feat: sglang update to 0.5.8 (#5655)

container/Dockerfile.sglang

@@ -42,7 +42,7 @@ ARG ENABLE_MEDIA_FFMPEG
 ARG CARGO_BUILD_JOBS
 
 ARG RUNTIME_IMAGE="lmsysorg/sglang"
-ARG RUNTIME_IMAGE_TAG="v0.5.7-runtime"
+ARG RUNTIME_IMAGE_TAG="v0.5.8-runtime"
 
 # SCCACHE configuration
 ARG USE_SCCACHE
@@ -544,7 +544,7 @@ RUN --mount=type=bind,source=.,target=/mnt/local_src \
         pip install --no-cache-dir --break-system-packages --force-reinstall --no-deps \
             nvidia-nccl-cu12==2.28.3 \
             nvidia-cudnn-cu12==9.16.0.29 \
-            nvidia-cutlass-dsl==4.3.0; \
+            nvidia-cutlass-dsl==4.3.5; \
     elif [ "$CUDA_MAJOR" = "13" ]; then \
         # CUDA 13: Install CuDNN for PyTorch 2.9.1 compatibility
         pip install --no-cache-dir --break-system-packages --force-reinstall --no-deps \

container/deps/requirements.txt

@@ -54,7 +54,7 @@ tensorboardX==2.6.2.2
 # Transformers version constraint for container builds
 # - vLLM 0.11.0: >=4.55.2, vLLM 0.11.2: >=4.56.0,<5
 # - TensorRT-LLM 1.2.0rc6.post2: ==4.57.1
-# - SGLang 0.5.7: ==4.57.1
+# - SGLang 0.5.8: ==4.57.1
 # Using >=4.56.0 to satisfy all frameworks
 transformers>=4.56.0
 types-aiofiles==25.1.0.20251011

docs/reference/support-matrix.md

@@ -17,7 +17,7 @@ The following table shows the backend framework versions included with each Dyna
 | **Dependency** | **main (ToT)** | **v0.8.1.post1** | **v0.8.1 (latest)** | **v0.8.0** | **v0.7.1** | **v0.7.0.post1** | **v0.7.0** |
 | :------------- | :------------- | :--------------- | :------------------ | :--------- | :--------- | :--------------- | :--------- |
 | vLLM           | `0.14.0`       | `0.12.0`         | `0.12.0`            | `0.12.0`   | `0.11.0`   | `0.11.0`         | `0.11.0`   |
-| SGLang         | `0.5.7`        | `0.5.6.post2`    | `0.5.6.post2`       | `0.5.6.post2` | `0.5.3.post4` | `0.5.3.post4` | `0.5.3.post4` |
+| SGLang         | `0.5.8`        | `0.5.6.post2`    | `0.5.6.post2`       | `0.5.6.post2` | `0.5.3.post4` | `0.5.3.post4` | `0.5.3.post4` |
 | TensorRT-LLM   | `1.2.0rc6.post2` | `1.2.0rc6.post2` | `1.2.0rc6.post1`  | `1.2.0rc6.post1` | `1.2.0rc3` | `1.2.0rc3`     | `1.2.0rc2` |
 | NIXL           | `0.9.0`        | `0.8.0`          | `0.8.0`             | `0.8.0`    | `0.8.0`    | `0.8.0`          | `0.8.0`    |
 

pyproject.toml

@@ -61,7 +61,7 @@ vllm = [
 
 sglang = [
     "uvloop",
-    "sglang==0.5.7",
+    "sglang==0.5.8",
     "nixl[cu12]<=0.9.0",
     "cupy-cuda12x>=13.0.0",
 ]