-
Notifications
You must be signed in to change notification settings - Fork 2k
Closed
tensorflow/tfjs-converter
#195Labels
Description
When I tried to implement a model converted from tensorflow frozen model, the execute() method gave me an Error said
tf-converter.esm.js?1e04:17
Uncaught (in promise) Error: The model contains control flow or dynamic shape ops, please use executeAsync method
at e.execute (tf-converter.esm.js?1e04:17)
at VueComponent._callee$ (TensorFlowJsTry.vue?dd89:50)
at tryCatch (runtime.js?4a57:62)
at Generator.invoke [as _invoke] (runtime.js?4a57:296)
at Generator.prototype.(:8080/anonymous function) [as next] (webpack-internal:///./node_modules/regenerator-runtime/runtime.js:114:21)
at step (asyncToGenerator.js?7b11:17)
at eval (asyncToGenerator.js?7b11:28)
But when I replaced the execute() method with executeAsync(), it gave me an Error said
tf-converter.esm.js?1e04:17
Uncaught (in promise) Error: The model does not contain control flow or dynamic shape ops, please use execute method for better performance.
at eval (tf-converter.esm.js?1e04:17)
at eval (tf-converter.esm.js?1e04:17)
at Object.eval [as next] (tf-converter.esm.js?1e04:17)
at eval (tf-converter.esm.js?1e04:17)
at new Promise (<anonymous>)
at __awaiter (tf-converter.esm.js?1e04:17)
at e.executeAsync (tf-converter.esm.js?1e04:17)
at VueComponent._callee$ (TensorFlowJsTry.vue?dd89:50)
at tryCatch (runtime.js?4a57:62)
at Generator.invoke [as _invoke] (runtime.js?4a57:296)
Any idea about what happened?