Update default branch post-25.04 (#8188)

mc-nv · dmitry-tokarev-nv · richardhuo-nv · web-flow · commit 4fc0a5ed464f · 2025-05-07T12:05:14.000-07:00
Co-authored-by: Dmitry Tokarev &lt;dtokarev@nvidia.com&gt;
Co-authored-by: richardhuo-nv &lt;rihuo@nvidia.com&gt;
diff --git a/Dockerfile.sdk b/Dockerfile.sdk
@@ -29,7 +29,7 @@
 #
 
 # Base image on the minimum Triton container
-ARG BASE_IMAGE=nvcr.io/nvidia/tritonserver:25.03-py3-min
+ARG BASE_IMAGE=nvcr.io/nvidia/tritonserver:25.04-py3-min
 
 ARG TRITON_CLIENT_REPO_SUBDIR=clientrepo
 ARG TRITON_PA_REPO_SUBDIR=perfanalyzerrepo
diff --git a/README.md b/README.md
@@ -29,8 +29,8 @@
 
 >[!WARNING]
 >You are currently on the `main` branch which tracks under-development progress
->towards the next release. The current release is version [2.56.0](https://github.com/triton-inference-server/server/releases/latest)
->and corresponds to the 25.03 container release on NVIDIA GPU Cloud (NGC).
+>towards the next release. The current release is version [2.57.0](https://github.com/triton-inference-server/server/releases/latest)
+>and corresponds to the 25.04 container release on NVIDIA GPU Cloud (NGC).
 
 # Triton Inference Server
 
diff --git a/TRITON_VERSION b/TRITON_VERSION
@@ -1 +1 @@
-2.57.0dev
+2.58.0dev
diff --git a/build.py b/build.py
@@ -71,14 +71,14 @@
 #
 
 DEFAULT_TRITON_VERSION_MAP = {
-    "release_version": "2.57.0dev",
-    "triton_container_version": "25.04dev",
-    "upstream_container_version": "25.03",
+    "release_version": "2.58.0dev",
+    "triton_container_version": "25.05dev",
+    "upstream_container_version": "25.04",
     "ort_version": "1.21.0",
     "ort_openvino_version": "2025.0.0",
     "standalone_openvino_version": "2025.0.0",
     "dcgm_version": "3.3.6",
-    "vllm_version": "0.7.3",
+    "vllm_version": "0.8.1",
     "rhel_py_version": "3.12.3",
 }
 
diff --git a/deploy/aws/values.yaml b/deploy/aws/values.yaml
@@ -27,7 +27,7 @@
 replicaCount: 1
 
 image:
-  imageName: nvcr.io/nvidia/tritonserver:25.03-py3
+  imageName: nvcr.io/nvidia/tritonserver:25.04-py3
   pullPolicy: IfNotPresent
   modelRepositoryPath: s3://triton-inference-server-repository/model_repository
   numGpus: 1
diff --git a/deploy/fleetcommand/Chart.yaml b/deploy/fleetcommand/Chart.yaml
@@ -26,7 +26,7 @@
 
 apiVersion: v1
 # appVersion is the Triton version; update when changing release
-appVersion: "2.56.0"
+appVersion: "2.57.0"
 description: Triton Inference Server (Fleet Command)
 name: triton-inference-server
 # version is the Chart version; update when changing anything in the chart
diff --git a/deploy/fleetcommand/values.yaml b/deploy/fleetcommand/values.yaml
@@ -27,7 +27,7 @@
 replicaCount: 1
 
 image:
-  imageName: nvcr.io/nvidia/tritonserver:25.03-py3
+  imageName: nvcr.io/nvidia/tritonserver:25.04-py3
   pullPolicy: IfNotPresent
   numGpus: 1
   serverCommand: tritonserver
@@ -47,13 +47,13 @@ image:
     #
     # To set model control mode, uncomment and configure below
     # TODO: Fix the following url, it is invalid
-    # See https://github.com/triton-inference-server/server/blob/r25.03/docs/model_management.md
+    # See https://github.com/triton-inference-server/server/blob/r25.04/docs/user_guide/model_management.md
     #  for more details
     #- --model-control-mode=explicit|poll|none
     #
     # Additional server args
     #
-    # see https://github.com/triton-inference-server/server/blob/r25.03/README.md
+    # see https://github.com/triton-inference-server/server/blob/r25.04/README.md
     #  for more details
 
