Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Torch Tensor RT example #2483

Merged
merged 3 commits into from
Jul 21, 2023
Merged

Torch Tensor RT example #2483

merged 3 commits into from
Jul 21, 2023

Conversation

agunapal
Copy link
Collaborator

@agunapal agunapal commented Jul 19, 2023 •
edited
Loading

Description

This PR

  • Enables Torch TensorRT to work without a custom handler
  • Example showing how to use Torch TensorRT with TorchServe

Fixes #(issue)

Type of change

Please delete options that are not relevant.

  • Bug fix (non-breaking change which fixes an issue)
  • Breaking change (fix or feature that would cause existing functionality to not work as expected)
  • New feature (non-breaking change which adds functionality)
  • This change requires a documentation update

Feature/Issue validation/testing

Before the fix

2023-07-19T21:15:42,182 [INFO ] main org.pytorch.serve.servingsdk.impl.PluginsManager - Initializing plugins manager...
2023-07-19T21:15:42,234 [INFO ] main org.pytorch.serve.metrics.configuration.MetricConfiguration - Successfully loaded metrics configuration from /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml
2023-07-19T21:15:42,309 [INFO ] main org.pytorch.serve.ModelServer - 
Torchserve version: 0.8.1
TS Home: /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages
Current directory: /home/ubuntu/serve
Temp directory: /tmp
Metrics config path: /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml
Number of GPUs: 1
Number of CPUs: 8
Max heap size: 7936 M
Python executable: /home/ubuntu/anaconda3/envs/torchserve/bin/python
Config file: N/A
Inference address: http://127.0.0.1:8080
Management address: http://127.0.0.1:8081
Metrics address: http://127.0.0.1:8082
Model Store: /home/ubuntu/serve/model_store
Initial Models: res50-trt-fp16=res50-trt-fp16.mar
Log dir: /home/ubuntu/serve/logs
Metrics dir: /home/ubuntu/serve/logs
Netty threads: 0
Netty client threads: 0
Default workers per model: 1
Blacklist Regex: N/A
Maximum Response Size: 6553500
Maximum Request Size: 6553500
Limit Maximum Image Pixels: true
Prefer direct buffer: false
Allowed Urls: [file://.*|http(s)?://.*]
Custom python dependency for model allowed: false
Enable metrics API: true
Metrics mode: log
Disable system metrics: false
Workflow Store: /home/ubuntu/serve/model_store
Model config: N/A
2023-07-19T21:15:42,314 [INFO ] main org.pytorch.serve.servingsdk.impl.PluginsManager -  Loading snapshot serializer plugin...
2023-07-19T21:15:42,328 [INFO ] main org.pytorch.serve.ModelServer - Loading initial models: res50-trt-fp16.mar
2023-07-19T21:15:43,023 [DEBUG] main org.pytorch.serve.wlm.ModelVersionedRefs - Adding new version 1.0 for model res50-trt-fp16
2023-07-19T21:15:43,023 [DEBUG] main org.pytorch.serve.wlm.ModelVersionedRefs - Setting default version to 1.0 for model res50-trt-fp16
2023-07-19T21:15:43,024 [INFO ] main org.pytorch.serve.wlm.ModelManager - Model res50-trt-fp16 loaded.
2023-07-19T21:15:43,024 [DEBUG] main org.pytorch.serve.wlm.ModelManager - updateModel: res50-trt-fp16, count: 1
2023-07-19T21:15:43,029 [DEBUG] W-9000-res50-trt-fp16_1.0 org.pytorch.serve.wlm.WorkerLifeCycle - Worker cmdline: [/home/ubuntu/anaconda3/envs/torchserve/bin/python, /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/model_service_worker.py, --sock-type, unix, --sock-name, /tmp/.ts.sock.9000, --metrics-config, /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml]
2023-07-19T21:15:43,030 [INFO ] main org.pytorch.serve.ModelServer - Initialize Inference server with: EpollServerSocketChannel.
2023-07-19T21:15:43,083 [INFO ] main org.pytorch.serve.ModelServer - Inference API bind to: http://127.0.0.1:8080
2023-07-19T21:15:43,084 [INFO ] main org.pytorch.serve.ModelServer - Initialize Management server with: EpollServerSocketChannel.
2023-07-19T21:15:43,085 [INFO ] main org.pytorch.serve.ModelServer - Management API bind to: http://127.0.0.1:8081
2023-07-19T21:15:43,085 [INFO ] main org.pytorch.serve.ModelServer - Initialize Metrics server with: EpollServerSocketChannel.
2023-07-19T21:15:43,085 [INFO ] main org.pytorch.serve.ModelServer - Metrics API bind to: http://127.0.0.1:8082
Model server started.
2023-07-19T21:15:43,254 [WARN ] pool-3-thread-1 org.pytorch.serve.metrics.MetricCollector - worker pid is not available yet.
2023-07-19T21:15:43,909 [INFO ] pool-3-thread-1 TS_METRICS - CPUUtilization.Percent:0.0|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,910 [INFO ] pool-3-thread-1 TS_METRICS - DiskAvailable.Gigabytes:65.70399475097656|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,910 [INFO ] pool-3-thread-1 TS_METRICS - DiskUsage.Gigabytes:127.93020629882812|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,910 [INFO ] pool-3-thread-1 TS_METRICS - DiskUtilization.Percent:66.1|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,910 [INFO ] pool-3-thread-1 TS_METRICS - GPUMemoryUtilization.Percent:0.0|#Level:Host,DeviceId:0|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,911 [INFO ] pool-3-thread-1 TS_METRICS - GPUMemoryUsed.Megabytes:0.0|#Level:Host,DeviceId:0|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,911 [INFO ] pool-3-thread-1 TS_METRICS - GPUUtilization.Percent:0.0|#Level:Host,DeviceId:0|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,911 [INFO ] pool-3-thread-1 TS_METRICS - MemoryAvailable.Megabytes:29600.74609375|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,912 [INFO ] pool-3-thread-1 TS_METRICS - MemoryUsed.Megabytes:1702.80078125|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:43,912 [INFO ] pool-3-thread-1 TS_METRICS - MemoryUtilization.Percent:6.7|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801343
