thomassutter / mopoe Goto Github PK
View Code? Open in Web Editor NEWCode release for ICLR 2021 paper "Generalised Multimodal ELBO"
Code release for ICLR 2021 paper "Generalised Multimodal ELBO"
Thanks for your great paper. I have a confusion about the PoE part and wonder if you can clarify it.
In the paper, Definition 1 shows that the posterior is made by PoE, but i don't realize how we did it.
I tried to find the answer in the code, and came down to poe
It seems to be adding mus and vars but not product it,what is the meaning of it?
I would apologize if it's a naive question, since i just study ML for a short time,and i will appreciate it if you could clarify it.
Thank a lot.
Thanks for producing this interesting paper. I have a confusion about the MoE part and wonder if you can clarify it.
In the paper, Eq (6) is the objective to be optimized, which involves KL(sum_i expert_i || prior)
. Suppose expert_i is Gaussian, how do you compute the KL between a mixture-of-gaussian and the prior? I don't think this has a closed form.
I tried to find the answer in the code, and came down to moe_fusion and
mixture_component_selection. It seems to be performing some sort of sampling. Is this the same as the importance sampling in MMVAE (Shi 2019)?
Any clarification would be much appreciated. Thank you.
File "MoPoE/eval_metrics/sample_quality.py", line 16, in
import prd_score.prd_score as prd
ModuleNotFoundError: No module named 'prd_score'
Hi Thomas,
I trained your code on the polymnist dataset, and got error every time in MoPoE/utils/TBLogger.py line 66 when the training epoch progress is 8.0%.
`File "/data/home/chenxiaoyu/code/VAE/MoPoE/utils/TBLogger.py", line 66, in write_lhood_logs
self.writer.add_scalars('Likelihoods/%s'%
File "/data/home/chenxiaoyu/anaconda3/envs/pytorch_1 12.1_py 9/lib/python3.9/site-packages/tensorboardX/writer.py", line 86, in add_scalars
fw = FileWriter(logdir=fw_tag)
File "/data/home/chenxiaoyu/anaconda3/envs/pytorch_1 12.1_py 9/lib/python3.9/site-packages/tensorboardX/writer.py", line 103, in init
self.event_writer = EventFileWriter(
File "/data/home/chenxiaoyu/anaconda3/envs/pytorch_1.12.1_py39/lib/python3.9/site-packages/tensorboardX/event_file_writer.py", line 105, in init
self._event_queue = multiprocessing.Queue(max_queue_size)
File "/data/ho e/chenxiaoyu/anaconda3/envs/pytorch_1.12.1_py39/lib/python3.9/multiprocessing/context.py", line 103, iQueue
return Queue(maxsize, ctx=self.get_con ext())
File "/data/home/chenxiaoyu/anaconda3/envs/pytorch_1.12.1_py39/lib/python3.9/multiprocessing/queues.py", line 42, in init
self. reader, self._writer = connection.Pipe(duplex=False)
File "/data/ho e/chenxiaoyu/anaconda3/envs/pytorch_1.12.1_py39/lib/python3.9/multiprocessing/connection.py", line 532in Pipe
fd1, fd2 = os.pipe()
OSError: [Errn 24] Too many open files`
Could you please give me some advices? @thomassutter
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ 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. ๐๐๐
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 โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.