cpu/">cpu</a><br/>
15 |   </body>
16 | </html>
17 | 


--------------------------------------------------------------------------------
/index/AVX/rocm5.4.2/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX/rocm5.5.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX/rocm5.5/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX/rocm5.6.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX2/cpu/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cpu/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx2-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx2-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/cu116/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cu116/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cu117/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cu118/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cu120/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cu121/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX2/cu122/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/wheels/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX2/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu116/">CUDA 11.6</a><br/>
 5 |     <a href="cu117/">CUDA 11.7</a><br/>
 6 |     <a href="cu118/">CUDA 11.8</a><br/>
 7 |     <a href="cu120/">CUDA 12.0</a><br/>
 8 |     <a href="cu121/">CUDA 12.1</a><br/>
 9 |     <a href="cu122/">CUDA 12.2</a><br/>
10 |     <a href="rocm5.4.2/">ROCm 5.4.2</a><br/>
11 |     <a href="rocm5.5/">ROCm 5.5</a><br/>
12 |     <a href="rocm5.5.1/">ROCm 5.5.1</a><br/>
13 |     <a href="rocm5.6.1/">ROCm 5.6.1</a><br/>
14 |     <a href="cpu/">cpu</a><br/>
15 |   </body>
16 | </html>
17 | 


--------------------------------------------------------------------------------
/index/AVX2/rocm5.4.2/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX2/rocm5.5.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX2/rocm5.5/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX2/rocm5.6.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/AVX512/cpu/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cpu/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpuavx512-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpuavx512-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/cu116/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cu116/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cu117/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cu118/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cu120/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cu121/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/AVX512/cu122/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/AVX512/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/AVX512/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu116/">CUDA 11.6</a><br/>
 5 |     <a href="cu117/">CUDA 11.7</a><br/>
 6 |     <a href="cu118/">CUDA 11.8</a><br/>
 7 |     <a href="cu120/">CUDA 12.0</a><br/>
 8 |     <a href="cu121/">CUDA 12.1</a><br/>
 9 |     <a href="cu122/">CUDA 12.2</a><br/>
10 |     <a href="cpu/">cpu</a><br/>
11 |   </body>
12 | </html>
13 | 


--------------------------------------------------------------------------------
/index/basic/cpu/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cpu/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/cpu/llama_cpp_python_ggml-0.1.78+cpubasic-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cpubasic-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/cu116/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cu116/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu116-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cu117/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu117-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cu118/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu118-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cu120/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu120-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cu121/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu121-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |     <a href="llama-cpp-python-ggml/">llama_cpp_python_ggml</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/basic/cu122/llama-cpp-python-ggml/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp37-cp37m-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp38-cp38-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp39-cp39-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp310-cp310-win_amd64.whl</a><br/>
