gpustack-runner 0.1.23.post1__tar.gz → 0.1.23.post2__tar.gz

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gpustack-runner
-Version: 0.1.23.post1
+Version: 0.1.23.post2
 Summary: GPUStack Runner is library for registering runnable accelerated backends and services in GPUStack.
 Project-URL: Homepage, https://github.com/gpustack/runner
 Project-URL: Bug Tracker, https://github.com/gpustack/gpustack/issues
@@ -58,8 +58,8 @@ The following table lists the supported accelerated backends and their correspon
 
 | CANN Version <br/> (Variant) | MindIE                   | vLLM                                                       | SGLang                 |
 |------------------------------|--------------------------|------------------------------------------------------------|------------------------|
-| 8.3 (A3/910C)                | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.6.post2`          |
-| 8.3 (910B)                   | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.6.post2`          |
+| 8.3 (A3/910C)                | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.7`, `0.5.6.post2` |
+| 8.3 (910B)                   | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.7`, `0.5.6.post2` |
 | 8.3 (310P)                   | `2.2.rc1`                |                                                            |                        |
 | 8.2 (A3/910C)                | `2.1.rc2`                | `0.10.2`, `0.10.1.1`                                       | `0.5.2`, `0.5.1.post3` |
 | 8.2 (910B)                   | `2.1.rc2`, ~~`2.1.rc1`~~ | `0.10.2`, `0.10.1.1`, <br/>`0.10.0`, `0.9.2`, <br/>`0.9.1` | `0.5.2`, `0.5.1.post3` |
@@ -88,12 +88,12 @@ The following table lists the supported accelerated backends and their correspon
 > - CUDA 12.6/12.4 supports Compute Capabilities:
     `7.5 8.0+PTX 8.9 9.0+PTX`.
 
-| CUDA Version <br/> (Variant) | vLLM                                                                                              | SGLang                                                        | VoxBox                 |
-|------------------------------|---------------------------------------------------------------------------------------------------|---------------------------------------------------------------|------------------------|
-| 12.9                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`                                                                 | `0.5.6.post2`                                                 |                        |
-| 12.8                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ | `0.5.6.post2`, `0.5.5.post3`, <br/>~~`0.5.5`~~, `0.5.4.post3` | `0.0.21`, ~~`0.0.20`~~ |
-| 12.6                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ | `0.5.6.post2`                                                 | `0.0.21`, ~~`0.0.20`~~ |
-| 12.4                         | ~~`0.11.0`~~, ~~`0.10.2`~~, <br/>~~`0.10.1.1`~~, ~~`0.10.0`~~                                     |                                                               | ~~`0.0.20`~~           |
+| CUDA Version <br/> (Variant) | vLLM                                                                                              | SGLang                                                                 | VoxBox                 |
+|------------------------------|---------------------------------------------------------------------------------------------------|------------------------------------------------------------------------|------------------------|
+| 12.9                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`                                                                 | `0.5.7`, `0.5.6.post2`                                                 |                        |
+| 12.8                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ | `0.5.7`, `0.5.6.post2`, `0.5.5.post3`, <br/>~~`0.5.5`~~, `0.5.4.post3` | `0.0.21`, ~~`0.0.20`~~ |
+| 12.6                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ |                                                                        | `0.0.21`, ~~`0.0.20`~~ |
+| 12.4                         | ~~`0.11.0`~~, ~~`0.10.2`~~, <br/>~~`0.10.1.1`~~, ~~`0.10.0`~~                                     |                                                                        | ~~`0.0.20`~~           |
 
 ### Hygon DTK
 
@@ -112,8 +112,8 @@ The following table lists the supported accelerated backends and their correspon
 
 > [!CAUTION]
 > Since v0.1.23:
-> Deprecated all services for ROCm 6.3.
-> Deprecated vLLM `0.11.0`.
+> - Deprecated all services for ROCm 6.3.
+> - Deprecated vLLM `0.11.0`.
 
 > [!NOTE]
 > - ROCm 7.0 supports LLVM targets:
@@ -128,11 +128,11 @@ The following table lists the supported accelerated backends and their correspon
 > - ROCm 6.4 SGLang supports `gfx942` only.
 > - ROCm 7.0 SGLang supports `gfx950` only.
 
