-
Hello ESM team,
Thank you for releasing the ESM3 code! I have a question regarding the training protocol you have described for the VQ-VAE model.
In the paper, you describe a two-stage training …
-
int requestedAlgoCount = CUDNN_CONVOLUTION_FWD_ALGO_COUNT;
int returnedAlgoCount = -1;
cudnnConvolutionFwdAlgoPerf_t fwd_results[2 * requestedAlgoCount];
checkCUDNN(cudnnFindConvolutionForwardAlg…
-
#### Issue description
I am trying to run some experiments with Variational Quantum Circuits.
Curiously training seems to be consistently **slower** when done on GPU, irrelevant of the used back…
-
I noticed in Table 12 of your paper, the hyperparameter ($\beta$) is set to a very low value, 1e-8, which suggests that the proposed code-based distillation process plays an almost negligible role dur…
-
:dancers: :dancers: :dancers: :dancers: :dancers:
:lollipop:
-
Looking at the description of the arguments of `diffeqsolve` closely, I realized that irrespective of choosing the solver to be `diffrax.ReversibleHeun`, the backward pass by default is `diffrax.Recur…
-
This corresponds to the encoder-decoder of the approach of time series forecasting (See #1 )
![image](https://github.mit.edu/storage/user/26647/files/cba7cf36-375a-44b9-82f4-c57193b2a96b)
Althou…
-
Currently, the `EigenmodeCoefficient` class of the adjoint solver defines the DFT field monitors **on the Yee grid**:
https://github.com/NanoComp/meep/blob/434f3fbf0ffe2334ce99cf105c6ec94ed44c95b7/…
-
first of all, it's a nice work definitely.
I use it to generate different texture for car models
it looks nice from far away
![image](https://github.com/LIU-Yuxin/SyncMVD/assets/17668601/d78d44fb-5…
-
CUDA supports:
https://github.com/kimlimjustin/xplorer/blob/master/src/Service/app.ts
https://github.com/launchbadge/sqlx
https://github.com/Jimver/cuda-toolkit
https://github.com/LLukas22/llm-r…