12 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-linux_x86_64.whl</a><br/>
13 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/basic/llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml-0.1.78+cu122-cp311-cp311-win_amd64.whl</a><br/>
14 |   </body>
15 | </html>
16 | 


--------------------------------------------------------------------------------
/index/basic/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu116/">CUDA 11.6</a><br/>
 5 |     <a href="cu117/">CUDA 11.7</a><br/>
 6 |     <a href="cu118/">CUDA 11.8</a><br/>
 7 |     <a href="cu120/">CUDA 12.0</a><br/>
 8 |     <a href="cu121/">CUDA 12.1</a><br/>
 9 |     <a href="cu122/">CUDA 12.2</a><br/>
10 |     <a href="rocm5.4.2/">ROCm 5.4.2</a><br/>
11 |     <a href="rocm5.5/">ROCm 5.5</a><br/>
12 |     <a href="rocm5.5.1/">ROCm 5.5.1</a><br/>
13 |     <a href="rocm5.6.1/">ROCm 5.6.1</a><br/>
14 |     <a href="cpu/">cpu</a><br/>
15 |   </body>
16 | </html>
17 | 


--------------------------------------------------------------------------------
/index/basic/rocm5.4.2/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/basic/rocm5.5.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/basic/rocm5.5/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/basic/rocm5.6.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python/">llama_cpp_python</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="AVX/">AVX</a><br/>
 5 |     <a href="AVX2/">AVX2</a><br/>
 6 |     <a href="AVX512/">AVX512</a><br/>
 7 |     <a href="basic/">basic</a><br/>
 8 |   </body>
 9 | </html>
10 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu117/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117avx-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu118/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118avx-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu120/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120avx-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu121/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121avx-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/cu122/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122avx-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu117/">CUDA 11.7</a><br/>
 5 |     <a href="cu118/">CUDA 11.8</a><br/>
 6 |     <a href="cu120/">CUDA 12.0</a><br/>
 7 |     <a href="cu121/">CUDA 12.1</a><br/>
 8 |     <a href="cu122/">CUDA 12.2</a><br/>
 9 |     <a href="rocm5.4.2/">ROCm 5.4.2</a><br/>
10 |     <a href="rocm5.5/">ROCm 5.5</a><br/>
11 |     <a href="rocm5.5.1/">ROCm 5.5.1</a><br/>
12 |     <a href="rocm5.6.1/">ROCm 5.6.1</a><br/>
13 |   </body>
14 | </html>
15 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/rocm5.4.2/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/rocm5.5.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/rocm5.5/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX/rocm5.6.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu117/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu118/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu120/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu121/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/cu122/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu117/">CUDA 11.7</a><br/>
 5 |     <a href="cu118/">CUDA 11.8</a><br/>
 6 |     <a href="cu120/">CUDA 12.0</a><br/>
 7 |     <a href="cu121/">CUDA 12.1</a><br/>
 8 |     <a href="cu122/">CUDA 12.2</a><br/>
 9 |     <a href="rocm5.4.2/">ROCm 5.4.2</a><br/>
10 |     <a href="rocm5.5/">ROCm 5.5</a><br/>
11 |     <a href="rocm5.5.1/">ROCm 5.5.1</a><br/>
12 |     <a href="rocm5.6.1/">ROCm 5.6.1</a><br/>
13 |   </body>
14 | </html>
15 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/rocm5.4.2/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/rocm5.5.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/rocm5.5/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/AVX2/rocm5.6.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu117/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu117/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu117basic-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu118/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu118/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu118basic-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu120/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu120/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu120basic-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu121/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu121/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu121basic-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu122/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |     <a href="llama-cpp-python-ggml-cuda/">llama_cpp_python_ggml_cuda</a>
