curl 404 ResourceNotFoundException

[ma-batita]

Hello, I am stuck with an error that I am not sure what does it mean. when I do curl "http://localhost:8080/models" I get : { "code": 404, "type": "ResourceNotFoundException", "message": "Requested resource is not found, please refer to API document." }

I make an .mar file for my model with torch-model-archiver -f \ --model-name=classifier \ --version=1.0 \ --serialized-file=pytorch_model.bin \ --handler=custom_handler.py \ --extra-files "config.json,index_to_name.json,special_tokens_map.json,tokenizer_config.json,tokenizer.json,training_args.bin,vocab.txt" \ --export-path=model_store All of those files are stored in the same directory.

When i run the serve torchserve --start --model-store model_store --models classifier=classifier.mar I dont get any error. normally when I do curl "http://localhost:8080/models" I will get my classifier but I instead I get that message.

is there anything that I am missing here? or should I add something? I want to mention that I am using a handler (custom_handler.py) from GoogleCloudPlatform. also, curl localhost:8080/ping give me Healthy Thanks!

[msaroufim]

I'm not too sure from your description if you got the behavior you're looking for with all torchserve models or just the custom one you tried out. In case it's a custom one I'd suggest you take a look at logs/model_log.log in the folder where you started torchserve as it may indicate some issue

[ma-batita]

Thank you for your response :) but I dont see anything else beside my files :/

[lxning]

@ma-batita I think you need use KFServing port

[Rishav-hub]

Getting the same issue checked the logs/model_log.log. Getting these logs.

2023-01-31T15:48:31,175 [INFO ] W-9005-mnist_1.0-stdout MODEL_LOG -  for CPU please run pip3 install --pre torch --extra-index-url https://download.pytorch.org/whl/nightly/cpu
2023-01-31T15:48:31,241 [INFO ] W-9014-mnist_1.0-stdout MODEL_LOG - /home/model-server/tmp/models/23f9575c3b024073932b98e2c9301e8e/compile.json is missing. PT 2.0 will not be used
2023-01-31T15:48:31,242 [INFO ] W-9014-mnist_1.0-stdout MODEL_LOG - dynamo/inductor are not installed.  
2023-01-31T15:48:31,243 [INFO ] W-9014-mnist_1.0-stdout MODEL_LOG -  For GPU please run pip3 install numpy --pre torch[dynamo] --force-reinstall --extra-index-url https://download.pytorch.org/whl/nightly/cu117
2023-01-31T15:48:31,243 [INFO ] W-9014-mnist_1.0-stdout MODEL_LOG -  for CPU please run pip3 install --pre torch --extra-index-url https://download.pytorch.org/whl/nightly/cpu
2023-01-31T15:48:31,289 [INFO ] W-9009-mnist_1.0-stdout MODEL_LOG - /home/model-server/tmp/models/23f9575c3b024073932b98e2c9301e8e/compile.json is missing. PT 2.0 will not be used
2023-01-31T15:48:31,290 [INFO ] W-9009-mnist_1.0-stdout MODEL_LOG - dynamo/inductor are not installed.  
2023-01-31T15:48:31,290 [INFO ] W-9009-mnist_1.0-stdout MODEL_LOG -  For GPU please run pip3 install numpy --pre torch[dynamo] --force-reinstall --extra-index-url https://download.pytorch.org/whl/nightly/cu117
2023-01-31T15:48:31,291 [INFO ] W-9009-mnist_1.0-stdout MODEL_LOG -  for CPU please run pip3 install --pre torch --extra-index-url https://download.pytorch.org/whl/nightly/cpu
2023-01-31T15:48:31,347 [INFO ] W-9011-mnist_1.0-stdout MODEL_LOG - /home/model-server/tmp/models/23f9575c3b024073932b98e2c9301e8e/compile.json is missing. PT 2.0 will not be used
2023-01-31T15:48:31,348 [INFO ] W-9011-mnist_1.0-stdout MODEL_LOG - dynamo/inductor are not installed.  
2023-01-31T15:48:31,349 [INFO ] W-9011-mnist_1.0-stdout MODEL_LOG -  For GPU please run pip3 install numpy --pre torch[dynamo] --force-reinstall --extra-index-url https://download.pytorch.org/whl/nightly/cu117
2023-01-31T15:48:31,350 [INFO ] W-9011-mnist_1.0-stdout MODEL_LOG -  for CPU please run pip3 install --pre torch --extra-index-url https://download.pytorch.org/whl/nightly/cpu

I think it is not supporting torch 2.0. But I checked the torch version in my environment it's 1.13+cpu.