AIWintermuteAI / Speech-to-Intent-Micro

An open-source, easily accessible package for training and deploying Speech-to-Intent models on microcontrollers and SBCs
Apache License 2.0
37 stars 7 forks source link

LSTM layer instead of Fully-Connected + Time Distribute #7

Open Tabrez-dev opened 7 months ago

Tabrez-dev commented 7 months ago

I was trying to create this model but i ran into some errors can u have a look.

`--------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[20], line 33
     29 intent_output = Dense(n_classes, activation='softmax', name='intent_output', use_bias = False)(x)
     31 model = Model(inputs=main_input, outputs=[intent_output, slot_output])
---> 33 optim = Adam(learning_rate=1e-3, weight_decay=1e-6)
     35 model.compile(optimizer = optim, loss='categorical_crossentropy', metrics='accuracy')
     36 model.summary()

File ~/anaconda3/envs/speech-to-intent-micro/lib/python3.8/site-packages/keras/optimizers/optimizer_v2/adam.py:110, in Adam.__init__(self, learning_rate, beta_1, beta_2, epsilon, amsgrad, name, **kwargs)
    102 def __init__(self,
    103              learning_rate=0.001,
    104              beta_1=0.9,
   (...)
    108              name='Adam',
    109              **kwargs):
--> 110   super(Adam, self).__init__(name, **kwargs)
    111   self._set_hyper('learning_rate', kwargs.get('lr', learning_rate))
    112   self._set_hyper('decay', self._initial_decay)

File ~/anaconda3/envs/speech-to-intent-micro/lib/python3.8/site-packages/keras/optimizers/optimizer_v2/optimizer_v2.py:351, in OptimizerV2.__init__(self, name, gradient_aggregator, gradient_transformers, **kwargs)
    349 for k in kwargs:
    350   if k not in allowed_kwargs:
--> 351     raise TypeError("Unexpected keyword argument "
    352                     f"passed to optimizer: {str(k)}. Allowed kwargs are "
    353                     f"{allowed_kwargs}.")
    354   # checks that all keyword arguments are non-negative.
    355   if kwargs[k] is not None and kwargs[k] < 0:

TypeError: Unexpected keyword argument passed to optimizer: weight_decay. Allowed kwargs are {'decay', 'lr', 'clipvalue', 'clipnorm', 'global_clipnorm'}.
`
Tabrez-dev commented 7 months ago

Im getting early stopping error after making it run somehow at 28 epoch is this normal?