6 |   </body>
7 | </html>
8 | 


--------------------------------------------------------------------------------
/index/textgen/basic/cu122/llama-cpp-python-ggml-cuda/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp38-cp38-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp38-cp38-linux_x86_64.whl</a><br/>
 5 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp38-cp38-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp38-cp38-win_amd64.whl</a><br/>
 6 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp39-cp39-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp39-cp39-linux_x86_64.whl</a><br/>
 7 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp39-cp39-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp39-cp39-win_amd64.whl</a><br/>
 8 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp310-cp310-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp310-cp310-linux_x86_64.whl</a><br/>
 9 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp310-cp310-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp310-cp310-win_amd64.whl</a><br/>
10 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp311-cp311-linux_x86_64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp311-cp311-linux_x86_64.whl</a><br/>
11 |     <a href="https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/download/textgen-webui/llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp311-cp311-win_amd64.whl">llama_cpp_python_ggml_cuda-0.1.78+cu122basic-cp311-cp311-win_amd64.whl</a><br/>
12 |   </body>
13 | </html>
14 | 


--------------------------------------------------------------------------------
/index/textgen/basic/index.html:
--------------------------------------------------------------------------------
 1 | <!DOCTYPE html>
 2 | <html>
 3 |   <body>
 4 |     <a href="cu117/">CUDA 11.7</a><br/>
 5 |     <a href="cu118/">CUDA 11.8</a><br/>
 6 |     <a href="cu120/">CUDA 12.0</a><br/>
 7 |     <a href="cu121/">CUDA 12.1</a><br/>
 8 |     <a href="cu122/">CUDA 12.2</a><br/>
 9 |     <a href="rocm5.4.2/">ROCm 5.4.2</a><br/>
10 |     <a href="rocm5.5/">ROCm 5.5</a><br/>
11 |     <a href="rocm5.5.1/">ROCm 5.5.1</a><br/>
12 |     <a href="rocm5.6.1/">ROCm 5.6.1</a><br/>
13 |   </body>
14 | </html>
15 | 


--------------------------------------------------------------------------------
/index/textgen/basic/rocm5.4.2/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/basic/rocm5.5.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/basic/rocm5.5/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/basic/rocm5.6.1/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="llama-cpp-python-cuda/">llama_cpp_python_cuda</a><br/>
5 |   </body>
6 | </html>
7 | 


--------------------------------------------------------------------------------
/index/textgen/index.html:
--------------------------------------------------------------------------------
1 | <!DOCTYPE html>
2 | <html>
3 |   <body>
4 |     <a href="AVX/">AVX</a><br/>
5 |     <a href="AVX2/">AVX2</a><br/>
6 |     <a href="basic/">basic</a><br/>
7 |   </body>
8 | </html>
9 | 


--------------------------------------------------------------------------------
/old_workflows/build-all-wheels.yml:
--------------------------------------------------------------------------------
 1 | name: Build All Wheels
 2 | 
 3 | on: workflow_dispatch
 4 | 
 5 | permissions:
 6 |   contents: write
 7 |   
 8 | jobs:
 9 |   run_workflow_0-1-62:
10 |     name: Build Wheels 0.1.62
11 |     uses: ./.github/workflows/build-wheels-0.1.62.yml
12 |     
13 |   run_workflow_0-1-66:
14 |     name: Build Wheels 0.1.66
15 |     needs: run_workflow_0-1-62
16 |     uses: ./.github/workflows/build-wheels-0.1.66.yml
17 |     
18 |   run_workflow_0-1-67:
19 |     name: Build Wheels 0.1.67
20 |     needs: run_workflow_0-1-66
21 |     uses: ./.github/workflows/build-wheels-0.1.67.yml
22 |     
23 |   run_workflow_0-1-68:
24 |     name: Build Wheels 0.1.68
25 |     needs: run_workflow_0-1-67
26 |     uses: ./.github/workflows/build-wheels-0.1.68.yml
27 |     
28 |   run_workflow_0-1-70:
29 |     name: Build Wheels 0.1.70
30 |     needs: run_workflow_0-1-68
31 |     uses: ./.github/workflows/build-wheels-0.1.70.yml
32 |     
33 |   run_workflow_0-1-71:
34 |     name: Build Wheels 0.1.71
35 |     needs: run_workflow_0-1-70
