Hi,
I got the same issue but I wanna use GPU. Does anyone have an idea how to solve it?
dkjdjdfdskln
Users receive reputation points by contributing quality content to the forums.
Posts include post and replies published on the forums.
Post authors can mark replies as Accepted to indicate successful solutions.
Apple Developer Forums admins can mark replies as Apple Recommended to indicate an approved solution
Post
Replies
Boosts
Views
Activity
tensorflow version: 2.6.0
tensorflow-macos somehow cannot be found too, so I installed this package by SYSTEM_VERSION_COMPAT=0 python -m pip install tensorflow-macos. similarly for tensorflow-metal
I posted this question again because I accepted my own "answers" by mistake. I cannot find a way to delete this post.
ouputs when uncommon following code:
from tensorflow.python.framework.ops import disable_eager_execution
disable_eager_execution()
2021-10-31 18:54:21.304180: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: SSE4.2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2021-10-31 18:54:21.305317: 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.
2021-10-31 18:54:21.306071: 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>)
2021-10-31 18:54:21.392880: 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.
2021-10-31 18:54:21.392913: 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>)
2021-10-31 18:54:21.486630: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.499261: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
Metal device set to: Apple M1 Pro
systemMemory: 16.00 GB
maxCacheSize: 5.33 GB
2021-10-31 18:54:21.560669: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.577903: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.671397: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.690163: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.725452: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.745059: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.762898: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)
2021-10-31 18:54:21.782170: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.796932: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-10-31 18:54:21.839 python[42307:1235107] -[MPSGraph adamUpdateWithLearningRateTensor:beta1Tensor:beta2Tensor:epsilonTensor:beta1PowerTensor:beta2PowerTensor:valuesTensor:momentumTensor:velocityTensor:gradientTensor:name:]: unrecognized selector sent to instance 0x6000036816c0
Train on 469 steps, validate on 79 steps
Epoch 1/10
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
/var/folders/mq/pnk708vd76xfjjjmy_4yn2lw0000gn/T/ipykernel_42307/2486947832.py in <module>
58
59 start_time = time.time()
---> 60 model.fit(
61 ds_train,
62 epochs=10,
/opt/anaconda3/lib/python3.8/site-packages/keras/engine/training_v1.py in fit(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_freq, max_queue_size, workers, use_multiprocessing, **kwargs)
775
776 func = self._select_training_loop(x)
--> 777 return func.fit(
778 self,
779 x=x,
/opt/anaconda3/lib/python3.8/site-packages/keras/engine/training_arrays_v1.py in fit(self, model, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_freq, **kwargs)
638 val_x, val_y, val_sample_weights = None, None, None
639
--> 640 return fit_loop(
641 model,
642 inputs=x,
/opt/anaconda3/lib/python3.8/site-packages/keras/engine/training_arrays_v1.py in model_iteration(model, inputs, targets, sample_weights, batch_size, epochs, verbose, callbacks, val_inputs, val_targets, val_sample_weights, shuffle, initial_epoch, steps_per_epoch, validation_steps, validation_freq, mode, validation_in_fit, prepared_feed_values_from_dataset, steps_name, **kwargs)
284 else:
285 actual_inputs = ins()
--> 286 batch_outs = f(actual_inputs)
287 except tf.errors.OutOfRangeError:
288 if is_dataset:
/opt/anaconda3/lib/python3.8/site-packages/keras/backend.py in __call__(self, inputs)
4027 feed_symbols != self._feed_symbols or self.fetches != self._fetches or
4028 session != self._session):
-> 4029 self._make_callable(feed_arrays, feed_symbols, symbol_vals, session)
4030
4031 fetched = self._callable_fn(*array_vals,
/opt/anaconda3/lib/python3.8/site-packages/keras/backend.py in _make_callable(self, feed_arrays, feed_symbols, symbol_vals, session)
3963 callable_opts.run_options.CopyFrom(self.run_options)
3964 # Create callable.
-> 3965 callable_fn = session._make_callable_from_options(callable_opts)
3966 # Cache parameters corresponding to the generated callable, so that
3967 # we can detect future mismatches and refresh the callable.