-| ROCm Version <br/> (Variant) | vLLM                                            | SGLang                       |
-|------------------------------|-------------------------------------------------|------------------------------|
-| 7.0                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~ | `0.5.6.post2`                |
-| 6.4                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2`     | `0.5.6.post2`, `0.5.5.post3` |
-| 6.3                          | ~~`0.10.1.1`~~, ~~`0.10.0`~~                    |                              |
+| ROCm Version <br/> (Variant) | vLLM                                            | SGLang                                |
+|------------------------------|-------------------------------------------------|---------------------------------------|
+| 7.0                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~ | `0.5.7`, `0.5.6.post2`                |
+| 6.4                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2`     | `0.5.7`, `0.5.6.post2`, `0.5.5.post3` |
+| 6.3                          | ~~`0.10.1.1`~~, ~~`0.10.0`~~                    |                                       |
 
 ## Directory Structure
 

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/README.md

@@ -38,8 +38,8 @@ The following table lists the supported accelerated backends and their correspon
 
 | CANN Version <br/> (Variant) | MindIE                   | vLLM                                                       | SGLang                 |
 |------------------------------|--------------------------|------------------------------------------------------------|------------------------|
-| 8.3 (A3/910C)                | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.6.post2`          |
-| 8.3 (910B)                   | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.6.post2`          |
+| 8.3 (A3/910C)                | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.7`, `0.5.6.post2` |
+| 8.3 (910B)                   | `2.2.rc1`                | `0.13.0`, `0.12.0`, `0.11.0`                               | `0.5.7`, `0.5.6.post2` |
 | 8.3 (310P)                   | `2.2.rc1`                |                                                            |                        |
 | 8.2 (A3/910C)                | `2.1.rc2`                | `0.10.2`, `0.10.1.1`                                       | `0.5.2`, `0.5.1.post3` |
 | 8.2 (910B)                   | `2.1.rc2`, ~~`2.1.rc1`~~ | `0.10.2`, `0.10.1.1`, <br/>`0.10.0`, `0.9.2`, <br/>`0.9.1` | `0.5.2`, `0.5.1.post3` |
@@ -68,12 +68,12 @@ The following table lists the supported accelerated backends and their correspon
 > - CUDA 12.6/12.4 supports Compute Capabilities:
     `7.5 8.0+PTX 8.9 9.0+PTX`.
 
-| CUDA Version <br/> (Variant) | vLLM                                                                                              | SGLang                                                        | VoxBox                 |
-|------------------------------|---------------------------------------------------------------------------------------------------|---------------------------------------------------------------|------------------------|
-| 12.9                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`                                                                 | `0.5.6.post2`                                                 |                        |
-| 12.8                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ | `0.5.6.post2`, `0.5.5.post3`, <br/>~~`0.5.5`~~, `0.5.4.post3` | `0.0.21`, ~~`0.0.20`~~ |
-| 12.6                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ | `0.5.6.post2`                                                 | `0.0.21`, ~~`0.0.20`~~ |
-| 12.4                         | ~~`0.11.0`~~, ~~`0.10.2`~~, <br/>~~`0.10.1.1`~~, ~~`0.10.0`~~                                     |                                                               | ~~`0.0.20`~~           |
+| CUDA Version <br/> (Variant) | vLLM                                                                                              | SGLang                                                                 | VoxBox                 |
+|------------------------------|---------------------------------------------------------------------------------------------------|------------------------------------------------------------------------|------------------------|
+| 12.9                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`                                                                 | `0.5.7`, `0.5.6.post2`                                                 |                        |
+| 12.8                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ | `0.5.7`, `0.5.6.post2`, `0.5.5.post3`, <br/>~~`0.5.5`~~, `0.5.4.post3` | `0.0.21`, ~~`0.0.20`~~ |
+| 12.6                         | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~, <br/>`0.10.2`, ~~`0.10.1.1`~~, <br/>~~`0.10.0`~~ |                                                                        | `0.0.21`, ~~`0.0.20`~~ |
+| 12.4                         | ~~`0.11.0`~~, ~~`0.10.2`~~, <br/>~~`0.10.1.1`~~, ~~`0.10.0`~~                                     |                                                                        | ~~`0.0.20`~~           |
 
 ### Hygon DTK
 
@@ -92,8 +92,8 @@ The following table lists the supported accelerated backends and their correspon
 
 > [!CAUTION]
 > Since v0.1.23:
-> Deprecated all services for ROCm 6.3.
-> Deprecated vLLM `0.11.0`.
+> - Deprecated all services for ROCm 6.3.
+> - Deprecated vLLM `0.11.0`.
 
