software-engineering/spark-vllm-docker
2
0
Fork 0
Code Issues Actions Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Eugene Rakhmatulin
2025-12-15 09:51:49 -08:00
parent 0606b1b984
commit 79f6a204d1
1 changed files with 19 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files

47
README.md
View File

@@ -12,6 +12,12 @@ The Dockerfile builds from the main branch of VLLM, so depending on when you run
## CHANGELOG
### 2025-12-15
Updated `build-and-copy.sh` flags:
- Renamed `--triton-sha` to `--triton-ref` to support branches and tags in addition to commit SHAs.
- Added `--vllm-ref <ref>`: Specify vLLM commit SHA, branch or tag (defaults to `main`).
### 2025-12-14
Converted to multi-stage Docker build with improved build times and reduced final image size. The builder stage is now separate from the runtime stage, excluding unnecessary build tools from the final image.
@@ -45,35 +51,19 @@ Applied patch to enable FastSafeTensors in cluster configuration (EXPERIMENTAL)
## 1\. Building the Docker Image
### Building Manually
The Dockerfile includes specific **Build Arguments** to allow you to selectively rebuild layers (e.g., update the vLLM source code without re-downloading PyTorch).
Using a provided build script is recommended, but if you want to build using `docker build` command, here are the supported build arguments:
### Option A: Standard Build (First Time)
| Argument | Default | Description |
| :--- | :--- | :--- |
| `CACHEBUST_DEPS` | `1` | Change this to force a re-download of PyTorch, FlashInfer, and system dependencies. |
| `CACHEBUST_VLLM` | `1` | Change this to force a fresh git clone and rebuild of vLLM source code. |
| `TRITON_REF` | `v3.5.1` | Triton commit SHA, branch, or tag to build. |
| `VLLM_REF` | `main` | vLLM commit SHA, branch, or tag to build. |
```bash
docker build -t vllm-node .
```
### Option B: Fast Rebuild (Update vLLM Source Only)
Use this if you want to pull the latest code from GitHub but keep the heavy dependencies (Torch, FlashInfer, system deps) cached.
```bash
docker build \
--build-arg CACHEBUST_VLLM=$(date +%s) \
-t vllm-node .
```
### Option C: Full Rebuild (Update All Dependencies)
Use this to force a re-download of PyTorch, FlashInfer, and system packages.
```bash
docker build \
--build-arg CACHEBUST_DEPS=$(date +%s) \
-t vllm-node .
```
### Option D: Using the Build Script (Recommended)
### Using the Build Script (Recommended)
The `build-and-copy.sh` script automates the build process and optionally copies the image to another node. This is the recommended method for building and deploying to multiple Spark nodes.
@@ -124,7 +114,7 @@ Using a different username:
**Build with specific Triton commit:**
```bash
./build-and-copy.sh --triton-sha abc123def456
./build-and-copy.sh --triton-ref abc123def456
```
**Copy existing image without rebuilding:**
@@ -140,7 +130,8 @@ Using a different username:
| `-t, --tag <tag>` | Image tag (default: 'vllm-node') |
| `--rebuild-deps` | Force rebuild all dependencies (sets CACHEBUST_DEPS) |
| `--rebuild-vllm` | Force rebuild vLLM source only (sets CACHEBUST_VLLM) |
| `--triton-sha <sha>` | Triton commit SHA (default: auto-detect latest main) |
| `--triton-ref <ref>` | Triton commit SHA, branch or tag (default: 'v3.5.1') |
| `--vllm-ref <ref>` | vLLM commit SHA, branch or tag (default: 'main') |
| `-h, --copy-to-host <host>` | Host address to copy the image to after building |
| `-u, --user <user>` | Username for SSH connection (default: current user) |
| `--no-build` | Skip building, only copy existing image (requires `--copy-to-host`) |