/opt/anaconda3/lib/python3.8/site-packages/tensorflow/python/client/session.py in _make_callable_from_options(self, callable_options)
1509 """
1510 self._extend_graph()
-> 1511 return BaseSession._Callable(self, callable_options)
1512
1513
/opt/anaconda3/lib/python3.8/site-packages/tensorflow/python/client/session.py in __init__(self, session, callable_options)
1467 compat.as_bytes(callable_options.SerializeToString()))
1468 try:
-> 1469 self._handle = tf_session.TF_SessionMakeCallable(
1470 session._session, options_ptr)
1471 finally:
RuntimeError: Caught an unknown exception!
there has a typo, I installed sklearn by using conda install scikit-learn . sorry for any misleading.
there has a typo, I installed sklearn by using conda install scikit-learn . sorry for any misleading.
I followed the official guideline (https://scikit-learn.org/stable/developers/advanced_installation.html), but somehow still cannot work.
so I took almost one day to solve this problem, combining the solutions from:
https://github.com/scikit-learn/scikit-learn/issues/19137
and https://github.com/scikit-learn-contrib/hdbscan/issues/457and it works.
python -m pip install --no-cache --no-use-pep517 pythran cython pybind11 gast"==0.4.0"
pyenv rehash
pip install --pre -i https://pypi.anaconda.org/scipy-wheels-nightly/simple scipy
python -m pip install --no-use-pep517 scikit-learn --no-binary :all:
Hi yuliya80,
Thank you for your help. Let me give more details:
My macOS version is 12.0.1.
the code sample is MINIST toy example, from tensorflow website: https://www.tensorflow.org/datasets/keras_example
I tried it again and occurs same error. what I can feel is the tensorflow-macos is still unstable..... but thank you at least give me a direction otherwise I am totally freak out....
It seems like the small batch size will reduce the performance of GPU https://developer.apple.com/forums/thread/685623
so I increased the batch size from 256 to 1024, which reduced the running time from 40 minutes to 10 minutes per epoch. However, again, one epoch only takes around 2 minutes with CPU.
I am so confused now, it seems like I need to increase the batch size from 1024 to 1024 * 5 so that the running time will be reduced to 2 minutes per epoch.....
Update: I found M1 chip is extremely slow on LSTM compared with CNN.
Here is the pip list. and I cannot run python -c "import sklearn" on my terminal.
Package Version
absl-py 0.10.0
aiohttp 3.7.4.post0
anyio 2.2.0
appnope 0.1.2
argon2-cffi 20.1.0
astroid 2.6.6
astunparse 1.6.3
async-generator 1.10
async-timeout 3.0.1
attrs 21.2.0
autopep8 1.5.6
Babel 2.9.1
backcall 0.2.0
beniget 0.3.0
biopython 1.79
bleach 4.0.0
blinker 1.4
Bottleneck 1.3.2
brotlipy 0.7.0
cached-property 1.5.2
cachetools 4.2.2