2023-07-19T21:15:44,246 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - s_name_part0=/tmp/.ts.sock, s_name_part1=9000, pid=37899
2023-07-19T21:15:44,247 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Listening on port: /tmp/.ts.sock.9000
2023-07-19T21:15:44,255 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Successfully loaded /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml.
2023-07-19T21:15:44,256 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - [PID]37899
2023-07-19T21:15:44,256 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Torch worker started.
2023-07-19T21:15:44,256 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Python runtime: 3.10.0
2023-07-19T21:15:44,256 [DEBUG] W-9000-res50-trt-fp16_1.0 org.pytorch.serve.wlm.WorkerThread - W-9000-res50-trt-fp16_1.0 State change null -> WORKER_STARTED
2023-07-19T21:15:44,259 [INFO ] W-9000-res50-trt-fp16_1.0 org.pytorch.serve.wlm.WorkerThread - Connecting to: /tmp/.ts.sock.9000
2023-07-19T21:15:44,265 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Connection accepted: /tmp/.ts.sock.9000.
2023-07-19T21:15:44,266 [INFO ] W-9000-res50-trt-fp16_1.0 org.pytorch.serve.wlm.WorkerThread - Flushing req.cmd LOAD to backend at: 1689801344266
2023-07-19T21:15:44,290 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - model_name: res50-trt-fp16, batchSize: 1
2023-07-19T21:15:44,893 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Enabled tensor cores
2023-07-19T21:15:44,907 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - ONNX enabled
2023-07-19T21:15:44,941 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Failed to load model res50-trt-fp16, exception 
2023-07-19T21:15:44,942 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Unknown type name '__torch__.torch.classes.tensorrt.Engine':
2023-07-19T21:15:44,942 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "code/__torch__/torchvision/models/resnet.py", line 4
2023-07-19T21:15:44,942 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   __parameters__ = []
2023-07-19T21:15:44,942 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   __buffers__ = []
2023-07-19T21:15:44,943 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   __torch___torchvision_models_resnet_ResNet_trt_engine_ : __torch__.torch.classes.tensorrt.Engine
2023-07-19T21:15:44,943 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -                                                            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ <--- HERE
2023-07-19T21:15:44,943 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   def forward(self_1: __torch__.torchvision.models.resnet.ResNet_trt,
2023-07-19T21:15:44,943 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     input_0: Tensor) -> Tensor:
2023-07-19T21:15:44,943 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Traceback (most recent call last):
2023-07-19T21:15:44,944 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "/home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/model_service_worker.py", line 131, in load_model
2023-07-19T21:15:44,944 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     service = model_loader.load(
2023-07-19T21:15:44,944 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "/home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/model_loader.py", line 135, in load
2023-07-19T21:15:44,944 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     initialize_fn(service.context)
2023-07-19T21:15:44,944 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "/home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/torch_handler/vision_handler.py", line 23, in initialize
2023-07-19T21:15:44,945 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     super().initialize(context)
2023-07-19T21:15:44,945 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "/home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/torch_handler/base_handler.py", line 172, in initialize
2023-07-19T21:15:44,945 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     self.model = self._load_torchscript_model(self.model_pt_path)
2023-07-19T21:15:44,945 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "/home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/torch_handler/base_handler.py", line 226, in _load_torchscript_model
2023-07-19T21:15:44,945 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     return torch.jit.load(model_pt_path, map_location=self.device)
2023-07-19T21:15:44,946 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "/home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/torch/jit/_serialization.py", line 162, in load
2023-07-19T21:15:44,946 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -     cpp_module = torch._C.import_ir_module(cu, str(f), map_location, _extra_files, _restore_shapes)  # type: ignore[call-arg]
2023-07-19T21:15:44,946 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - RuntimeError: 
2023-07-19T21:15:44,946 [DEBUG] W-9000-res50-trt-fp16_1.0 org.pytorch.serve.wlm.WorkerThread - sent a reply, jobdone: true
2023-07-19T21:15:44,946 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG - Unknown type name '__torch__.torch.classes.tensorrt.Engine':
2023-07-19T21:15:44,946 [INFO ] W-9000-res50-trt-fp16_1.0 org.pytorch.serve.wlm.WorkerThread - Backend response time: 656
2023-07-19T21:15:44,946 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   File "code/__torch__/torchvision/models/resnet.py", line 4
2023-07-19T21:15:44,947 [INFO ] W-9000-res50-trt-fp16_1.0-stdout MODEL_LOG -   __parameters__ = []

