TypeError: cannot pickle '_thread.RLock' object

[GOURAVSUMAN]

*TypeError: cannot pickle '_thread.RLock' object
TypeError Traceback (most recent call last)
in
----> 1 pickle.dump(learner, open('model.pkl','wb'))
2
3 model = pickle.load(open('model.pkl','rb'))
4 print(model.predict([['GST TDS RETURN', 'Tax teturn filled', 'Good bless u']]))

TypeError: cannot pickle '_thread.RLock' object

[stacy-ojwang]

i am having the same issue. Can someone please assist.

[tkmamidi]

Maybe follow this -
https://medium.com/fintechexplained/how-to-save-trained-machine-learning-models-649c3ad1c018

[itsrawlinz-jeff]

Same issue..Who's got a solution

[rohitnishad613]

I have the same issue too.

[HeinrichWizardKreuser]

has anyone found a fix for this yet? I'm not even using multiprocessing

[Gourav-infobiz]

has anyone found a fix for this yet? I'm not even using multiprocessing

Can you please share full error details and the code where you got this error.

[tkmamidi]

Maybe follow this -
https://medium.com/fintechexplained/how-to-save-trained-machine-learning-models-649c3ad1c018

As mentioned in this link, you can save and load keras models without using pickle.

Save the model:

# serialize to JSON
json_file = model.to_json()
with open(json_file_path, "w") as file:
   file.write(json_file)
# serialize weights to HDF5
model.save_weights(h5_file)

Load the model:

# load json and create model
file = open(json_file, 'r')
model_json = file.read()
file.close()
loaded_model = model_from_json(model_json)
# load weights
loaded_model.load_weights(h5_file)

Hope this helps!

[tkmamidi]

This is as other way to do it -

model.save("my_model")
model.save_weights("weights.h5")

from tensorflow import keras
my_model = keras.models.load_model('my_model')
my_model.load_weights("weights.h5")

[sushreebarsa]

@GOURAVSUMAN could you please try as per the above comment and let us know if this issue still persists ? Thanks !

[sushreebarsa]

Closing this issue due to lack of recent activity .Please feel free to reopen if you still have concern.Thanks!

[Lord-O-Binary]

This is as other way to do it -

model.save("my_model")
model.save_weights("weights.h5")

from tensorflow import keras
my_model = keras.models.load_model('my_model')
my_model.load_weights("weights.h5")

I can confirm saving as json with weights file fixes the threading issue during deepcopy