-
(update) as I mentioned in this [comment](https://github.com/tenstorrent-metal/tt-metal/issues/5168#issuecomment-1940293930), this issue is related to the TT_METAL_DEVICE_PROFILER env variable.
*…
-
hello, when i coverted my onnx model to TensorRT by the command,
`./trtexec --onnx=model.onnx --saveEngine=model.engine`
i got big diff between pytorch result and trt result. i located the problem w…
-
Boyan and I performance-tested the FFF-BERT (on HuggingFace) against a vanilla BERT of similar size, and found that it performs maybe 15% more slowly on my M2 mac.
https://gist.github.com/p-i-/3556…
p-i- updated
10 months ago
-
### 问题描述 Issue Description
安装https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/guides/hardware_support/npu/install_cn.html#wheel文档说明,安装完paddlepaddle==2.6.1 的CPU版,再cd PaddleCustomDevice/ba…
-
The test mentioned below is on the branch eland/whisper-perf
Unfortunately the size of the tensor_a.pt and tensor_b.pt are too large to put into git so they are attached as zip files to be downloaded…
-
For computational efficiency, I had specified c to have an identity link in the brms custom family, but coded it manually in the likelihood function on the log scale (relative to the model equations).…
-
-
https://github.com/wtnthu/FaCoR/
Criei um `requirements.txt` a partir do `environment.yml` porque não gosto de usar conda. Abaixo os detalhes
```
➜ FaCoR git:(main) ✗ workon facor
➜ FaCoR gi…
-
Why does the library use `sequence_length` as the first dimension of the input-tensor as opposed to the `batch_size`?
Is this just a choice of convention from RNNs or is the difference performance…
-
### 🐛 Describe the bug
Error with attention mask in torch.nn.TransformerEncoderLayer
Easy example:
```
if __name__=="__main__":
import torch
model = torch.nn.TransformerEncoderLaye…