 > [!NOTE]
 > - ROCm 7.0 supports LLVM targets:
@@ -108,11 +108,11 @@ The following table lists the supported accelerated backends and their correspon
 > - ROCm 6.4 SGLang supports `gfx942` only.
 > - ROCm 7.0 SGLang supports `gfx950` only.
 
-| ROCm Version <br/> (Variant) | vLLM                                            | SGLang                       |
-|------------------------------|-------------------------------------------------|------------------------------|
-| 7.0                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~ | `0.5.6.post2`                |
-| 6.4                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2`     | `0.5.6.post2`, `0.5.5.post3` |
-| 6.3                          | ~~`0.10.1.1`~~, ~~`0.10.0`~~                    |                              |
+| ROCm Version <br/> (Variant) | vLLM                                            | SGLang                                |
+|------------------------------|-------------------------------------------------|---------------------------------------|
+| 7.0                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, ~~`0.11.0`~~ | `0.5.7`, `0.5.6.post2`                |
+| 6.4                          | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2`     | `0.5.7`, `0.5.6.post2`, `0.5.5.post3` |
+| 6.3                          | ~~`0.10.1.1`~~, ~~`0.10.0`~~                    |                                       |
 
 ## Directory Structure
 

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/gpustack_runner/_version.py

@@ -27,8 +27,8 @@ version_tuple: VERSION_TUPLE
 __commit_id__: COMMIT_ID
 commit_id: COMMIT_ID
 
-__version__ = version = '0.1.23.post1'
-__version_tuple__ = version_tuple = (0, 1, 23, 'post1')
+__version__ = version = '0.1.23.post2'
+__version_tuple__ = version_tuple = (0, 1, 23, 'post2')
 try:
     from ._version_appendix import git_commit
     __commit_id__ = commit_id = git_commit

gpustack_runner-0.1.23.post2/gpustack_runner/_version_appendix.py

@@ -0,0 +1 @@
+git_commit = "6d771e2"

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/gpustack_runner/cmds/images.py

@@ -17,7 +17,7 @@ from typing import TYPE_CHECKING
 import requests
 from dataclasses_json import dataclass_json
 
-from gpustack_runner import BackendRunners, list_backend_runners
+from gpustack_runner import BackendRunners, envs, list_backend_runners
 
 from .__types__ import SubCommand
 
@@ -1094,6 +1094,9 @@ def list_images(**kwargs) -> list[PlatformedImage]:
             name = img.name
             if not name:
                 continue
+            if namespace := envs.GPUSTACK_RUNNER_DEFAULT_IMAGE_NAMESPACE:
+                name = name.replace("gpustack/", f"{namespace}/")
+                img.name = name
             if name not in image_names_index:
                 image_names_index[name] = len(images)
                 images.append(img)

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/gpustack_runner/runner.py.json

