keras-onnx
keras-onnx copied to clipboard
onnx
Convert tf.keras/Keras models to ONNX
Hi Team, I was trying to convert a model with conv1d and padding type "same" when I countered below error: ``` Traceback (most recent call last): File "test_model.py", line 11,...
hello， here I first tried to convert my keras model to onnx format. when I check the nodes in graph，it seems all conv2d and batchnorm are merged in to one...
I'm trying to convert a keras h5 model(create with tf1.12) to ONNX: ```python from tensorflow.keras.models import load_model model = load_model(my_model_file) onnx_model = keras2onnx.convert_keras(model, model.name) onnx_model.save_model('model.onnx') ``` which report a error...
when I use tf.gather in tensorflow==1.14.0, it showed this warning Any solutions to solve this?
bert4keras 0.7.7 keras 2.3.1 tensorflow 1.14.0 keras2onnx 1.7.0 i want to convert the simbert pretrained model to onnx format with following code: `bert = build_transformer_model( config_path, checkpoint_path, with_pool='linear', application='unilm', keep_tokens=keep_tokens,...
I am using Ubuntu 18 and Tensorflow 2.5, ONNX 1.7, and Keras 2.4. I'm trying to convert Keras code to Onnx. I have Python code I tried to convert it...
Hello, I converted yolov4 keras model from [TensorFlow-2.x-YOLOv3](https://github.com/pythonlessons/TensorFlow-2.x-YOLOv3) to .onnx model, everything worked fine. Now following this tutorial [https://mxnet.apache.org/versions/1.5.0/tutorials/onnx/super_resolution.html](https://mxnet.apache.org/versions/1.5.0/tutorials/onnx/super_resolution.html) I tried to convert to mx model, and receiving the following...
from transformers import BertTokenizer, TFBertForSequenceClassification import tensorflow as tf tokenizer = BertTokenizer.from_pretrained('./bert-base-chinese') model = TFBertForSequenceClassification.from_pretrained('./bert-base-chinese') inputs = tokenizer("rch能保证是不是回因为环境差异带来运行错误。", return_tensors="tf") print(inputs) logits = model(inputs) print(logits) import keras2onnx onnx_model = keras2onnx.convert_keras(model, model.name)...
Hi, I have a bidirectional LSTM trained keras model (.h5) which is as under: [bidirectional_LSTM.zip](https://github.com/onnx/keras-onnx/files/6267667/bidirectional_LSTM.zip) I use a simple script for conversion: ```from tensorflow.python.keras import backend as K from tensorflow.python.keras.models...
Hello, I think there is a bug with the tensorflow max_pool_with_argmax op. When running the op with onnxruntime, I find that the op returns the pooled values twice instead of...
