## Stuck at Postfix Configuration during Lambda Stack install  
**Mar 4, 2025**  
Adding `I_AGREE_TO_THE_CUDNN_LICENSE=1` to the top of the installation script does it.

---

## Stuck at Postfix Configuration during Lambda Stack install  
**Mar 3, 2025**  
Dear Lambda folks & community,  
As part of the installation process, an ancient error appears when the in-terminal dialog for "Postfix Configuration" describing mail server configuration type options is not responsive to keyboard input. Arrows get typed as `^[B`, etc. and nothing happens on Tab.

---

## Is Lambda Stack free for use on non-LambdaLabs computers  
**Nov 23, 2024**  
Hi there,  
It’s not quite clear what the terms of using Lambda Stack on computers not purchased from Lambda Labs. I have my personal Quad workstation from Lambda Labs so this question hasn’t arisen for me before, but now I am asking on behalf of a startup I have joined.

---

## Quad workstation won't boot after long disuse  
**Feb 20, 2022**  
Hi there guys,  
I was abroad for many months and my QUAD no longer boots. I checked if the UPS had a problem by bypassing it but no. I opened the box to clean it, but that did fix anything. It’s just a dark screen. Only once did it get to setup, but I didn’t have a procedure to follow and exited.

---

## Lambda Stack seems to be wiped out after Ubuntu update  
**Jul 5, 2021**  
Hi Mark,  
Thanks for the reply. The commands did nothing for me. It looks like the stack was re-installed, but `nvidia-smi` is not found:  
```bash  
ivogeorg@turnaround:~$ sudo rm -f /etc/apt/sources.list.d/{graphics,nvidia,cuda}*;  
COLUMNS=200 dpkg -l |  
awk '/cuda|lib(accinj64|cu(blas|dart|dnn|fft|inj|pti...'
```

---

## Getting unhandled cuda error when trying to use TF distributed strategy for training on 4 GPUs  
**Jun 11, 2020**  
After fixing [this other error](https://deeptalk.lambda.ai/t/started-getting-tensorflow-profiler-error-after-cuda-update/1676) which was blocking replication of this one, this error has also gone away. It is possible that the software updates shown in the other error description were the solution.

---

## Solved  
The following worked for me (thanks, Jeremy@Lambda):  
In file `/etc/modprobe.d/nvidia-common-kernel.conf`, opened with `sudo`, write  
`options nvidia "NVreg_RestrictProfilingToAdminUsers=0"`  
Run the command  
```bash  
sudo update-initramfs -u -k all  
```  
Reboot.

---

## Brief  
Attempted to run distributed training on all 4 GPUs for the first time, using TF’s simple MirroredStrategy (which uses NCCL all-reduce). Errors cascade down from `CUPTI_ERROR_INSUFFICIENT_PRIVILEGES`. This is a brand new problem after the Ubuntu Updater updated some NVIDIA AMD packages today.

---

## How and where to add a missing library for Caffe2  
**Jul 25, 2018**  
Hi guys,  
How is Caffe2 installed as part of Lambda Stack, and how can I administer the installation (and the stack installations in general)? I get the following when I try to run a test for FB’s Detectron, for which I did all the installation steps except the Caffe2 part.