certifi 2021.5.30
cffi 1.14.5
chardet 4.0.0
charset-normalizer 2.0.4
clang 5.0
click 8.0.3
cloudpickle 2.0.0
conda 4.10.3
conda-package-handling 1.7.3
cryptography 3.4.7
cycler 0.11.0
Cython 0.29.24
decorator 5.1.0
defusedxml 0.7.1
deprecation 2.1.0
dill 0.3.4
entrypoints 0.3
et-xmlfile 1.1.0
flake8 3.9.0
flatbuffers 1.12
future 0.18.2
gast 0.4.0
google-auth 1.33.0
google-auth-oauthlib 0.4.1
google-pasta 0.2.0
googleapis-common-protos 1.53.0
grpcio 1.38.1
gym 0.21.0
h5py 3.1.0
idna 3.2
importlib-metadata 4.8.1
ipykernel 5.3.4
ipython 7.26.0
ipython-genutils 0.2.0
isort 5.9.3
jdcal 1.4.1
jedi 0.17.2
Jinja2 3.0.2
joblib 1.1.0
json5 0.9.6
jsonschema 3.2.0
jupyter-client 7.0.1
jupyter-core 4.7.1
jupyter-http-over-ws 0.0.8
jupyter-lsp 1.5.0
jupyter-packaging 0.10.4
jupyter-server 1.6.4
jupyterlab 3.1.7
jupyterlab-lsp 3.9.1
jupyterlab-pygments 0.1.2
jupyterlab-server 2.8.2
keras 2.6.0
Keras-Preprocessing 1.1.2
kiwisolver 1.3.2
lazy-object-proxy 1.6.0
Markdown 3.3.4
MarkupSafe 2.0.1
matplotlib 3.4.2
matplotlib-inline 0.1.2
mccabe 0.6.1
mistune 0.8.4
msgpack 1.0.2
msgpack-numpy 0.4.7.1
multidict 5.1.0
nbclassic 0.2.6
nbclient 0.5.3
nbconvert 6.1.0
nbformat 5.1.3
nest-asyncio 1.5.1
networkx 2.6.3
notebook 6.4.5
numexpr 2.7.3
numpy 1.19.5
oauthlib 3.1.1
olefile 0.46
openpyxl 3.0.5
opt-einsum 3.3.0
packaging 21.0
pandas 1.3.1
pandocfilters 1.4.3
parso 0.7.0
pexpect 4.8.0
pickleshare 0.7.5
Pillow 8.4.0
pip 21.3
pluggy 0.13.1
ply 3.11
prometheus-client 0.11.0
promise 2.3
prompt-toolkit 3.0.20
protobuf 3.17.2
ptyprocess 0.7.0
pyasn1 0.4.8
pyasn1-modules 0.2.8
pybind11 2.8.1
pycodestyle 2.6.0
pycosat 0.6.3
pycparser 2.20
pydocstyle 6.1.1
pyflakes 2.2.0
pyglet 1.5.21
Pygments 2.10.0
PyJWT 2.1.0
pylint 2.9.6
pyOpenSSL 21.0.0
pyparsing 3.0.4
pyrsistent 0.18.0
PySocks 1.7.1
python-dateutil 2.8.2
python-jsonrpc-server 0.4.0
python-language-server 0.36.2
pythran 0.9.11
pytz 2021.3
pyzmq 20.0.0
requests 2.26.0
requests-oauthlib 1.3.0
rope 0.19.0
rsa 4.7.2
ruamel-yaml-conda 0.15.100
scikit-learn 1.0.1
scipy 1.8.0.dev0+1921.c30fefc
seaborn 0.11.2
Send2Trash 1.8.0
setuptools 52.0.0.post20210825
six 1.15.0
sniffio 1.2.0
snowballstemmer 2.1.0
tensorboard 2.6.0
tensorboard-data-server 0.6.0
tensorboard-plugin-wit 1.6.0
tensorflow-datasets 4.4.0
tensorflow-estimator 2.7.0
tensorflow-macos 2.6.0
tensorflow-metadata 1.4.0
tensorflow-metal 0.2.0
Tensorforce 0.6.5
termcolor 1.1.0
terminado 0.9.4
testpath 0.5.0
threadpoolctl 3.0.0
toml 0.10.2
tomlkit 0.7.0
tornado 6.1
tqdm 4.62.3
traitlets 5.1.0
typing-extensions 3.7.4.3
ujson 4.0.2
urllib3 1.26.7
wcwidth 0.2.5
webencodings 0.5.1
Werkzeug 2.0.2
wheel 0.35.1
wrapt 1.12.1
yapf 0.31.0
yarl 1.6.3
zipp 3.6.0
I reinstalled the sklearn and scipy, after that sklearn can be imported correctly.