@@ -21,6 +21,17 @@
     "docker_image": "gpustack/runner:cann8.3-a3-mindie2.2.rc1",
     "deprecated": false
   },
+  {
+    "backend": "cann",
+    "backend_version": "8.3",
+    "original_backend_version": "8.3.rc2",
+    "backend_variant": "a3",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/arm64",
+    "docker_image": "gpustack/runner:cann8.3-a3-sglang0.5.7",
+    "deprecated": false
+  },
   {
     "backend": "cann",
     "backend_version": "8.3",
@@ -230,6 +241,17 @@
     "docker_image": "gpustack/runner:cann8.3-910b-mindie2.2.rc1",
     "deprecated": false
   },
+  {
+    "backend": "cann",
+    "backend_version": "8.3",
+    "original_backend_version": "8.3.rc2",
+    "backend_variant": "910b",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/arm64",
+    "docker_image": "gpustack/runner:cann8.3-910b-sglang0.5.7",
+    "deprecated": false
+  },
   {
     "backend": "cann",
     "backend_version": "8.3",
@@ -626,6 +648,28 @@
     "docker_image": "gpustack/runner:corex4.2-vllm0.8.3",
     "deprecated": false
   },
+  {
+    "backend": "cuda",
+    "backend_version": "12.9",
+    "original_backend_version": "12.9.1",
+    "backend_variant": "",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/amd64",
+    "docker_image": "gpustack/runner:cuda12.9-sglang0.5.7",
+    "deprecated": false
+  },
+  {
+    "backend": "cuda",
+    "backend_version": "12.9",
+    "original_backend_version": "12.9.1",
+    "backend_variant": "",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/arm64",
+    "docker_image": "gpustack/runner:cuda12.9-sglang0.5.7",
+    "deprecated": false
+  },
   {
     "backend": "cuda",
     "backend_version": "12.9",
@@ -714,6 +758,28 @@
     "docker_image": "gpustack/runner:cuda12.9-vllm0.11.2",
     "deprecated": false
   },
+  {
+    "backend": "cuda",
+    "backend_version": "12.8",
+    "original_backend_version": "12.8.1",
+    "backend_variant": "",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/amd64",
+    "docker_image": "gpustack/runner:cuda12.8-sglang0.5.7",
+    "deprecated": false
+  },
+  {
+    "backend": "cuda",
+    "backend_version": "12.8",
+    "original_backend_version": "12.8.1",
+    "backend_variant": "",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/arm64",
+    "docker_image": "gpustack/runner:cuda12.8-sglang0.5.7",
+    "deprecated": false
+  },
   {
     "backend": "cuda",
     "backend_version": "12.8",
@@ -1341,6 +1407,17 @@
     "docker_image": "gpustack/runner:maca3.0-vllm0.9.1",
     "deprecated": false
   },
+  {
+    "backend": "rocm",
+    "backend_version": "7.0",
+    "original_backend_version": "7.0.2",
+    "backend_variant": "",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/amd64",
+    "docker_image": "gpustack/runner:rocm7.0-sglang0.5.7",
+    "deprecated": false
+  },
   {
     "backend": "rocm",
     "backend_version": "7.0",
@@ -1396,6 +1473,17 @@
     "docker_image": "gpustack/runner:rocm7.0-vllm0.11.0",
     "deprecated": true
   },
+  {
+    "backend": "rocm",
+    "backend_version": "6.4",
+    "original_backend_version": "6.4.4",
+    "backend_variant": "",
+    "service": "sglang",
+    "service_version": "0.5.7",
+    "platform": "linux/amd64",
+    "docker_image": "gpustack/runner:rocm6.4-sglang0.5.7",
+    "deprecated": false
+  },
   {
     "backend": "rocm",
     "backend_version": "6.4",

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/pack/cann/Dockerfile

@@ -58,9 +58,9 @@ ARG VLLM_ASCEND_VERSION
 ARG VLLM_TORCH_VERSION=2.8.0
 ARG VLLM_MOONCAKE_VERSION=0.3.7.post2
 ARG SGLANG_BASE_IMAGE=gpustack/runner:cann${CANN_VERSION}-${CANN_ARCHS}-python${PYTHON_VERSION}
-ARG SGLANG_VERSION=0.5.6.post2
+ARG SGLANG_VERSION=0.5.7
 ARG SGLANG_TORCH_VERSION=2.8.0
-ARG SGLANG_KERNEL_VERSION=20251209
+ARG SGLANG_KERNEL_VERSION=2025.12.31
 ARG SGLANG_VLLM_VERSION=0.9.2
 
 #
@@ -1023,7 +1023,7 @@ EOT
             -r /tmp/requirements.txt
     fi
     uv pip install --verbose \
-        https://sglang-ascend.obs.cn-east-3.myhuaweicloud.com/sglang/torch_npu/torch_npu-2.8.0.post2.dev20251113-cp311-cp311-manylinux_2_28_$(uname -m).whl
+        https://sglang-ascend.obs.cn-east-3.myhuaweicloud.com/sglang/torch_npu/torch_npu-2.8.0.post2.dev20251224-cp311-cp311-manylinux_2_28_$(uname -m).whl
     uv pip install \
         numpy==1.26.4 scipy==1.13.1
 

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/pack/cuda/Dockerfile

@@ -115,9 +115,9 @@ ARG VLLM_FLASHATTENTION_VERSION=2.8.3
 ARG VLLM_LMCACHE_VERSION=0.3.11
 ARG VLLM_MOONCAKE_VERSION=0.3.7.post2
 ARG SGLANG_BASE_IMAGE=vllm
-ARG SGLANG_VERSION=0.5.6.post2
+ARG SGLANG_VERSION=0.5.7
 ARG SGLANG_BUILD_BASE_IMAGE=vllm-build
-ARG SGLANG_KERNEL_VERSION=0.3.19
+ARG SGLANG_KERNEL_VERSION=0.3.20
 