36 |     uses: ./.github/workflows/build-wheels-0.1.71.yml
37 |     
38 |   run_workflow_0-1-72:
39 |     name: Build Wheels 0.1.72
40 |     needs: run_workflow_0-1-71
41 |     uses: ./.github/workflows/build-wheels-0.1.72.yml
42 |     
43 |   run_workflow_0-1-73:
44 |     name: Build Wheels 0.1.73
45 |     needs: run_workflow_0-1-72
46 |     uses: ./.github/workflows/build-wheels-0.1.73.yml
47 |     
48 |   run_workflow_0-1-74:
49 |     name: Build Wheels 0.1.74
50 |     needs: run_workflow_0-1-73
51 |     uses: ./.github/workflows/build-wheels-0.1.74.yml
52 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheel-rocm-windows.yml:
--------------------------------------------------------------------------------
  1 | name: Build ROCm Windows Wheel
  2 | 
  3 | on:
  4 |   workflow_dispatch:
  5 |     inputs:
  6 |       version:
  7 |         description: 'Version tag of llama-cpp-python to build: v0.1.79'
  8 |         default: 'v0.1.79'
  9 |         required: false
 10 |         type: string
 11 |   workflow_call:
 12 |     inputs:
 13 |       version:
 14 |         description: 'Version tag of llama-cpp-python to build: v0.1.79'
 15 |         default: 'v0.1.79'
 16 |         required: false
 17 |         type: string
 18 | 
 19 | permissions:
 20 |   contents: write
 21 | 
 22 | jobs:
 23 |   build_libs:
 24 |     name: Build ROCm Lib
 25 |     runs-on: windows-latest
 26 |     defaults:
 27 |       run:
 28 |         shell: pwsh
 29 | 
 30 |     steps:
 31 |       - uses: actions/checkout@v3
 32 |         with:
 33 |           repository: 'abetlen/llama-cpp-python'
 34 |           ref: ${{ inputs.version }}
 35 |           submodules: 'recursive'
 36 |           
 37 |       - name: Install ROCm SDK
 38 |         run: |
 39 |           curl -LO https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-23.Q3-Win10-Win11-For-HIP.exe
 40 |           Start-Process 'AMD-Software-PRO-Edition-23.Q3-Win10-Win11-For-HIP.exe' -ArgumentList '-install' -NoNewWindow -Wait
 41 |           echo "C:\Program Files\AMD\ROCm\5.5\bin" >> $env:GITHUB_PATH
 42 |           echo 'ROCM_PATH=C:\Program Files\AMD\ROCm\5.5' >> $env:GITHUB_ENV
 43 |           echo 'HIP_PATH=C:\Program Files\AMD\ROCm\5.5' >> $env:GITHUB_ENV
 44 |           echo "ROCM_VERSION=5.5.1" >> $env:GITHUB_ENV
 45 |           
 46 |       - uses: actions/setup-python@v3
 47 |         with:
 48 |           python-version: "3.10"
 49 |           
 50 |       - name: Install Dependencies
 51 |         run: |
 52 |           python -m pip install cmake ninja
 53 |         
 54 |       - name: Build Lib
 55 |         run: |
 56 |           $env:CC = 'C:\Program Files\AMD\ROCm\5.5\bin\clang.exe'
 57 |           $env:CXX = 'C:\Program Files\AMD\ROCm\5.5\bin\clang++.exe'
 58 |           $env:CMAKE_PREFIX_PATH = 'C:\Program Files\AMD\ROCm\5.5'
 59 |           $env:VERBOSE = '1'
 60 |           mkdir 'build'
 61 |           Set-Location '.\vendor\llama.cpp'
 62 |           cmake -B build -G "Ninja" -DLLAMA_HIPBLAS=ON -DBUILD_SHARED_LIBS=ON '-DGPU_TARGETS=gfx803;gfx900;gfx906:xnack-;gfx908:xnack-;gfx90a:xnack+;gfx90a:xnack-;gfx1010;gfx1012;gfx1030;gfx1100;gfx1101;gfx1102'
 63 |           cmake --build build --config Release --target llama
 64 |           Copy-Item '.\build\bin\llama.dll' '..\..\build'
 65 |         
 66 |       - uses: actions/upload-artifact@v3
 67 |         with:
 68 |           name: 'win-rocm-lib'
 69 |           path: ./build/llama.dll
 70 |           
 71 |   build_wheel:
 72 |     name: Build ROCm Wheels
 73 |     runs-on: windows-latest
 74 |     needs: build_libs
 75 |     strategy:
 76 |       matrix:
 77 |         pyver: ["3.8", "3.9", "3.10", "3.11"]
 78 |     defaults:
 79 |       run:
 80 |         shell: pwsh
 81 |     env:
 82 |       PCKGVER: ${{ inputs.version }}
 83 | 
 84 |     steps:
 85 |       - uses: actions/checkout@v3
 86 |         with:
 87 |           repository: 'abetlen/llama-cpp-python'
 88 |           ref: ${{ inputs.version }}
 89 |           
 90 |       - uses: actions/download-artifact@v3
 91 |         with:
 92 |           name: 'win-rocm-lib'
 93 |           path: ./llama_cpp
 94 |           
 95 |       - uses: actions/setup-python@v3
 96 |         with:
 97 |           python-version: ${{ matrix.pyver }}
 98 |           
 99 |       - name: Install Dependencies
100 |         run: |
101 |           python -m pip install build wheel cmake scikit-build ninja
102 |         
103 |       - name: Build Wheel
104 |         run: |
105 |           $packageVersion = [version]$env:PCKGVER.TrimStart('v')
