You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
tf2.16 is the last version supports keras v2, and tf2.17 is fully based on keras v3, which is not officially supported yet in the main branch. Try PR #1116 if you want to use keras v3.
Quick summary
This is the error that appears:
10 layer = parse_default_keras_layer(keras_layer, input_names)
11
---> 12 layers['input_shape'] = keras_layer['config']['batch_input_shape'][1:]
13
14 dtype = keras_layer['config']['dtype']
KeyError: 'batch_input_shape'
Details
I'm using python3.12.3, tensorflow 2.17, but when I use an environment with python3.10 and tensorflow 2.14 I get some error
Steps to reproduce
Just run the command from the quickstart itself:
config=hls4ml.utils.config_from_keras_model( model, granularity='name' )
Expected behavior
it would have executed the configuration
Actual behavior
10 layer = parse_default_keras_layer(keras_layer, input_names)
11
---> 12 layer['input_shape'] = keras_layer['config']['batch_input_shape'][1:]
13
14 dtype = keras_layer['config']['dtype']
KeyError: 'batch_input_shape'
The text was updated successfully, but these errors were encountered: