I am trying to replicates the code from this page.
At my workplace we have access to transformers and pytorch library but cannot connect to internet from our python environment. Could anyone help with how we could get the script working after manually downloading files to my machine?
my specific questions are -
should I go to the location bert-base-uncased at main and download all the files? Do I have put them in a folder with a specific name?
How should I change the below code
# Load pre-trained model tokenizer (vocabulary)
tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
# Tokenize our sentence with the BERT tokenizer.
tokenized_text = tokenizer.tokenize(marked_text)
How should I change the below code
# Load pre-trained model (weights)
model = BertModel.from_pretrained('bert-base-uncased',
output_hidden_states = True, # Whether the model returns all hidden-states.
)
Please let me know if anyone has done this…thanks
###update1
I went to the link and manually downloaded all files to a folder and specified path of that folder in my code. Tokenizer works but this line model = BertModel.from_pretrained('bert-base-uncased', output_hidden_states = True, # Whether the model returns all hidden-states. ) fails. Any idea what should i do? I noticed that 4 big files when downloaded have very strange name...should I rename them to same names as shown on the above page? Do I need to download any other files?
the error message is OSErrr: unable to load weights from pytorch checkpoint file for bert-base-uncased2/ at bert-base-uncased/pytorch_model.bin If you tried to load a pytroch model from a TF 2 checkpoint, please set from_tf=True
clone the model repo for downloading all the files
git lfs install
git clone https://huggingface.co/bert-base-uncased
# if you want to clone without large files – just their pointers
# prepend your git clone with the following env var:
GIT_LFS_SKIP_SMUDGE=1
git usage:
download git from here https://git-scm.com/downloads
paste these to your cli(terminal):
a. git lfs install
b. git clone https://huggingface.co/bert-base-uncased
wait for download, it will take time. if you want monitor your web performance
find the current directory simply pasting cd to your cli and get the file path(e.g "C:/Users/........./bert-base-uncased" )
use it as:
from transformers import BertModel, BertTokenizer
model = BertModel.from_pretrained("C:/Users/........./bert-base-uncased")
tokenizer = BertTokenizer.from_pretrained("C:/Users/........./bert-base-uncased")
Manual download, without git:
Download all the files from here https://huggingface.co/bert-base-uncased/tree/main
Put them in a folder named "yourfoldername"
use it as:
model = BertModel.from_pretrained("C:/Users/........./yourfoldername")
tokenizer = BertTokenizer.from_pretrained("C:/Users/........./yourfoldername")
For only model(manual download, without git):
just click the download button here and download only pytorch pretrained model. its about 420mb
https://huggingface.co/bert-base-uncased/blob/main/pytorch_model.bin
download config.json file from here https://huggingface.co/bert-base-uncased/tree/main
put both of them in a folder named "yourfilename"
use it as:
model = BertModel.from_pretrained("C:/Users/........./yourfilename")
Answering "###update1" for the error: 'OSErrr: unable to load weights from pytorch checkpoint file for bert-base-uncased2/ at bert-base-uncased/pytorch_model.bin If you tried to load a pytroch model from a TF 2 checkpoint, please set from_tf=True'
Please try this methods from -> https://huggingface.co/transformers/model_doc/bert.html
from transformers import BertTokenizer, BertForMaskedLM
import torch
tokenizer = BertTokenizer.from_pretrained("C:/Users/........./bert-base-uncased")
model = BertForMaskedLM.from_pretrained("C:/Users/........./bert-base-uncased")
inputs = tokenizer("The capital of France is [MASK].", return_tensors="pt")
labels = tokenizer("The capital of France is Paris.", return_tensors="pt")["input_ids"]
outputs = model(**inputs, labels=labels)
loss = outputs.loss
logits = outputs.logits
if this works we understand that there is nothing wrong with filesystem or foldernames.
If it works try to get hiddenstate after(note that bert model already returns hiddenstate as explained: " The bare Bert Model transformer outputting raw hidden-states without any specific head on top." so you dont need to use "output_hidden_states = True,")
from transformers import BertTokenizer, BertModel
import torch
tokenizer = BertTokenizer.from_pretrained("C:/Users/........./bert-base-uncased")
model = BertModel.from_pretrained("C:/Users/........./bert-base-uncased")
inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
outputs = model(**inputs)
last_hidden_states = outputs.last_hidden_state
if this not works try to load pytorch model with one of these methods
# Load all tensors onto the CPU
torch.load("C:/Users/........./bert-base-uncased/pytorch_model.bin", map_location=torch.device('cpu'))
# Load all tensors onto GPU 1
torch.load("C:/Users/........./bert-base-uncased/pytorch_model.bin", map_location=lambda storage, loc: storage.cuda(1))
if pytorch load method is not worked, we understand that there is pytorch version compatibility problem between pytorch 1.4.0 and released bert pytorch model. Or maybe your pytorch_model.bin file not downloaded very well. And please pay attention when pytorch 1.4.0 released the last python was python3.4
Related
I have this model: https://github.com/williamyang1991/DualStyleGAN and try to convert it to CoreML. So far I create copy of original Colab notebook and append at the end two blocks:
!pip install coremltools
import coremltools as ct
and
##title Convert inverted image.
inverted_latent = torch.Tensor(result_latents[0][4]).cuda().unsqueeze(0).unsqueeze(1)
with torch.no_grad():
net.eval()
[sampled_src, sampled_dst] = net(inverted_latent, input_is_latent=True)[0]
traced_model = torch.jit.trace(net, inverted_latent)
mlmodel = ct.convert(traced_model, inputs=[ct.ImageType(name="input", shape=inverted_latent.shape,bias=[-1,-1,-1],scale=2.0/255.0)])
mlmodel.save("modelsaved.mlmodel")
To run it, you should put any image with face to /content and in /usr/local/lib/python3.7/dist-packages/torchvision/transforms/functional.py
replace round method at 545, 546 lines with np.round
But then it fails at
mlmodel = ct.convert(...
with:
RuntimeError: PyTorch convert function for op 'pythonop' not implemented.
I suggest that there the way to rewrite this module with methods that could be convert, am I right? But I can't to figure out how to find the source of this module.
So my question is:
If I think in a right way, how I can find the source of module?
And if I wrong, please advise me the right way to do it.
The code starts by loading the model into PyTorch.
The code then converts the model into CoreML format and saves it to a .mlmodel file.The code below will take the existing PyTorch model and convert it into a CoreML model with input and output features.
The outputs are saved in the file example.mlmodel which can be opened in Xcode or any other development environment that supports CoreML models.
import torch
import coremltools
model = torch.load('MyPyTorchModel.pt')
coreml_model = coremltools.converters.pytorch.from_pytorch(model,
input_features=
['input'],
output_features=
['output'])
coreml_model.save('MyCoreMLModel.mlmodel')
This is my environment and parameter of xgboost when I saved a xgboost model.
environment
xgboost version 0.90
parameter
objective: gpu:binary:logistic
I want to load a saved xgboost model in version 1.0.0. However, this error massage is occurred.
XGBoostError: [12:51:46] C:\Users\Administrator\workspace\xgboost-win64_release_1.0.0\src\objective\objective.cc:26: Unknown objective function: `gpu:binary:logistic`
This is because gpu:binary:logistic is now deprecated. When I loaded my model in version 0.90, this warnings is occurred.
[13:45:14] WARNING: C:/Jenkins/workspace/xgboost-win64_release_0.90/src/objective/regression_obj.cu:170: gpu:binary:logistic is now deprecated, use binary:logistic instead.
I tried to apply set_params for replacing gpu:binary:logistic to binary:logistic and saved a model. But, it didn't work.
import xgboost as xgb
model = xgb.XGBClassifier()
model.load_model('xgb.model')
params = {'objective':'binary:logistic'}
model.set_params(**params)
I tried to replace 'gpu:binary:logistic' to 'binary:'logistic' from a saved model file, but OSError was occurred.
temp = open('test.model','rb')
lines = temp.readlines()
lines[0] = lines[0].replace(b'gpu:',b'')
new_model = open('test_replace.model','wb')
for line in lines:
new_model.write(line)
new_model.close()
And OSError was appeared when I loaded model.
model = xgb.XGBClassifier()
model.load_model('test_replace.model')
OSError: [WinError -529697949] Windows Error 0xe06d7363
This is a saved xgboost model. This file was saved as model format
b'\x00\x00\x00\x80\x19\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x13\x00\x00\x00\x00\x00\x00\x00gpu:binary:logistic\x06\x00\x00\x00\x00\x00\x00\x00gbtree\xfb\x00\x00\x00\x01\x00\x00\x00\x19\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x19\x00\x00\x00\x00\x00
I can't re-train my model. For this reason, I have to fix this problem from a saved model.
How to fix this problem?
The name of the objective function is stored as a string in the model file. Simply perform a binary-aware replacement gpu:binary:logistic -> binary:logistic on your model file!
Should be doable using command-line tools, see for example this SO thread: binary sed replacement
Upon a successful replacement, the size of the model file should shrink by 4 bytes (the length of the gpu: prefix).
I trained a Pipeline model, which uses CountVectorizer, TfidfTransformer, OneVsRestClassifier and also a GridSearchCV.
Now I want to save it into a tflite file, to use it on my Android app.
For a Sequential model (where my tflite file was created successfully), I did:
sequential_model = Sequential()
...
# train and fit the model
...
h5_file = "h5_model.h5"
tflite_file = "tflite_model.tflite"
sequential_model.save(h5_file)
converter = tf.lite.TFLiteConverter.from_keras_model_file(h5_file)
tflite_model = converter.convert()
open(tflite_file, "wb").write(tflite_model)
All good to save Sequential model into a tflite file.
Well, Pipeline has no attribute "save", unlike a Sequential model, so I tried saving the Pipeline model with joblib and then with pickle, but none of them worked.
Let's say that pipeline_model is my trained model (the one described in the first sentence).
pb_file = 'pipeline_model.pb'
# I also tried with other extensions, like h5, hdf5, sav, pkl
joblib.dump(pipeline_model, filename)
# or with pickle equivalent and pkl extension
# pickle.dump(pipeline_model, open(pb_file, 'wb'))
Now the pb file is created and I want to create a tflite one. Since it's not a Keras model, I can't use from_keras_model_file, so I tried instead with from_saved_model.
pb_file = 'pipeline_model.pb'
tflite_file = "tflite_model.tflite"
converter = tf.lite.TFLiteConverter.from_saved_model(pb_file)
tflite_model = converter.convert()
open(tflite_file, "wb").write(tflite_model)
It generates the error on line of converter = ...:
OSError: SavedModel file does not exist at: pb_file.pb/{saved_model.pbtxt|saved_model.pb}
I tried running it on Kaggle, Colab, PyCharm IDE, with both versions of tensorflow (1 and 2), with different file extensions and nothing seems to work.
I also noticed that TFLiteConverter contains the methods from_frozen_graph and from_session, but these two requires an extra parameter, so I don't think these could be the solution.
So, how can I obtain my tflite file from the trained Pipeline model? Please, if you find any solution, tell me the library versions that you used, since there could be a different behaviour on different libs.
Context:
I have a simple classifier based on tf.estimator.DNNClassifier that takes text and output probabilities over an intent tags. I am able to train an export the model to a serveable as well as serve the serveable using tensorflow serving. The problem is this servable is too big (around 1GB) and so I wanted to try some tensorflow graph transforms to try to reduce the size of the files being served.