 service:
diff --git a/deploy/gcp/values.yaml b/deploy/gcp/values.yaml
@@ -27,7 +27,7 @@
 replicaCount: 1
 
 image:
-  imageName: nvcr.io/nvidia/tritonserver:25.03-py3
+  imageName: nvcr.io/nvidia/tritonserver:25.04-py3
   pullPolicy: IfNotPresent
   modelRepositoryPath: gs://triton-inference-server-repository/model_repository
   numGpus: 1
diff --git a/deploy/gke-marketplace-app/benchmark/perf-analyzer-script/triton_client.yaml b/deploy/gke-marketplace-app/benchmark/perf-analyzer-script/triton_client.yaml
@@ -33,7 +33,7 @@ metadata:
   namespace: default
 spec:
   containers:
-  - image: nvcr.io/nvidia/tritonserver:25.03-py3-sdk
+  - image: nvcr.io/nvidia/tritonserver:25.04-py3-sdk
     imagePullPolicy: Always
     name: nv-triton-client
     securityContext:
diff --git a/deploy/gke-marketplace-app/server-deployer/build_and_push.sh b/deploy/gke-marketplace-app/server-deployer/build_and_push.sh
@@ -27,9 +27,9 @@
 
 export REGISTRY=gcr.io/$(gcloud config get-value project | tr ':' '/')
 export APP_NAME=tritonserver
-export MAJOR_VERSION=2.56
-export MINOR_VERSION=2.56.0
-export NGC_VERSION=25.03-py3
+export MAJOR_VERSION=2.57
+export MINOR_VERSION=2.57.0
+export NGC_VERSION=25.04-py3
 
 docker pull nvcr.io/nvidia/$APP_NAME:$NGC_VERSION
 
diff --git a/deploy/gke-marketplace-app/server-deployer/chart/triton/Chart.yaml b/deploy/gke-marketplace-app/server-deployer/chart/triton/Chart.yaml
@@ -25,7 +25,7 @@
 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 apiVersion: v1
-appVersion: "2.56"
+appVersion: "2.57"
 description: Triton Inference Server
 name: triton-inference-server
-version: 2.56.0
+version: 2.57.0
diff --git a/deploy/gke-marketplace-app/server-deployer/chart/triton/values.yaml b/deploy/gke-marketplace-app/server-deployer/chart/triton/values.yaml
@@ -31,14 +31,14 @@ maxReplicaCount: 3
 tritonProtocol: HTTP
 # HPA GPU utilization autoscaling target
 HPATargetAverageValue: 85
-modelRepositoryPath: gs://triton_sample_models/25.03
-publishedVersion: '2.56.0'
+modelRepositoryPath: gs://triton_sample_models/25.04
+publishedVersion: '2.57.0'
 gcpMarketplace: true
 
 image:
   registry: gcr.io
   repository: nvidia-ngc-public/tritonserver
-  tag: 25.03-py3
+  tag: 25.04-py3
   pullPolicy: IfNotPresent
   # modify the model repository here to match your GCP storage bucket
   numGpus: 1
diff --git a/deploy/gke-marketplace-app/server-deployer/data-test/schema.yaml b/deploy/gke-marketplace-app/server-deployer/data-test/schema.yaml
@@ -27,7 +27,7 @@
 x-google-marketplace:
   schemaVersion: v2
   applicationApiVersion: v1beta1
-  publishedVersion: '2.56.0'
+  publishedVersion: '2.57.0'
   publishedVersionMetadata:
     releaseNote: >-
       Initial release.
diff --git a/deploy/gke-marketplace-app/server-deployer/schema.yaml b/deploy/gke-marketplace-app/server-deployer/schema.yaml
@@ -27,7 +27,7 @@
 x-google-marketplace:
   schemaVersion: v2
   applicationApiVersion: v1beta1
-  publishedVersion: '2.56.0'
+  publishedVersion: '2.57.0'
   publishedVersionMetadata:
     releaseNote: >-
       Initial release.
@@ -89,7 +89,7 @@ properties:
   modelRepositoryPath:
     type: string
     title: Bucket where models are stored. Please make sure the user/service account to create the GKE app has permission to this GCS bucket. Read Triton documentation on configs and formatting details, supporting TensorRT, TensorFlow, Pytorch, Onnx ... etc.
-    default: gs://triton_sample_models/25.03
+    default: gs://triton_sample_models/25.04
   image.ldPreloadPath:
     type: string
     title: Leave this empty by default. Triton allows users to create custom layers for backend such as TensorRT plugin or Tensorflow custom ops, the compiled shared library must be provided via LD_PRELOAD environment variable.
diff --git a/deploy/gke-marketplace-app/trt-engine/README.md b/deploy/gke-marketplace-app/trt-engine/README.md
@@ -33,7 +33,7 @@
 ```
 docker run --gpus all -it --network host \
     --shm-size=1g --ulimit memlock=-1 --ulimit stack=67108864 \
-    -v ~:/scripts nvcr.io/nvidia/tensorrt:25.03-py3
+    -v ~:/scripts nvcr.io/nvidia/tensorrt:25.04-py3
 
 pip install onnx six torch tf2onnx tensorflow
 
@@ -57,7 +57,7 @@ mkdir -p engines
 
 python3 builder.py -m models/fine-tuned/bert_tf_ckpt_large_qa_squad2_amp_128_v19.03.1/model.ckpt -o engines/bert_large_int8_bs1_s128.engine -b 1 -s 128 -c models/fine-tuned/bert_tf_ckpt_large_qa_squad2_amp_128_v19.03.1/ -v models/fine-tuned/bert_tf_ckpt_large_qa_squad2_amp_128_v19.03.1/vocab.txt --int8 --fp16 --strict --calib-num 1 -iln -imh
 
-gsutil cp bert_large_int8_bs1_s128.engine gs://triton_sample_models/25.03/bert/1/model.plan
+gsutil cp bert_large_int8_bs1_s128.engine gs://triton_sample_models/25.04/bert/1/model.plan
 ```
 
-For each Triton upgrade, container version used to generate the model, and the model path in GCS `gs://triton_sample_models/25.03/` should be updated accordingly with the correct version.
+For each Triton upgrade, container version used to generate the model, and the model path in GCS `gs://triton_sample_models/25.04/` should be updated accordingly with the correct version.
diff --git a/deploy/k8s-onprem/values.yaml b/deploy/k8s-onprem/values.yaml
@@ -30,7 +30,7 @@ tags:
   openshift: false
 
 image:
-  imageName: nvcr.io/nvidia/tritonserver:25.03-py3
+  imageName: nvcr.io/nvidia/tritonserver:25.04-py3
   pullPolicy: IfNotPresent
   modelRepositoryServer: < Replace with the IP Address of your file server >
   modelRepositoryPath: /srv/models
diff --git a/deploy/oci/values.yaml b/deploy/oci/values.yaml
@@ -27,7 +27,7 @@
 replicaCount: 1
 
 image:
-  imageName: nvcr.io/nvidia/tritonserver:25.03-py3
+  imageName: nvcr.io/nvidia/tritonserver:25.04-py3
   pullPolicy: IfNotPresent
   modelRepositoryPath: s3://https://<OCI_NAMESPACE>.compat.objectstorage.<OCI_REGION>.oraclecloud.com:443/triton-inference-server-repository
   numGpus: 1
diff --git a/docs/introduction/release_notes.md b/docs/introduction/release_notes.md
@@ -25,9 +25,9 @@
 # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 -->
-# [Triton Inference Server Release 25.03](https://docs.nvidia.com/deeplearning/triton-inference-server/release-notes/rel-25-03.html#rel-25-03)
+# [Triton Inference Server Release 25.04](https://docs.nvidia.com/deeplearning/triton-inference-server/release-notes/rel-25-04.html#rel-25-04)
 
