diff --git a/.github/workflows/pr-test.yml b/.github/workflows/pr-test.yml index d474d7425..710250b2e 100644 --- a/.github/workflows/pr-test.yml +++ b/.github/workflows/pr-test.yml @@ -37,7 +37,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -59,7 +59,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -81,7 +81,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -92,7 +92,7 @@ jobs: python3 test_data_parallelism.py - name: Evaluate MLA accuracy (TP=2) - timeout-minutes: 10 + timeout-minutes: 20 run: | cd test/srt python3 test_mla.py @@ -120,7 +120,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -157,7 +157,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -188,7 +188,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -219,7 +219,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh @@ -243,7 +243,7 @@ jobs: - name: Install dependencies env: - FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu121/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu121/torch2.4/flashinfer' }} + FLASHINFER_REPO: ${{ inputs.version == 'nightly' && 'https://flashinfer.ai/whl/nightly/cu124/torch2.4/flashinfer' || 'https://flashinfer.ai/whl/cu124/torch2.4/flashinfer' }} run: | bash scripts/ci_install_dependency.sh diff --git a/docs/start/install.md b/docs/start/install.md index c297d1073..5de126401 100644 --- a/docs/start/install.md +++ b/docs/start/install.md @@ -5,7 +5,7 @@ You can install SGLang using any of the methods below. ## Method 1: With pip ``` pip install --upgrade pip -pip install "sglang[all]" --find-links https://flashinfer.ai/whl/cu121/torch2.4/flashinfer/ +pip install "sglang[all]" --find-links https://flashinfer.ai/whl/cu124/torch2.4/flashinfer/ ``` Note: Please check the [FlashInfer installation doc](https://docs.flashinfer.ai/installation.html) to install the proper version according to your PyTorch and CUDA versions. @@ -17,7 +17,7 @@ git clone -b v0.4.1.post3 https://github.com/sgl-project/sglang.git cd sglang pip install --upgrade pip -pip install -e "python[all]" --find-links https://flashinfer.ai/whl/cu121/torch2.4/flashinfer/ +pip install -e "python[all]" --find-links https://flashinfer.ai/whl/cu124/torch2.4/flashinfer/ ``` Note: Please check the [FlashInfer installation doc](https://docs.flashinfer.ai/installation.html) to install the proper version according to your PyTorch and CUDA versions. diff --git a/python/sglang/README.md b/python/sglang/README.md index c0a61e236..6221cdb2c 100644 --- a/python/sglang/README.md +++ b/python/sglang/README.md @@ -11,5 +11,5 @@ - `check_env.py`: Check the environment variables. - `global_config.py`: The global configs and constants. - `launch_server.py`: The entry point for launching the local server. -- `llama3_eval.py`: Llama 3.1 evaluation with meta-llama dataset. +- `llama3_eval.py`: Evaluation of Llama 3.1 using the Meta Llama dataset. - `utils.py`: Common utilities. diff --git a/scripts/ci_install_dependency.sh b/scripts/ci_install_dependency.sh index 5fa202a32..26c34879e 100755 --- a/scripts/ci_install_dependency.sh +++ b/scripts/ci_install_dependency.sh @@ -4,13 +4,13 @@ set -euxo pipefail # Install the dependency in CI. # Use repo from environment variable, passed from GitHub Actions -FLASHINFER_REPO="${FLASHINFER_REPO:-https://flashinfer.ai/whl/cu121/torch2.4/flashinfer}" +FLASHINFER_REPO="${FLASHINFER_REPO:-https://flashinfer.ai/whl/cu124/torch2.4/flashinfer}" SCRIPT_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )" bash "${SCRIPT_DIR}/killall_sglang.sh" pip install --upgrade pip -pip install -e "python[all]" --find-links https://flashinfer.ai/whl/cu121/torch2.4/flashinfer/ +pip install -e "python[all]" --find-links https://flashinfer.ai/whl/cu124/torch2.4/flashinfer/ # Force reinstall flashinfer pip install flashinfer==0.1.6 --find-links ${FLASHINFER_REPO} --force-reinstall --no-deps