106 |           $setup = Get-Content 'setup.py' -raw
107 |           if ($packageVersion -lt [version]'0.1.78') {$newsetup = $setup.Replace("packages=[`"llama_cpp`", `"llama_cpp.server`"],","packages=[`"llama_cpp`", `"llama_cpp.server`"],`n    package_data={'llama_cpp': ['llama.dll']},")}
108 |           if ($packageVersion -gt [version]'0.1.77') {$newsetup = $setup.Replace('package_data={"llama_cpp": ["py.typed"]},','package_data={"llama_cpp": ["py.typed", "llama.dll"]},')}
109 |           New-Item 'setup.py' -itemType File -value $newsetup -force
110 |           python setup.py --skip-cmake bdist_wheel egg_info --tag-build=+rocm5.5.1
111 |           
112 |       - name: Upload files to a GitHub release
113 |         id: upload-release
114 |         uses: svenstaro/upload-release-action@2.6.1
115 |         continue-on-error: true
116 |         with:
117 |           file: ./dist/*.whl
118 |           tag: rocm
119 |           file_glob: true
120 |           make_latest: false
121 |           overwrite: true
122 |         
123 |       - uses: actions/upload-artifact@v3
124 |         if: steps.upload-release.outcome == 'failure'
125 |         with:
126 |           name: 'win-rocm-wheels'
127 |           path: ./dist/*.whl
128 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.62.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.62
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.62'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           $env:CUDAFLAGS = '-arch=all'
70 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
71 |           
72 |       - name: Upload files to a GitHub release
73 |         id: upload-release
74 |         uses: svenstaro/upload-release-action@2.6.1
75 |         continue-on-error: true
76 |         with:
77 |           file: ./dist/*.whl
78 |           tag: ${{ matrix.releasetag }}
79 |           file_glob: true
80 |           make_latest: false
81 |           overwrite: true
82 |         
83 |       - uses: actions/upload-artifact@v3
84 |         if: steps.upload-release.outcome == 'failure'
85 |         with:
86 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
87 |           path: ./dist/*.whl
88 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.66.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.66
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.66'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl
87 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.67.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.67
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.67'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl
87 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.68.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.68
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.68'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl
87 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.69.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.69
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.69'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'
67 |           if ([version]$env:CUDAVER -ge [version]'12.0') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'}
68 |           if ([version]$env:CUDAVER -lt [version]'11.8') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86'}
69 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
70 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
71 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
72 |           
73 |       - name: Upload files to a GitHub release
74 |         id: upload-release
75 |         uses: svenstaro/upload-release-action@2.6.1
76 |         continue-on-error: true
77 |         with:
78 |           file: ./dist/*.whl
79 |           tag: ${{ matrix.releasetag }}
80 |           file_glob: true
81 |           make_latest: false
82 |           overwrite: true
83 |         
84 |       - uses: actions/upload-artifact@v3
85 |         if: steps.upload-release.outcome == 'failure'
86 |         with:
87 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
88 |           path: ./dist/*.whl
89 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.70.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.70
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.70'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'
67 |           if ([version]$env:CUDAVER -ge [version]'12.0') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'}
68 |           if ([version]$env:CUDAVER -lt [version]'11.8') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86'}
