I’m trying to import a saved Keras model from disk:
from keras._tf_keras.keras.models import load_model
model = load_model('data/model.keras')
print(model.summary())
(Note: keras is imported like this because everything other solution I tried gave an Import 'keras._' could not be resolved error.)
When I run this using py in a virtual environment:
(env) $ py test.py
2024-07-12 18:40:21.267652: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`.
2024-07-12 18:40:21.268097: I external/local_tsl/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2024-07-12 18:40:21.270249: I external/local_tsl/tsl/cuda/cudart_stub.cc:32] Could not find cuda drivers on your machine, GPU will not be used.
2024-07-12 18:40:21.278556: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:479] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-07-12 18:40:21.294538: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:10575] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-07-12 18:40:21.294580: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1442] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
2024-07-12 18:40:21.304639: I tensorflow/core/platform/cpu_feature_guard.cc:210] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
2024-07-12 18:40:21.785069: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
/home/user/projects/python/bonk_bot/env/lib/python3.12/site-packages/keras/src/saving/saving_lib.py:576: UserWarning: Skipping variable loading for optimizer 'adam', because it has 22 variables whereas the saved optimizer has 2 variables.
saveable.load_own_variables(weights_store.get(inner_path))
Model: "sequential"
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓
┃ Layer (type) ┃ Output Shape ┃ Param # ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩
│ conv3d (Conv3D) │ (None, 1, 1, 158, 32) │ 15,073,312 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ activation (Activation) │ (None, 1, 1, 158, 32) │ 0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ conv3d_1 (Conv3D) │ (None, 1, 1, 79, 64) │ 131,136 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ activation_1 (Activation) │ (None, 1, 1, 79, 64) │ 0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ conv3d_2 (Conv3D) │ (None, 1, 1, 79, 64) │ 110,656 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ activation_2 (Activation) │ (None, 1, 1, 79, 64) │ 0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ flatten (Flatten) │ (None, 5056) │ 0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense (Dense) │ (None, 512) │ 2,589,184 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ dense_1 (Dense) │ (None, 6) │ 3,078 │
└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘
Total params: 53,722,100 (204.93 MB)
Trainable params: 17,907,366 (68.31 MB)
Non-trainable params: 0 (0.00 B)
Optimizer params: 35,814,734 (136.62 MB)
None
Everything works fine and the model prints the summary as expected (apart from the verbose output it gives out at the beginning). However, VSCode gives an error when opened:
This happens for every function I try to use on model. The editor I am currently using is Code-OSS version 1.91.0, and my linter is Pyright. I made sure that the editor was using the correct interpreter inside the Python virtual environment. I doubt it’s a Keras problem since the editor does this. Does anyone know why my editor thinks it’s an error and how to fix it?