| # BEGIN GENERAL GGUF METADATA | |
| id: ichigo-whispervq # Model ID unique between models | |
| model: ichigo-whispervq # Model ID which is used for request construct - should be unique between models (author / quantization) | |
| name: Ichigo WhisperVQ | |
| version: 1 # metadata.version | |
| # END GENERAL METADATA | |
| # BEGIN INFERENCE PARAMETERS | |
| # BEGIN REQUIRED | |
| load_model: # method to load python model through API | |
| method: post | |
| path: /loadmodel | |
| transform_request: "" # jinja2 template to transform request | |
| transform_response: "" # jinja2 template to transform response | |
| destroy: # method to destroy python process through API | |
| method: delete | |
| path: /detroy | |
| health_check: # method to destroy python process through API | |
| method: get | |
| path: /health | |
| inference: # method to do inference python model through API | |
| method: post | |
| path: /inference | |
| transform_request: "" | |
| transform_response: "" | |
| extra_endpoints: # untilities methods | |
| - method: post | |
| path: /tokenize/wav | |
| transform_request: "" | |
| transform_response: "" | |
| - method: get | |
| path: /supported_formats | |
| # END REQUIRED | |
| # BEGIN OPTIONAL | |
| # END OPTIONAL | |
| # END INFERENCE PARAMETERS | |
| # BEGIN SERVER START PARAMETERS | |
| # BEGIN REQUIRED | |
| files: /home/thuan/cortexcpp/models/cortex.so/whispervq/fp16 | |
| port: 3348 | |
| log_path: whisper.log | |
| log_level: INFO | |
| environment: whispervq # python environment to run model | |
| script: app.py | |
| command: ["python"] # this is the base command, cortex will automatic find the correct location of python in env and add params when execute command | |
| engine: python-engine | |
| # END REQUIRED | |
| # BEGIN OPTIONAL | |
| extra_params: | |
| device_id: "0" | |
| package_dir: "" # the package directory to be searched | |
| # END OPTIONAL | |
| # END SERVER START PARAMETERS | |