matlock08 / triton-clients

Python clients for consuming triton inference server models

Hi there 👋

My name is Jose Castellanos

I'm a Solution Architect from Mexico

🔨 What I do

🖥 I'm currently working as Digital Transformation Manager at IBM

🐦 I shared content on Twitter about AI, IoT and dev.

---

Working with

• NVIDIA DeepStream, OMNIVERSE, TAO
• Typescript
• Python
• Java
• AWS
• NodeJS
• Express

📫 You can find me

• Twitter 🐦
• LinkedIn 💼

triton-clients's Issues

failed to load model

@matlock08 Hi, I enter the command like yours docker run --gpus=1 --rm -p8000:8000 -p8001:8001 -p8002:8002 -v models:/models nvcr.io/nvidia/tritonserver:21.04-py3 tritonserver --model-repository=/models
but I can not load models correctly.
| maskrcnn_onnx | 1 | UNAVAILABLE: Internal: failed to stat file /models/maskrcnn_onnx/1/model.onnx |
| yolo_v4 | 1 | UNAVAILABLE: Internal: failed to stat file /models/yolo_v4/1/model.onnx |

I0508 09:44:41.933218 1 server.cc:234] Waiting for in-flight requests to complete.
I0508 09:44:41.933223 1 server.cc:249] Timeout 30: Found 0 live models and 0 in-flight non-inference requests
error: creating server: Internal - failed to load all models

How should I modify it ? thanks.