After the fix

2023-07-19T21:13:58,627 [INFO ] main org.pytorch.serve.servingsdk.impl.PluginsManager - Initializing plugins manager...
2023-07-19T21:13:58,680 [INFO ] main org.pytorch.serve.metrics.configuration.MetricConfiguration - Successfully loaded metrics configuration from /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml
2023-07-19T21:13:58,757 [INFO ] main org.pytorch.serve.ModelServer - 
Torchserve version: 0.8.1
TS Home: /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages
Current directory: /home/ubuntu/serve
Temp directory: /tmp
Metrics config path: /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml
Number of GPUs: 1
Number of CPUs: 8
Max heap size: 7936 M
Python executable: /home/ubuntu/anaconda3/envs/torchserve/bin/python
Config file: N/A
Inference address: http://127.0.0.1:8080
Management address: http://127.0.0.1:8081
Metrics address: http://127.0.0.1:8082
Model Store: /home/ubuntu/serve/model_store
Initial Models: res50=res50-trt-fp16.mar
Log dir: /home/ubuntu/serve/logs
Metrics dir: /home/ubuntu/serve/logs
Netty threads: 0
Netty client threads: 0
Default workers per model: 1
Blacklist Regex: N/A
Maximum Response Size: 6553500
Maximum Request Size: 6553500
Limit Maximum Image Pixels: true
Prefer direct buffer: false
Allowed Urls: [file://.*|http(s)?://.*]
Custom python dependency for model allowed: false
Enable metrics API: true
Metrics mode: log
Disable system metrics: false
Workflow Store: /home/ubuntu/serve/model_store
Model config: N/A
2023-07-19T21:13:58,762 [INFO ] main org.pytorch.serve.servingsdk.impl.PluginsManager -  Loading snapshot serializer plugin...
2023-07-19T21:13:58,776 [INFO ] main org.pytorch.serve.ModelServer - Loading initial models: res50-trt-fp16.mar
2023-07-19T21:13:59,464 [DEBUG] main org.pytorch.serve.wlm.ModelVersionedRefs - Adding new version 1.0 for model res50
2023-07-19T21:13:59,464 [DEBUG] main org.pytorch.serve.wlm.ModelVersionedRefs - Setting default version to 1.0 for model res50
2023-07-19T21:13:59,465 [INFO ] main org.pytorch.serve.wlm.ModelManager - Model res50 loaded.
2023-07-19T21:13:59,465 [DEBUG] main org.pytorch.serve.wlm.ModelManager - updateModel: res50, count: 1
2023-07-19T21:13:59,470 [DEBUG] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerLifeCycle - Worker cmdline: [/home/ubuntu/anaconda3/envs/torchserve/bin/python, /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/model_service_worker.py, --sock-type, unix, --sock-name, /tmp/.ts.sock.9000, --metrics-config, /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml]
2023-07-19T21:13:59,472 [INFO ] main org.pytorch.serve.ModelServer - Initialize Inference server with: EpollServerSocketChannel.
2023-07-19T21:13:59,525 [INFO ] main org.pytorch.serve.ModelServer - Inference API bind to: http://127.0.0.1:8080
2023-07-19T21:13:59,526 [INFO ] main org.pytorch.serve.ModelServer - Initialize Management server with: EpollServerSocketChannel.
2023-07-19T21:13:59,527 [INFO ] main org.pytorch.serve.ModelServer - Management API bind to: http://127.0.0.1:8081
2023-07-19T21:13:59,527 [INFO ] main org.pytorch.serve.ModelServer - Initialize Metrics server with: EpollServerSocketChannel.
2023-07-19T21:13:59,528 [INFO ] main org.pytorch.serve.ModelServer - Metrics API bind to: http://127.0.0.1:8082
Model server started.
2023-07-19T21:13:59,676 [WARN ] pool-3-thread-1 org.pytorch.serve.metrics.MetricCollector - worker pid is not available yet.
