Keras Update Stream¶
Overview¶
Compatible Pyspark only
Python node : launch with punchplatform-pyspark.sh or the punch UI job-editor
with environment pyspark
.
This is a demo node that take a trained model as input, continue to train it with some provided data, and submit the final model to the output stream. This aim to demonstrate the capability to re-train model within our pipelines.
This node have no configuration, and only work with the other keras nodes provided with the pml-pyspark.
Example configuration :¶
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 | { type: keras_update_train component: updater subscribe: [ { component: data_input stream: x_train } { component: data_input stream: y_train } { component: data_input stream: x_test } { component: data_input stream: y_test } { component: model_load stream: model } ] publish: [ { stream: model } ] } |
Example of complete pipeline :¶
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 | { job: [ { type: keras_data_input component: data_input publish: [ { stream: x_train } { stream: y_train } { stream: x_test } { stream: y_test } ] } { type: keras_load_model component: model_load settings:{ path: my_model.h5 }, publish: [ { stream: model } ] } { type: keras_update_train component: updater subscribe: [ { component: data_input stream: x_train } { component: data_input stream: y_train } { component: data_input stream: x_test } { component: data_input stream: y_test } { component: model_load stream: model } ] publish: [ { stream: model } ] } { type: keras_save_model component: model_save settings:{ path: my_model.h5 }, subscribe:[ { component: updater stream: model } ] } ] } |