Package Version
absl-py 0.10.0
aiohttp 3.7.4.post0
anyio 2.2.0
appnope 0.1.2
argon2-cffi 20.1.0
astroid 2.6.6
astunparse 1.6.3
async-generator 1.10
async-timeout 3.0.1
attrs 21.2.0
autopep8 1.5.6
Babel 2.9.1
backcall 0.2.0
beniget 0.3.0
biopython 1.79
bleach 4.0.0
blinker 1.4
Bottleneck 1.3.2
brotlipy 0.7.0
cached-property 1.5.2
cachetools 4.2.2
certifi 2021.5.30
cffi 1.14.5
chardet 4.0.0
charset-normalizer 2.0.4
clang 5.0
click 8.0.3
cloudpickle 2.0.0
conda 4.10.3
conda-package-handling 1.7.3
cryptography 3.4.7
cycler 0.11.0
Cython 0.29.24
decorator 5.1.0
defusedxml 0.7.1
deprecation 2.1.0
dill 0.3.4
entrypoints 0.3
et-xmlfile 1.1.0
flake8 3.9.0
flatbuffers 1.12
future 0.18.2
gast 0.4.0
google-auth 1.33.0
google-auth-oauthlib 0.4.1
google-pasta 0.2.0
googleapis-common-protos 1.53.0
grpcio 1.38.1
gym 0.21.0
h5py 3.1.0
idna 3.2
importlib-metadata 4.8.1
ipykernel 5.3.4
ipython 7.26.0
ipython-genutils 0.2.0
isort 5.9.3
jdcal 1.4.1
jedi 0.17.2
Jinja2 3.0.2
joblib 1.1.0
json5 0.9.6
jsonschema 3.2.0
jupyter-client 7.0.1
jupyter-core 4.7.1
jupyter-http-over-ws 0.0.8
jupyter-lsp 1.5.0
jupyter-packaging 0.10.4
jupyter-server 1.6.4
jupyterlab 3.1.7
jupyterlab-lsp 3.9.1
jupyterlab-pygments 0.1.2
jupyterlab-server 2.8.2
keras 2.6.0
Keras-Preprocessing 1.1.2
kiwisolver 1.3.2
lazy-object-proxy 1.6.0
Markdown 3.3.4
MarkupSafe 2.0.1
matplotlib 3.4.2
matplotlib-inline 0.1.2
mccabe 0.6.1
mistune 0.8.4
msgpack 1.0.2
msgpack-numpy 0.4.7.1
multidict 5.1.0
nbclassic 0.2.6
nbclient 0.5.3
nbconvert 6.1.0
nbformat 5.1.3
nest-asyncio 1.5.1
networkx 2.6.3
notebook 6.4.5
numexpr 2.7.3
numpy 1.19.5
oauthlib 3.1.1
olefile 0.46
openpyxl 3.0.5
opt-einsum 3.3.0
packaging 21.0
pandas 1.3.1
pandocfilters 1.4.3
parso 0.7.0
pexpect 4.8.0
pickleshare 0.7.5
Pillow 8.4.0
pip 21.3
pluggy 0.13.1
ply 3.11
prometheus-client 0.11.0
promise 2.3
prompt-toolkit 3.0.20
protobuf 3.17.2
ptyprocess 0.7.0
pyasn1 0.4.8
pyasn1-modules 0.2.8
pybind11 2.8.1
pycodestyle 2.6.0
pycosat 0.6.3
pycparser 2.20
pydocstyle 6.1.1
pyflakes 2.2.0
pyglet 1.5.21
Pygments 2.10.0
PyJWT 2.1.0
pylint 2.9.6
pyOpenSSL 21.0.0
pyparsing 3.0.4
pyrsistent 0.18.0
PySocks 1.7.1
python-dateutil 2.8.2
python-jsonrpc-server 0.4.0
python-language-server 0.36.2
pythran 0.9.11
pytz 2021.3
pyzmq 20.0.0
requests 2.26.0
requests-oauthlib 1.3.0
rope 0.19.0
rsa 4.7.2
ruamel-yaml-conda 0.15.100
scikit-learn 1.0.1
scipy 1.7.1
seaborn 0.11.2
Send2Trash 1.8.0
setuptools 52.0.0.post20210825
six 1.15.0
sniffio 1.2.0
snowballstemmer 2.1.0
tensorboard 2.6.0
tensorboard-data-server 0.6.0
tensorboard-plugin-wit 1.6.0
tensorflow-datasets 4.4.0
tensorflow-estimator 2.7.0
tensorflow-macos 2.6.0
tensorflow-metadata 1.4.0
tensorflow-metal 0.2.0
Tensorforce 0.6.5
termcolor 1.1.0
terminado 0.9.4
testpath 0.5.0
threadpoolctl 3.0.0
toml 0.10.2
tomlkit 0.7.0
tornado 6.1
tqdm 4.62.3
traitlets 5.1.0
typing-extensions 3.7.4.3
ujson 4.0.2
urllib3 1.26.7
wcwidth 0.2.5
webencodings 0.5.1
Werkzeug 2.0.2
wheel 0.35.1
wrapt 1.12.1
yapf 0.31.0
yarl 1.6.3
zipp 3.6.0
Update: I ran exactly the same LSTM code on Macbook Pro M1 Pro and Macbook Pro 2017, It turns out M1 Pro costs 6 hrs for one epoch, and 2017 model only needs 158s.