69 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
70 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
71 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
72 |           
73 |       - name: Upload files to a GitHub release
74 |         id: upload-release
75 |         uses: svenstaro/upload-release-action@2.6.1
76 |         continue-on-error: true
77 |         with:
78 |           file: ./dist/*.whl
79 |           tag: ${{ matrix.releasetag }}
80 |           file_glob: true
81 |           make_latest: false
82 |           overwrite: true
83 |         
84 |       - uses: actions/upload-artifact@v3
85 |         if: steps.upload-release.outcome == 'failure'
86 |         with:
87 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
88 |           path: ./dist/*.whl
89 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.71.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.71
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.71'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'
67 |           if ([version]$env:CUDAVER -ge [version]'12.0') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'}
68 |           if ([version]$env:CUDAVER -lt [version]'11.8') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86'}
69 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
70 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
71 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
72 |           
73 |       - name: Upload files to a GitHub release
74 |         id: upload-release
75 |         uses: svenstaro/upload-release-action@2.6.1
76 |         continue-on-error: true
77 |         with:
78 |           file: ./dist/*.whl
79 |           tag: ${{ matrix.releasetag }}
80 |           file_glob: true
81 |           make_latest: false
82 |           overwrite: true
83 |         
84 |       - uses: actions/upload-artifact@v3
85 |         if: steps.upload-release.outcome == 'failure'
86 |         with:
87 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
88 |           path: ./dist/*.whl
89 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.72.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.72
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: 'v0.1.72'
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'
67 |           if ([version]$env:CUDAVER -ge [version]'12.0') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=52;61-real;70-real;72-real;75-real;80-real;86-real;89-real;90'}
68 |           if ([version]$env:CUDAVER -lt [version]'11.8') {$env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=35-real;37-real;52;61-real;70-real;72-real;75-real;80-real;86'}
69 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
70 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
71 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
72 |           
73 |       - name: Upload files to a GitHub release
74 |         id: upload-release
75 |         uses: svenstaro/upload-release-action@2.6.1
76 |         continue-on-error: true
77 |         with:
78 |           file: ./dist/*.whl
79 |           tag: ${{ matrix.releasetag }}
80 |           file_glob: true
81 |           make_latest: false
82 |           overwrite: true
83 |         
84 |       - uses: actions/upload-artifact@v3
85 |         if: steps.upload-release.outcome == 'failure'
86 |         with:
87 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
88 |           path: ./dist/*.whl
89 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.73.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.73
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: v0.1.73
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl
87 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.74.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.74
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: v0.1.74
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl
87 | 


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.76.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.76
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: v0.1.76
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl


--------------------------------------------------------------------------------
/old_workflows/build-wheels-0.1.77.yml:
--------------------------------------------------------------------------------
 1 | name: Build Wheels 0.1.77
 2 | 
 3 | on:
 4 |   workflow_dispatch:
 5 |   workflow_call:
 6 | 
 7 | permissions:
 8 |   contents: write
 9 | 
10 | jobs:
11 |   build_wheels:
12 |     name: ${{ matrix.os }} ${{ matrix.pyver }} ${{ matrix.cuda }} ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
