Comments (7)
We did not use relation matching for half KG. Can you give the exact commands you used? Did you use the pretrained KG Embeddings or trained them again?
from embedkgqa.
OK, thanks! I used the pretrained KG ebmeddings you provided, and I train them again, I find the hit@1 is just 0.2. These embedding I all try, but result is not good, .....
my command,
python3 main.py --mode train --relation_dim 200 --hidden_dim 256
--gpu 2 --freeze 0 --batch_size 128 --validate_every 5 --hops 2 --lr 0.0005 --entdrop 0.1 --reldrop 0.2 --scoredrop 0.2
--decay 1.0 --model ComplEx --patience 5 --ls 0.0 --kg_type half
from embedkgqa.
I think the good or bad about pretrained embedding is very important, I used your code and data in metaQA half KG, get the 0.2 hit@1 while full kg is 1.0 hit@1, so the result in half kg is too low, can you tell me the result about hit@1 in metaQA half KG?
from embedkgqa.
OK, thanks! I used the pretrained KG ebmeddings you provided, and I train them again,
Do you mean you are using pretrained embeddings or training them again? For pretrained, I am able to reproduce the following:
python3 main.py --mode train --relation_dim 200 --hidden_dim 256 \
--gpu 4 --freeze 0 --batch_size 256 --validate_every 5 --hops 1 --kg_type half --lr 0.001 --entdrop 0.2 --reldrop 0.3 --scoredrop 0.3 \
--decay 1.0 --model ComplEx --patience 5 --ls 0.1 --l3_reg 0.0001
0.82 hit@1 within 10 epochs (still not converged)
Similarly, for 2 hop (which takes longer to converge), I got the following within 5 epochs. It will take ~30 epochs to reach accuracy reported in paper but it's definitely not 0.2
python3 main.py --mode train --relation_dim 200 --hidden_dim 256 \
--gpu 3 --freeze 0 --batch_size 256 --validate_every 5 --hops 2 --kg_type half --lr 0.001 --entdrop 0.2 --reldrop 0.3 --scoredrop 0.3 \
--decay 1.0 --model ComplEx --patience 5 --ls 0.1 --l3_reg 0.0001
Maybe you could try by cloning a fresh copy of the repo and try the above commands? Or if you are retraining the KG embeddings, could you give me the commands you are using along with how you are creating the embedding files from the model checkpoint?
from embedkgqa.
OK, I think I found the reason, my parameters is not same with you! Thanks for your reply!! If I have probems again, I will contact you again! Thanks!
from embedkgqa.
By the way, can you apply the repretrained half KG model file in websqp and command, just like above?
from embedkgqa.
By the way, can you apply the repretrained half KG model file in websqp and command, just like above?
Yes I think you should be able to do that
from embedkgqa.
Related Issues (20)
- [MetaQA] Training data contains knowledge base triplets as questions HOT 2
- relation in fbwq_full
- relation matching
- MetaQA relation match HOT 1
- About ComplEx score calculation HOT 1
- The MetaQA dataset HOT 2
- Question with "half KG" protocol HOT 2
- How to build your own pruning_ train.txt
- Hello I have some questiones about the code such as what is the meaning of "best_valid" HOT 1
- Do you use the folder "train_embeddings"?
- I have some question in the relation matching
- When I run RoBERTa / main.py, running to ' creating model ' GPU takes up 0 and takes several hours to create the model.
- Is it possible to share the pdf version or ppt version slides
- How to use eval? How can I use pre trained model for QA? HOT 1
- RoBERTa used for question embedding
- Pretrained models missing HOT 1
- How to set up fbwq_full?
- miss files HOT 2
- About dataset
- pretrained_models.zip HOT 7
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from embedkgqa.