Hello , I've got a crash when trying to quantize mnist model with tf

<a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/us

Hi <a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="

<a class="user-mention notranslate" data-hovercard-type="user" data-hovercard-url="/us

Are you satisfied with the resolution of your issue? <a href="https://docs.google.

tflite converter crashes with version 2.16.1 about tensorflow HOT 5 CLOSED

omontf commented on April 30, 2024

tflite converter crashes with version 2.16.1

from tensorflow.

Comments (5)

tilakrayal commented on April 30, 2024

@omontf,
By default Tensorflow v2.16 uses keras v3.0 which might be the reason for the error you are facing. As the work-around, Could you please try to install keras 2 where the code was executed without any issue/error. Kindly find the gist of it here.

!pip install -U tf_keras # Keras 2
import os
os.environ["TF_USE_LEGACY_KERAS"] = "1"

Thank you!

from tensorflow.

omontf commented on April 30, 2024

Hi @tilakrayal

I tried your suggestion.
I confirm that it works well with Keras v2.

Thank you !

Olivier

from tensorflow.

tilakrayal commented on April 30, 2024

@omontf,
Glad the issue is resolved for you, Could you please feel free to move this issue to closed status. Thank you!

from tensorflow.

google-ml-butler commented on April 30, 2024

Are you satisfied with the resolution of your issue?
Yes
No

from tensorflow.

savvn001 commented on April 30, 2024

Same issue here. macOS M2.
Above solution is not working for me.

2024-04-03 18:55:04.158262: I metal_plugin/src/device/metal_device.cc:1154] Metal device set to: Apple M2
2024-04-03 18:55:04.158289: I metal_plugin/src/device/metal_device.cc:296] systemMemory: 16.00 GB
2024-04-03 18:55:04.158295: I metal_plugin/src/device/metal_device.cc:313] maxCacheSize: 5.33 GB
2024-04-03 18:55:04.158337: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:305] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.
2024-04-03 18:55:04.158350: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:271] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 0 MB memory) -> physical PluggableDevice (device: 0, name: METAL, pci bus id: <undefined>)
WARNING: All log messages before absl::InitializeLog() is called are written to STDERR
W0000 00:00:1712166904.332563   84530 tf_tfl_flatbuffer_helpers.cc:390] Ignored output_format.
W0000 00:00:1712166904.332624   84530 tf_tfl_flatbuffer_helpers.cc:393] Ignored drop_control_dependency.
2024-04-03 18:55:04.333233: I tensorflow/cc/saved_model/reader.cc:83] Reading SavedModel from: mypath
2024-04-03 18:55:04.334679: I tensorflow/cc/saved_model/reader.cc:51] Reading meta graph with tags { serve }
2024-04-03 18:55:04.334692: I tensorflow/cc/saved_model/reader.cc:146] Reading SavedModel debug info (if present) from: mypath
2024-04-03 18:55:04.352055: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:388] MLIR V1 optimization pass is not enabled
2024-04-03 18:55:04.354424: I tensorflow/cc/saved_model/loader.cc:234] Restoring SavedModel bundle.
2024-04-03 18:55:04.420357: I tensorflow/cc/saved_model/loader.cc:218] Running initialization op on SavedModel bundle at path: mypath
2024-04-03 18:55:04.440781: I tensorflow/cc/saved_model/loader.cc:317] SavedModel load for tags { serve }; Status: success: OK. Took 107549 microseconds.
2024-04-03 18:55:04.459571: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:268] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.
loc(fused["ReadVariableOp:", "sequential_4_1/conv1d_24_1/Reshape/ReadVariableOp@__inference_serving_default_142497"]): error: missing attribute 'value'
LLVM ERROR: Failed to infer result type(s).
zsh: abort      python3 convert_model.py

from tensorflow.

Recommend Projects