13 |     runs-on: ${{ matrix.os }}
14 |     strategy:
15 |       matrix:
16 |         os: [ubuntu-20.04, windows-latest]
17 |         pyver: ["3.7", "3.8", "3.9", "3.10", "3.11"]
18 |         cuda: ["11.6.2", "11.7.1", "11.8.0", "12.0.1", "12.1.1", "12.2.0"]
19 |         releasetag: ["AVX","wheels","AVX512"]
20 |     defaults:
21 |       run:
22 |         shell: pwsh
23 |     env:
24 |       CUDAVER: ${{ matrix.cuda }}
25 |       AVXVER: ${{ matrix.releasetag }}
26 | 
27 |     steps:
28 |       - uses: actions/checkout@v3
29 |         with:
30 |           repository: 'abetlen/llama-cpp-python'
31 |           ref: v0.1.77
32 |           submodules: 'recursive'
33 |           
34 |       - uses: actions/setup-python@v3
35 |         with:
36 |           python-version: ${{ matrix.pyver }}
37 |         
38 |       - name: Setup Mamba
39 |         uses: conda-incubator/setup-miniconda@v2.2.0
40 |         with:
41 |           activate-environment: "build"
42 |           python-version: ${{ matrix.pyver }}
43 |           miniforge-variant: Mambaforge
44 |           miniforge-version: latest
45 |           use-mamba: true
46 |           add-pip-as-python-dependency: true
47 |           auto-activate-base: false
48 |           
49 |       - name: Install Dependencies
50 |         run: |
51 |           $cudaVersion = $env:CUDAVER
52 |           $cudaChannels = ''
53 |           $cudaNum = [int]$cudaVersion.substring($cudaVersion.LastIndexOf('.')+1)
54 |           while ($cudaNum -ge 0) { $cudaChannels += '-c nvidia/label/cuda-' + $cudaVersion.Remove($cudaVersion.LastIndexOf('.')+1) + $cudaNum + ' '; $cudaNum-- }
55 |           mamba install -y 'cuda' $cudaChannels.TrimEnd().Split()
56 |           python -m pip install build wheel
57 |           
58 |       - name: Build Wheel
59 |         run: |
60 |           $cudaVersion = $env:CUDAVER.Remove($env:CUDAVER.LastIndexOf('.')).Replace('.','')
61 |           $env:CUDA_PATH = $env:CONDA_PREFIX
62 |           $env:CUDA_HOME = $env:CONDA_PREFIX
63 |           if ($IsLinux) {$env:LD_LIBRARY_PATH = $env:CONDA_PREFIX + '/lib:' + $env:LD_LIBRARY_PATH}
64 |           $env:VERBOSE = '1'
65 |           $env:FORCE_CMAKE = '1'
66 |           $env:CMAKE_ARGS = '-DLLAMA_CUBLAS=on -DCMAKE_CUDA_ARCHITECTURES=all'
67 |           if ($env:AVXVER -eq 'AVX') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX2=off'}
68 |           if ($env:AVXVER -eq 'AVX512') {$env:CMAKE_ARGS = $env:CMAKE_ARGS + ' -DLLAMA_AVX512=on'}
69 |           python -m build --wheel -C--build-option=egg_info "-C--build-option=--tag-build=+cu$cudaVersion"
70 |           
71 |       - name: Upload files to a GitHub release
72 |         id: upload-release
73 |         uses: svenstaro/upload-release-action@2.6.1
74 |         continue-on-error: true
75 |         with:
76 |           file: ./dist/*.whl
77 |           tag: ${{ matrix.releasetag }}
78 |           file_glob: true
79 |           make_latest: false
80 |           overwrite: true
81 |         
82 |       - uses: actions/upload-artifact@v3
83 |         if: steps.upload-release.outcome == 'failure'
84 |         with:
85 |           name: ${{ matrix.releasetag == 'wheels' && 'AVX2' || matrix.releasetag }}
86 |           path: ./dist/*.whl


--------------------------------------------------------------------------------
/workflows.md:
--------------------------------------------------------------------------------
 1 | All workflows are configured to accept a llama-cpp-python release tag to build a specific version of the package.  
 2 | For the most part, they are written to account for changes in every version since 0.1.62.
 3 | 
 4 | Primary workflows used for new llama-cpp-python releases
 5 | ----
 6 | - `build-wheels.yml`
 7 |   - This workflow will build around 192 wheels for various CUDA, Python and CPU configurations. After this, it will call the `build-wheels-cpu.yml` workflow.
 8 | - `build-wheels-full-release.yml`
 9 |   - This workflow calls these workflows in order: `build-wheels.yml build-wheels-oobabooga.yml build-wheels-rocm-full.yml build-wheels-macos.yml`
10 |   - Somewhere around 370 wheels are produced in total, last I checked. This number will likely increase as additional builds, such as MacOS Metal, are eventually included.
11 | - `build-wheels-prioritized-release.yml`
12 |   - This workflow is much like `build-wheels-full-release.yml`, except `build-wheels.yml` and `build-wheels-oobabooga.yml` are incorporated into the workflow instead of being called due to minor modifications.
