-
Hi,
I am trying cell2location for my data.
followed the tutorial and got an error at estimation of reference signature.
here is execution code.
My computer has only cpu, I do not use gpu opt…
-
I am trying to run the below pre-train script using [MASS-Summarization](https://github.com/microsoft/MASS/tree/master/MASS-summarization) and fairseq and I get the below error. Please advise?
Wind…
-
I tried training the setup in order to replicate the results provided but the training is being stopped after few epochs (in exact after 101 epoch, it had saved epoch 99 checkpoint). The exact error I…
-
### Problem
After inputting the unnormalised raw counts, I try to train the regression model using adata_ref. However, it gave back ValueError.
### the data input
adata_ref.X
### the para…
-
Using this Trainer, [CoreMod][0] & [ZEreshkigal AllinOneMod][1] seems to cause a problem where the sound bugs away and if being shot it will cause a black screen that most of the time freezes the game…
-
Hello,
Since the beta 1.2 there is a bug and the mod doesn't work anymore. Could you update it ?
There is also a bug with the command max stat if you can do something with your mod.
Thank you :)
-
### Bug description
When compiling a lightning model to torchscript, a seemingly unrelated error is produced:
```
Traceback (most recent call last):
File "/home/train.py", line 148, in
t…
-
Can we use PL's QuantizationAwareTraining and ModelPruning callbacks during NeMo model training?
To quickly test this, I tried to add the QAT callback to the trainer in speech_to_text_ctc.py script…
-
Hi. I'm using this image "us-docker.pkg.dev/vertex-ai/training/tf-cpu.2-9:latest" in "Custom training job" ( Vertex AI). I get the error printed below. I think the problem is the version of tensorflow…
-
1. PaddleClas版本以及PaddlePaddle版本:请您提供您使用的版本号或分支信息,如PaddleClas 2.5和PaddlePaddle_gpu 2.4
2. 涉及的其他产品使用的版本号:no
3. 训练环境信息:
a. 具体操作系统,如Linux
b. Python版本号,如Python3.8
c. CUDA/cuDNN版本, 如CUDA11.7/c…