-The Triton Inference Server container image, release 25.03, is available
+The Triton Inference Server container image, release 25.04, is available
 on [NGC](https://ngc.nvidia.com/catalog/containers/nvidia:tritonserver) and
 is open source
 on [GitHub](https://github.com/triton-inference-server/server). Release notes can
diff --git a/python/openai/README.md b/python/openai/README.md
@@ -51,7 +51,7 @@
 docker run -it --net=host --gpus all --rm \
   -v ${HOME}/.cache/huggingface:/root/.cache/huggingface \
   -e HF_TOKEN \
-  nvcr.io/nvidia/tritonserver:25.03-vllm-python-py3
+  nvcr.io/nvidia/tritonserver:25.04-vllm-python-py3
 ```
 
 2. Launch the OpenAI-compatible Triton Inference Server:
diff --git a/qa/common/gen_jetson_trt_models b/qa/common/gen_jetson_trt_models
@@ -34,7 +34,7 @@
 # Make all generated files accessible outside of container
 umask 0000
 # Set the version of the models
-TRITON_VERSION=${TRITON_VERSION:=25.03}
+TRITON_VERSION=${TRITON_VERSION:=25.04}
 # Set the CUDA device to use
 CUDA_DEVICE=${RUNNER_ID:=0}
 # Set TensorRT image
diff --git a/qa/common/gen_qa_custom_ops b/qa/common/gen_qa_custom_ops
@@ -37,7 +37,7 @@
 ##
 ############################################################################
 
-TRITON_VERSION=${TRITON_VERSION:=25.03}
+TRITON_VERSION=${TRITON_VERSION:=25.04}
 NVIDIA_UPSTREAM_VERSION=${NVIDIA_UPSTREAM_VERSION:=$TRITON_VERSION}
 PYTORCH_IMAGE=${PYTORCH_IMAGE:=nvcr.io/nvidia/pytorch:$NVIDIA_UPSTREAM_VERSION-py3}
 UBUNTU_IMAGE=${UBUNTU_IMAGE:=ubuntu:24.04}
diff --git a/qa/common/gen_qa_model_repository b/qa/common/gen_qa_model_repository
@@ -48,7 +48,7 @@
 ##
 ############################################################################
 
-TRITON_VERSION=${TRITON_VERSION:=25.03}
+TRITON_VERSION=${TRITON_VERSION:=25.04}
 
 # ONNX. Use ONNX_OPSET 0 to use the default for ONNX version
 ONNX_VERSION=1.16.1
diff --git a/src/python/tritonfrontend/_api/_kservegrpc.py b/src/python/tritonfrontend/_api/_kservegrpc.py
@@ -79,6 +79,7 @@ class Options:
         infer_compression_level: Union[
             int, Grpc_compression_level
         ] = Grpc_compression_level.NONE
+        infer_thread_count: int = Field(2, ge=0)
         infer_allocation_pool_size: int = Field(8, ge=0)
         max_response_pool_size: int = Field(2_147_483_647, ge=0)
         forward_header_pattern: str = ""
diff --git a/src/python/tritonfrontend/_api/_kservegrpc.pyi b/src/python/tritonfrontend/_api/_kservegrpc.pyi
@@ -56,6 +56,7 @@ class KServeGrpc:
         max_connection_age_ms: int
         max_connection_age_grace_ms: int
         infer_compression_level: int | Grpc_compression_level
+        infer_thread_count: int
         infer_allocation_pool_size: int
         max_response_pool_size: int
         forward_header_pattern: str

Original file line number	Diff line number	Diff line change
`@@ -29,7 +29,7 @@`
`29`	`29`	`#`
`30`	`30`
`31`	`31`	`# Base image on the minimum Triton container`
`32`		`-ARG BASE_IMAGE=nvcr.io/nvidia/tritonserver:25.03-py3-min`
	`32`	`+ARG BASE_IMAGE=nvcr.io/nvidia/tritonserver:25.04-py3-min`
`33`	`33`
`34`	`34`	`ARG TRITON_CLIENT_REPO_SUBDIR=clientrepo`
`35`	`35`	`ARG TRITON_PA_REPO_SUBDIR=perfanalyzerrepo`