 #
 # Stage Bake Runtime
@@ -1264,11 +1264,6 @@ RUN --mount=type=bind,from=vllm-build-vllm,source=/,target=/vllm,rw <<EOF
     uv pip install \
         /vllm/workspace/*.whl
 
-    # Uninstall
-    ## Remove Run:AI Model Streamer to avoid peak memory usage in model loading.
-    uv pip uninstall \
-        runai-model-streamer || true
-
     # Cleanup
     rm -rf /var/tmp/* \
         && rm -rf /tmp/*

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/pack/matrix.yaml

@@ -96,8 +96,8 @@ rules:
       - "sglang"
     args:
       - "CUDA_VERSION=12.8.1"
-      - "SGLANG_BASE_IMAGE=gpustack/runner:cuda12.8-vllm0.12.0"
-      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:cuda12.8-vllm0.12.0"
+      - "SGLANG_BASE_IMAGE=gpustack/runner:cuda12.8-vllm0.13.0"
+      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:cuda12.8-vllm0.13.0"
   ## NVIDIA CUDA 12.9.1, using PyTorch +cu129.
   ##
   - backend: "cuda"
@@ -106,8 +106,8 @@ rules:
       - "sglang"
     args:
       - "CUDA_VERSION=12.9.1"
-      - "SGLANG_BASE_IMAGE=gpustack/runner:cuda12.9-vllm0.12.0"
-      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:cuda12.9-vllm0.12.0"
+      - "SGLANG_BASE_IMAGE=gpustack/runner:cuda12.9-vllm0.13.0"
+      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:cuda12.9-vllm0.13.0"
 
   #
   # Hygon DTK
@@ -152,8 +152,8 @@ rules:
     args:
       - "ROCM_VERSION=7.0.2"
       - "VLLM_TORCH_SOURCE=radeon"
-      - "SGLANG_BASE_IMAGE=gpustack/runner:rocm7.0-vllm0.12.0"
-      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:rocm7.0-vllm0.12.0"
+      - "SGLANG_BASE_IMAGE=gpustack/runner:rocm7.0-vllm0.13.0"
+      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:rocm7.0-vllm0.13.0"
   ##
   - backend: "rocm"
     services:
@@ -164,5 +164,5 @@ rules:
     args:
       - "ROCM_VERSION=6.4.4"
       - "VLLM_AITER_VERSION=0.1.7.post2"
-      - "SGLANG_BASE_IMAGE=gpustack/runner:rocm6.4-vllm0.12.0"
-      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:rocm6.4-vllm0.12.0"
+      - "SGLANG_BASE_IMAGE=gpustack/runner:rocm6.4-vllm0.13.0"
+      - "SGLANG_BUILD_BASE_IMAGE=gpustack/runner:rocm6.4-vllm0.13.0"

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/pack/rocm/Dockerfile

@@ -83,7 +83,7 @@ ARG VLLM_AITER_VERSION=0.1.7.post5
 ARG VLLM_LMCACHE_VERSION=0.3.11
 ARG VLLM_MOONCAKE_VERSION=0.3.7.post2
 ARG SGLANG_BASE_IMAGE=vllm
-ARG SGLANG_VERSION=0.5.6.post2
+ARG SGLANG_VERSION=0.5.7
 ARG SGLANG_BUILD_BASE_IMAGE=vllm-build
 
 # Stage Bake Runtime
@@ -1349,12 +1349,22 @@ RUN --mount=type=bind,from=sglang-build-sglang,source=/,target=/sglang,rw \
 
     # Install
     uv pip install \
-        "$(ls /sglang/workspace/sglang-*.whl)[all_hip]"
+        "$(ls /sglang/workspace/sglang-*.whl)[srt_hip]"
+    if [[ "${TARGETARCH}" == "amd64" ]]; then
+        # Install SGLang Diffusion Extension
+        uv pip install \
+            "$(ls /sglang/workspace/sglang-*.whl)[diffusion]"
+    fi
     uv pip install \
         "$(ls /sglang/workspace/sgl_kernel-*.whl)"
     uv pip install --force-reinstall \
         /sglangrouter/workspace/*.whl
 
