-
Hi everyone:
I am researching source code of caffe about `BaseConvolutionLayer::backward_cpu_gemm` function. I found that the BP algorithm of convolution, in the book, need to rotate the weight 18…
-
Hello.
I am trying to run the energy calculations of the conventional output stationary systolic array where the input activations and weights are skewed to exploit the data reuse by moving the dat…
-
right now, if you have weights constructed using an `idVariable` or `ids` list, the `plot` method fails.
This is because it is using the `iloc` method to do lookups based on iteration indices.
…
-
Hi all,
Thanks for your work! I am trying to fine-tune the layout model you provide on my data, but I get the following warning while launching the fine-tuning command:
```
$ python -m models.bas…
-
My code stops at second steps:
(img2img-turbo) D:\Charis\img2img-turbo-main>accelerate launch --main_process_port 29501 src/train_cyclegan_turbo.py --pretrained_model_name_or_path="stabilityai/sd-t…
-
Hi, is there any way to generate the output attention maps of `model.transformer.decoder.layers[i].cross_attn` layer? when I follow the referenced functions, I finally get stuck in `MSDA.ms_deform_att…
-
Hello! I recently started working with timeloop, and I noticed that both `construct_workloads.py` and `cnn_layers.py` in `timeloop-accelergy-exercises/workspace/baseline_designs/scripts` are configura…
xpww updated
7 months ago
-
#### 问题描述:
在微调InternLM XComposer过程中,遇到了如下报错信息。报错指出期望的标量类型为Float,但实际找到的是BFloat16。不使用DeepSpeed直接使用python运行无报错
#### 报错位置及代码片段:
```python
{'loss': 14.9467, 'grad_norm': 94.78041076660156, 'learning_…
-
Hi Peter,
In practice, it is frequent to perform a two-stage analysis:
1. get one BLUP per individual via a somewhat sophisticated mixed model (e.g. to account for spatial and/or temporal covari…
-
Hi,
would you be keen on storing the data in some open formats alongside rda so we could link to it from Python? We have a `geodatasets` package that holds metadata and some tooling to cache the da…