Error when using the finetuned model generated with example data

[ArmandMichaud]

Hello,
I'm trying to use finetuning, with the example for now, but I get an error. Finetuning didn't create an error, but using the generated model did. I don't get any errors when using a pre-trained model or a newly trained model.

python3 AIONER_Run.py -i ../example/input/ -m bioformer-softmax-es-finetune.h5 -v ../vocab/AnEM_label.vocab -e ALL -o example/output/

`
...
Total params: 42,598,935
Trainable params: 42,598,935
Non-trainable params: 0

---

Traceback (most recent call last):
File "AIONER_Run.py", line 252, in
NER_main_path(args.inpath, para_set, args.outpath, args.model)
File "AIONER_Run.py", line 194, in NER_main_path
nn_model.load_model(modelfile)
File "/home/amichaud/Documents/main/AIONER/src/model_ner.py", line 128, in load_model
self.model.load_weights(model_file)
File "/home/amichaud/.local/lib/python3.8/site-packages/tensorflow/python/keras/engine/training.py", line 2211, in load_weights
hdf5_format.load_weights_from_hdf5_group(f, self.layers)
File "/home/amichaud/.local/lib/python3.8/site-packages/tensorflow/python/keras/saving/hdf5_format.py", line 708, in load_weights_from_hdf5_group
K.batch_set_value(weight_value_tuples)
File "/home/amichaud/.local/lib/python3.8/site-packages/tensorflow/python/util/dispatch.py", line 201, in wrapper
return target(*args, **kwargs)
File "/home/amichaud/.local/lib/python3.8/site-packages/tensorflow/python/keras/backend.py", line 3576, in batch_set_value
x.assign(np.asarray(value, dtype=dtype(x)))
File "/home/amichaud/.local/lib/python3.8/site-packages/tensorflow/python/ops/resource_variable_ops.py", line 858, in assign
self._shape.assert_is_compatible_with(value_tensor.shape)
File "/home/amichaud/.local/lib/python3.8/site-packages/tensorflow/python/framework/tensor_shape.py", line 1134, in assert_is_compatible_with
raise ValueError("Shapes %s and %s are incompatible" % (self, other))
ValueError: Shapes (128, 19) and (128, 23) are incompatible

[lingluodlut]

I think you used the unmatch lable.vocab files. The bioformer-softmax-es-finetune.h5 model should use ../vocab/AIO_label.vocab, as this model is trained to predict (Gene, Chemical, Disease, Variant, Species, CellLine, ALL) entity types, not AnEM.