-
## 🐛 Bug
Using the **--print-alignment** argument makes the generation up to 3x times slower (Depends on the batch size). For example, generating translation for my test set took 47.9s (134.54 sent…
-
Since you added OpenGL I thought I'd try and feedback for v0.9alpha1.
I have a dedicated CAD GPU (nvidia Quadro) which is quite friendly with OpenGL. The only difference I can notice is that pannin…
-
OS: `Darwin Kernel Version 19.0.0: Wed Sep 25 20:18:50 PDT 2019; root:xnu-6153.11.26~2/RELEASE_X86_64`.
If add latest version (`0.14.0`) with `features = ["tensorflow_gpu"]` to `Cargo.toml`, most o…
-
```
Aparapi's current goal is to be yet another
easily-write-a-crappy-GPGPU-implementation framework. These are a dime a dozen,
are useless, ignored by mainstream developers, and shunned by HPC deve…
-
I tried using two separate workflows and I've encountered the issue anyways.
When looking at my RAM (not VRAM) during the generations, it's always rising, but never goes down. Even with 16GB + 10GB …
-
I have pulled this project out of a thin air and thought that columnar datastore can be one upped by graph data store, and on top of it with indexes ( distributed ).
The cool part is it's just an opt…
-
Hi,
Thank you for this amazing package and it's very convenient to use. In my recent work we proposed a [differentiable LP solver](https://arxiv.org/pdf/2004.14539.pdf) which can be easily implemen…
-
Hi, thanks for your great work! I am wondering why during the training, the original CSPN is used. Because as you mentioned in the paper, the new implemented one is much faster.
-
how to calculate Euclidean distance using GPU ?
I want to calculate Euclidean distance between 10000000x512 matrix and 10000000x512 matrix .
Can I use GPU ?? and how to do this??
zsz00 updated
5 years ago
-
Hello, I am currently using the AMD Instinct MI50 GPU to train models. It has 26 Tflops of fp16 and 13 Tflops of fp32 compute power, but it lacks tensor cores.
My experiments on PyTorch indicate th…