13 |   - This workflow is configured to build the wheels used by [text-generation-webui](https://github.com/oobabooga/text-generation-webui) first. This is because the long runtime of the workflow (currently 5 - 6 hours) was causing significant delays in updating the project.
14 | - `build-wheels-cpu.yml`
15 |   - This workflow builds CPU-only wheels for all of the CPU configurations supported by the other workflows.
16 |   - It was made because the wheels in the main repo are only built to support the default configuration of `AVX2`.
17 | 
18 | ~~These workflows, and their dependents, were recently optimized to significantly reduce run times from 6 hours for the longest down to around 2 hours.~~  
19 | Workflow optimizations have been made incompatible with llama-cpp-python 0.2.X+ due to abetlen switching the build backend to one that does not support modifications of the build process.  
20 | Copies of the optimized workflows can be found in the `old_workflows` directory.
21 | 
22 | Renamed package workflows
23 | ----
24 | These workflows produced renamed packages under different namespaces to allow for simultaneous installation with the main package.
25 | - `build-wheels-oobabooga*.yml`
26 |   - This workflow builds wheels with packages renamed to `llama_cpp_python_cuda`.
27 |   - These wheels are to allow applications to simultaneously support both CPU and CUDA builds of llama-cpp-python.
28 |   - As the name implies, this was made for text-generation-webui.
29 | - `build-wheels-ggml*.yml`
30 |   - This workflow was made to produce wheels for llama-cpp-python 0.1.78 under the name of `llama_cpp_python_ggml`.
31 |   - This allows applications to maintain support for GGML while updating to newer versions of llama-cpp-python.
32 |   - Intended to be a temporary measure until more models are converted to GGUF.
33 | 
34 | Configuration-specific workflows
35 | ----
36 | - `*-basic.yml`
37 | - `*-avx.yml`
38 | 
39 | These are copies of other workflows with build matrices limited to specific configurations.  
40 | For the most part, I made these to rebuild previous versions of llama-cpp-python as needed to support new configurations that were added to the main workflows.
41 | 
42 | Batch build workflows
43 | ----
44 | - `build-wheels-batch-*.yml`
45 | 
46 | These workflows accept a comma-separated string of llama-cpp-python release tags.  
47 | They then use Powershell to parse the input and construct a JSON string that is used to form a job matrix.  
48 | Associated workflows are then called as needed to build each version.  
49 | Only one workflow is executed at a time due to the large number of jobs that can be generated.
50 | 
51 | Experimental workflows used for more specialized builds
52 | ----
53 | - `build-wheel-rocm.yml`
54 |   - This workflow builds Linux and Windows wheels for AMD GPUs using ROCm.
55 |   - Linux wheels are built using these ROCm versions: `5.4.2  5.5  5.6.1`
56 |   - Currently considered experimental until someone with an AMD GPU can confirm if the resulting wheels work.
57 | - `build-wheels-oobabooga-rocm.yml`
58 |   - This workflow is much like the previous. It additionally builds `llama_cpp_python_cuda` wheels.
59 | - `build-wheels-rocm-full.yml`
60 |   - This workflow is essentially a combination of the previous 2.
61 | - `build-wheels-macos.yml`
62 |   - This workflow builds wheels with MacOS Metal support for MacOS 11, 12 and 13.
63 |   - Builds separate wheels for Intel and Apple Silicon CPU support.
64 |   - Is currently experimental and may not produce functional Metal wheels. I do not have a Mac to test with, so I can only go off of build logs.
65 | 
66 | Utility workflows
67 | ----
68 | - `deploy-index.yml`
69 |   - This workflow is for deploying the package index to GitHub Pages.
70 |   - It is configured to run automatically when the index's html files are altered.
71 | - `build-wheels-test.yml`
72 |   - This workflow is entirely for testing new workflow code and is changed frequently as needed.
73 | 


--------------------------------------------------------------------------------