wengong-jin / nips17-rexgen Goto Github PK
View Code? Open in Web Editor NEWPredicting Organic Reaction Outcomes with Weisfeiler-Lehman Network (NIPS 2017)
License: MIT License
Predicting Organic Reaction Outcomes with Weisfeiler-Lehman Network (NIPS 2017)
License: MIT License
Hi,
I am wondering if you plan on releasing a TensorFlow 2 version (or PyTorch) anytime soon? I guess chances are slim since this repo has not been updated for a couple of years, but asking can never hurt.
This is not an issue but having an apply model demo code would be great ;-)
I get the following error when I tried to run nntrain.py from core-wln-global.
To Reproduce
Run the following command:
python nntrain.py --train ../data/train.txt --hidden 320 --depth 8 --save_dir .
Error Description
Traceback (most recent call last):
File "nntrain.py", line 65, in <module>
atom_hiddens, _ = rcnn_wl_last(graph_inputs, batch_size=batch_size, hidden_size=hidden_size, depth=depth)
File "/home/sachin/Desktop/nips17-rexgen/USPTO/core-wln-global/models.py", line 32, in rcnn_wl_last
atom_features = tf.nn.relu(linearND(input_atom, hidden_size, "atom_embedding", init_bias=None))
File "/home/sachin/Desktop/nips17-rexgen/USPTO/core-wln-global/utils/nn.py", line 81, in linearND
target_shape = tf.concat(0, [X_shape, [output_size]])
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/ops/array_ops.py", line 1121, in concat
dtype=dtypes.int32).get_shape().assert_is_compatible_with(
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1048, in convert_to_tensor
as_ref=False)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1144, in internal_convert_to_tensor
ret = conversion_func(value, dtype=dtype, name=name, as_ref=as_ref)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/ops/array_ops.py", line 971, in _autopacking_conversion_function
return _autopacking_helper(v, dtype, name or "packed")
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/ops/array_ops.py", line 923, in _autopacking_helper
return gen_array_ops.pack(elems_as_tensors, name=scope)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/ops/gen_array_ops.py", line 4689, in pack
"Pack", values=values, axis=axis, name=name)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/framework/op_def_library.py", line 787, in _apply_op_helper
op_def=op_def)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/util/deprecation.py", line 488, in new_func
return func(*args, **kwargs)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 3272, in create_op
op_def=op_def)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1790, in __init__
control_input_ops)
File "/home/sachin/anaconda3/envs/reaction_pred/lib/python2.7/site-packages/tensorflow/python/framework/ops.py", line 1629, in _create_c_op
raise ValueError(str(e))
ValueError: Dimension 0 in both shapes must be equal, but are 2 and 1. Shapes are [2] and [1].
From merging shape 0 with other shapes. for 'encoder/concat/concat_dim' (op: 'Pack') with input shapes: [2], [1].
Hi jin,
Thanks for your work. I am a little bit confused about the data processing of this dataset. I've read about the old issue, but I still don't know how to find those wrong reactions whose atom mapping are incorrect.
Can you tell me the details of removing wrong atom-mapping reactions? And what's name of this atom-mapping tool? Thank you, sincerely.
Hi,
I was trying to reproduce the reaction center prediction result shown in the paper with core-wln-global/nntrain.py
, but the top 10 accuracy on the test dataset is about 5% lower than what you report in the report.
To reproduce the result, I used hidden=300
and depth=3
, which gives me 756k parameters. Everything else is set to the nntrain.py
default. Did I miss anything?
Thanks,
Wes
Hi,
Thanks for your work. And recently I am trying to reproduce the result of your paper.
I am a little bit confused about the dataset. There are 1,808,938 records in Lowe’s grants set, but I noticed you selected 480K records from them.
Could you please publish the data processing code or describe the algorithm?
Thanks in advance.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.