Problem:
I understand how to take the saved_model.pb and use freeze_model.py to create a new .pb file that can be used to call transforms on. The result of these transforms (a .pb file as well) is not a servable and cannot be used with tensorflow serving.
How can a developer go from:
saved model -> graph transforms -> back to a servable
There's documentation that suggests that this is certainly possible, but its not at all intuitive from the docs as to how to do this.
What I've Tried:
import tensorflow as tf
from tensorflow.saved_model import simple_save
from tensorflow.saved_model import signature_constants
from tensorflow.saved_model import tag_constants
from tensorflow.tools.graph_transforms import TransformGraph
with tf.Session(graph=tf.Graph()) as sess_meta:
meta_graph_def = tf.saved_model.loader.load(
sess_meta,
[tag_constants.SERVING],
"/model/path")
graph_def = meta_graph_def.graph_def
other_graph_def = TransformGraph(
graph_def,
["Placeholder"],
["dnn/head/predictions/probabilities"],
["quantize_weights"])
with tf.Graph().as_default():
graph = tf.get_default_graph()
tf.import_graph_def(other_graph_def)
in_tensor = graph.get_tensor_by_name(
"import/Placeholder:0")
out_tensor = graph.get_tensor_by_name(
"import/dnn/head/predictions/probabilities:0")
inputs = {"inputs": in_tensor}
outputs = {"outputs": out_tensor}
simple_save(sess_meta, "./new", inputs, outputs)
My idea was to load the servable, extract the graph_def from the meta_graph_def, transform the graph_def and then try to recreate the servable. This seems to be the incorrect approach.
Is there a way to successfully perform transforms (to reduce file size at inference) on a graph from an exported servable, and then recreate a servable with the transformed graph?
Thanks.
Update (2018-08-28):
Found contrib.meta_graph_transform() which looks promising.
Update (2018-12-03):
A related github issue I opened that seems to be resolved in a detailed blog post which is listed at the end of the ticket.
I'm trying to build a CNN with Tensorflow (r1.4) based on the API tf.estimator. It's a canned model. The idea is to train and evaluate the network with estimator in python and use the prediction in C++ without estimator by loading a pb file generated after the training.
My first question is, is it possible?
If yes, the training part works and the prediction part works too (with pb file generated without estimator) but it doesn't work when I load a pb file from estimator.
I got this error : "Data loss: Can't parse saved_model.pb as binary proto"
My pyhon code to export my model :
feature_spec = {'input_image': parsing_ops.FixedLenFeature(dtype=dtypes.float32, shape=[1, 48 * 48])}
export_input_fn = tf.estimator.export.build_parsing_serving_input_receiver_fn(feature_spec)
input_fn = tf.estimator.inputs.numpy_input_fn(self.eval_features,
self.eval_label,
shuffle=False,
num_epochs=1)
eval_result = self.model.evaluate(input_fn=input_fn, name='eval')
exporter = tf.estimator.FinalExporter('save_model', export_input_fn)
exporter.export(estimator=self.model, export_path=MODEL_DIR,
checkpoint_path=self.model.latest_checkpoint(),
eval_result=eval_result,
is_the_final_export=True)
It doesn't work neither with tf.estimator.Estimator.export_savedmodel()
If one of you knows an explicit tutorial on estimator with canned model and how to export it, I'm interested
Please look at this issue on github, it looks like you have the same problem. Apparently (at least when using estimator.export_savedmodel) you should load the graph with LoadSavedModel instead of ReadBinaryProto, because it's not saved as a graphdef file.
You'll find here a bit more instructions about how to use it:
const string export_dir = ...
SavedModelBundle bundle;
...
LoadSavedModel(session_options, run_options, export_dir, {kSavedModelTagTrain},
&bundle);
I can't seem to find the SavedModelBundle documentation for c++ to use it afterwards, but it's likely close to the same class in Java, in which case it basically contains the session and the graph you'll be using.