2023-07-19T21:14:00,283 [INFO ] pool-3-thread-1 TS_METRICS - CPUUtilization.Percent:0.0|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,284 [INFO ] pool-3-thread-1 TS_METRICS - DiskAvailable.Gigabytes:65.70406341552734|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,285 [INFO ] pool-3-thread-1 TS_METRICS - DiskUsage.Gigabytes:127.93013763427734|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,285 [INFO ] pool-3-thread-1 TS_METRICS - DiskUtilization.Percent:66.1|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,285 [INFO ] pool-3-thread-1 TS_METRICS - GPUMemoryUtilization.Percent:0.0|#Level:Host,DeviceId:0|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,285 [INFO ] pool-3-thread-1 TS_METRICS - GPUMemoryUsed.Megabytes:0.0|#Level:Host,DeviceId:0|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,286 [INFO ] pool-3-thread-1 TS_METRICS - GPUUtilization.Percent:0.0|#Level:Host,DeviceId:0|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,286 [INFO ] pool-3-thread-1 TS_METRICS - MemoryAvailable.Megabytes:29626.125|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,286 [INFO ] pool-3-thread-1 TS_METRICS - MemoryUsed.Megabytes:1677.9296875|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,287 [INFO ] pool-3-thread-1 TS_METRICS - MemoryUtilization.Percent:6.7|#Level:Host|#hostname:ip-172-31-7-107,timestamp:1689801240
2023-07-19T21:14:00,710 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - s_name_part0=/tmp/.ts.sock, s_name_part1=9000, pid=37148
2023-07-19T21:14:00,711 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Listening on port: /tmp/.ts.sock.9000
2023-07-19T21:14:00,719 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Successfully loaded /home/ubuntu/anaconda3/envs/torchserve/lib/python3.10/site-packages/ts/configs/metrics.yaml.
2023-07-19T21:14:00,719 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - [PID]37148
2023-07-19T21:14:00,719 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Torch worker started.
2023-07-19T21:14:00,719 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Python runtime: 3.10.0
2023-07-19T21:14:00,720 [DEBUG] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerThread - W-9000-res50_1.0 State change null -> WORKER_STARTED
2023-07-19T21:14:00,723 [INFO ] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerThread - Connecting to: /tmp/.ts.sock.9000
2023-07-19T21:14:00,728 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Connection accepted: /tmp/.ts.sock.9000.
2023-07-19T21:14:00,730 [INFO ] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerThread - Flushing req.cmd LOAD to backend at: 1689801240730
2023-07-19T21:14:00,753 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - model_name: res50, batchSize: 1
2023-07-19T21:14:01,361 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Enabled tensor cores
2023-07-19T21:14:01,375 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - ONNX enabled
2023-07-19T21:14:01,634 [INFO ] W-9000-res50_1.0-stdout MODEL_LOG - Torch TensorRT enabled
2023-07-19T21:14:03,138 [WARN ] W-9000-res50_1.0-stderr MODEL_LOG - WARNING: [Torch-TensorRT] - TensorRT was linked against cuDNN 8.6.0 but loaded cuDNN 8.5.0
2023-07-19T21:14:03,144 [WARN ] W-9000-res50_1.0-stderr MODEL_LOG - WARNING: [Torch-TensorRT] - TensorRT was linked against cuDNN 8.6.0 but loaded cuDNN 8.5.0
2023-07-19T21:14:03,144 [WARN ] W-9000-res50_1.0-stderr MODEL_LOG - WARNING: [Torch-TensorRT] - CUDA lazy loading is not enabled. Enabling it can significantly reduce device memory usage. See `CUDA_MODULE_LOADING` in https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#env-vars
2023-07-19T21:14:03,158 [DEBUG] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerThread - sent a reply, jobdone: true
2023-07-19T21:14:03,158 [INFO ] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerThread - Backend response time: 2405
2023-07-19T21:14:03,159 [DEBUG] W-9000-res50_1.0 org.pytorch.serve.wlm.WorkerThread - W-9000-res50_1.0 State change WORKER_STARTED -> WORKER_MODEL_LOADED

