dreasysnail / inset Goto Github PK
View Code? Open in Web Editor NEWINSET: Sentence Infilling with Inter-sentential Transformer
INSET: Sentence Infilling with Inter-sentential Transformer
Hi Yichen,
Thanks for your released source code.
Would it be also possible to provide the pre-processed datasets for facilitating future research? Looking forward to hearing from you.
Best,
Dong
Have you release the code for sentence infilling with lexical-constraint? Could you give some instructions about this subtask?
Hi authors,
thanks for the sharing this nice work!
I am curious about how to evaluate the generated intermediate sentence when there is no human-labeled reference.
thanks for any tips.
looking forward to your reply.
Has the relevant code and data set been released recently?
Thank you for sharing your code and trained model.
I ran INSET_test.py using the README.md file as a reference, but I got the following error about loading the BertModel.
(LSP) $ python3 INSET_test.py
decode type: beam
Model name 'bert-base-uncased' was not found in model name list (bert-base-uncased, bert-large-uncased, bert-base-cased, bert-large-cased, bert-base-multilingual-uncased, bert-base-multilingual-cased, bert-base-chinese). We assumed 'https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-uncased-vocab.txt' was a path or url but couldn't find any file associated to this path or url.
Model name 'bert-base-uncased' was not found in model name list (bert-base-uncased, bert-large-uncased, bert-base-cased, bert-large-cased, bert-base-multilingual-uncased, bert-base-multilingual-cased, bert-base-chinese). We assumed 'https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-uncased.tar.gz' was a path or url but couldn't find any file associated to this path or url.
Traceback (most recent call last):
File "INSET_test.py", line 54, in <module>
model_bert = BertModel.from_pretrained('bert-base-uncased', state_dict=torch.load('models/BERT-pretrain-1-step-5000.pkl')).cuda()
AttributeError: 'NoneType' object has no attribute 'cuda'
To be sure, I checked the behavior of the version of "pytorch_pretrained_bert" in the conda virtual environment, and I got the same error.
(LSP) $ ipython
Python 3.6.9 |Anaconda, Inc.| (default, Jul 30 2019, 19:07:31)
Type 'copyright', 'credits' or 'license' for more information
IPython 7.8.0 -- An enhanced Interactive Python. Type '?' for help.
In [1]: import torch
In [2]: from pytorch_pretrained_bert import *
Better speed can be achieved with apex installed from https://www.github.com/nvidia/apex.
In [3]: model = BertModel.from_pretrained('bert-base-uncased')
Model name 'bert-base-uncased' was not found in model name list (bert-base-uncased, bert-large-uncased, bert-base-cased, bert-large-cased, bert-base-multilingual-uncased, bert-base-multilingual-cased, bert-base-chinese). We assumed 'https://s3.amazonaws.com/models.huggingface.co/bert/bert-base-uncased.tar.gz' was a path or url but couldn't find any file associated to this path or url.
Possibly, do I need to download "bert-base-uncased" from "pytorch_pretrained_bert" beforehand to get this code to work?
I would appreciate it if you know and could tell me how to avoid the error.
Hi, the link to the pre-trained checkpoints seems to be broken. Could you please set it up again? Thank you!
Hi, thanks for the well-documented code.
Can the code used for generating sents_derep_bert_train_mask.json, sents_derep_bert_train.json, sents_derep_gpt_train.json, sents_derep_gpt_test.json, trip_cut_train_denoising.json, trip_derep_val.json
from tripadvisor_review_processed_uncut.json
also be released?
Hello, I'm more interested in this paper, but the preprocessed data set, that URL crashed, how should I get the data set
Is there a way to generate more than just one middle sentence? Perhaps 3 or more? Is there also a way to feed more than three sentences to the preceeding and following contexts? Thank you for your help!
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.