+    # Uninstall
+    ## Remove Run:AI Model Streamer to avoid peak memory usage in model loading.
+    uv pip uninstall \
+        runai-model-streamer || true
+
     # Cleanup
     rm -rf /var/tmp/* \
         && rm -rf /tmp/*

{gpustack_runner-0.1.23.post1 → gpustack_runner-0.1.23.post2}/tests/gpustack_runner/fixtures/test_list_runners_by_backend.json

@@ -27,6 +27,17 @@
         "docker_image": "gpustack/runner:cann8.3-a3-mindie2.2.rc1",
         "deprecated": false
       },
+      {
+        "backend": "cann",
+        "backend_version": "8.3",
+        "original_backend_version": "8.3.rc2",
+        "backend_variant": "a3",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/arm64",
+        "docker_image": "gpustack/runner:cann8.3-a3-sglang0.5.7",
+        "deprecated": false
+      },
       {
         "backend": "cann",
         "backend_version": "8.3",
@@ -236,6 +247,17 @@
         "docker_image": "gpustack/runner:cann8.3-910b-mindie2.2.rc1",
         "deprecated": false
       },
+      {
+        "backend": "cann",
+        "backend_version": "8.3",
+        "original_backend_version": "8.3.rc2",
+        "backend_variant": "910b",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/arm64",
+        "docker_image": "gpustack/runner:cann8.3-910b-sglang0.5.7",
+        "deprecated": false
+      },
       {
         "backend": "cann",
         "backend_version": "8.3",
@@ -648,6 +670,28 @@
       "backend": "cuda"
     },
     [
+      {
+        "backend": "cuda",
+        "backend_version": "12.9",
+        "original_backend_version": "12.9.1",
+        "backend_variant": "",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/amd64",
+        "docker_image": "gpustack/runner:cuda12.9-sglang0.5.7",
+        "deprecated": false
+      },
+      {
+        "backend": "cuda",
+        "backend_version": "12.9",
+        "original_backend_version": "12.9.1",
+        "backend_variant": "",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/arm64",
+        "docker_image": "gpustack/runner:cuda12.9-sglang0.5.7",
+        "deprecated": false
+      },
       {
         "backend": "cuda",
         "backend_version": "12.9",
@@ -736,6 +780,28 @@
         "docker_image": "gpustack/runner:cuda12.9-vllm0.11.2",
         "deprecated": false
       },
+      {
+        "backend": "cuda",
+        "backend_version": "12.8",
+        "original_backend_version": "12.8.1",
+        "backend_variant": "",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/amd64",
+        "docker_image": "gpustack/runner:cuda12.8-sglang0.5.7",
+        "deprecated": false
+      },
+      {
+        "backend": "cuda",
+        "backend_version": "12.8",
+        "original_backend_version": "12.8.1",
+        "backend_variant": "",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/arm64",
+        "docker_image": "gpustack/runner:cuda12.8-sglang0.5.7",
+        "deprecated": false
+      },
       {
         "backend": "cuda",
         "backend_version": "12.8",
@@ -1387,6 +1453,17 @@
       "backend": "rocm"
     },
     [
+      {
+        "backend": "rocm",
+        "backend_version": "7.0",
+        "original_backend_version": "7.0.2",
+        "backend_variant": "",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/amd64",
+        "docker_image": "gpustack/runner:rocm7.0-sglang0.5.7",
+        "deprecated": false
+      },
       {
         "backend": "rocm",
         "backend_version": "7.0",
@@ -1442,6 +1519,17 @@
         "docker_image": "gpustack/runner:rocm7.0-vllm0.11.0",
         "deprecated": true
       },
+      {
+        "backend": "rocm",
+        "backend_version": "6.4",
+        "original_backend_version": "6.4.4",
+        "backend_variant": "",
+        "service": "sglang",
+        "service_version": "0.5.7",
+        "platform": "linux/amd64",
+        "docker_image": "gpustack/runner:rocm6.4-sglang0.5.7",
+        "deprecated": false
+      },
       {
         "backend": "rocm",
         "backend_version": "6.4",

gpustack_runner-0.1.23.post1/gpustack_runner/_version_appendix.py

@@ -1 +0,0 @@
-git_commit = "b4fa7a6"