Checklist:

  • Did you have fun?
  • Have you added tests that prove your fix is effective or that this feature works?
  • Has code been commented, particularly in hard-to-understand areas?
  • Have you made corresponding changes to the documentation?

@codecov
Copy link

codecov bot commented Jul 19, 2023 •
edited
Loading

Codecov Report

Merging #2483 (f877efc) into master (7e5857f) will increase coverage by 0.01%.
The diff coverage is 80.00%.

❗ Current head f877efc differs from pull request most recent head 05ed115. Consider uploading reports for the commit 05ed115 to get more accurate results

@@            Coverage Diff             @@
##           master    #2483      +/-   ##
==========================================
+ Coverage   71.89%   71.90%   +0.01%     
==========================================
  Files          78       78              
  Lines        3654     3659       +5     
  Branches       58       58              
==========================================
+ Hits         2627     2631       +4     
- Misses       1023     1024       +1     
  Partials        4        4
Impacted Files Coverage Δ
ts/torch_handler/base_handler.py 56.54% <80.00%> (+0.56%) ⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

@msaroufim msaroufim self-requested a review July 20, 2023 00:20
@agunapal agunapal merged commit b998f8c into master Jul 21, 2023
12 of 13 checks passed
@agunapal agunapal deleted the issues/import_trt branch July 21, 2023 19:20
@sachanub sachanub mentioned this pull request Jul 25, 2023
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@msaroufim msaroufim msaroufim approved these changes

@lxning lxning lxning approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
3 participants
@agunapal @msaroufim @lxning