Antoine Pultier

Results 166 comments of Antoine Pultier

@mikedizon I quickly made a very simple helm chart that you can find there: https://github.com/SINTEF/helm-charts/tree/main/charts/timescaledb-simple

I updated the comment with links to a random poll service. It might do the job :-)

I had a similar issue and it's because we both use `queue.Queue` that blocks the thread. when you do `msg = message_queue.get(block=True, timeout=None)`, you wait for a message to come...

As a workaround, I used the builtin `uri` module. The Azure API is quite simple to use with the shared access tokens. I used something like this, with my SAS...

> Bump. Following this You can also add a thumb up on the first message and click the "subscribe" button 🙂

It works but is it expected to get worse results? ## MPS on macbook m1: ## Cuda (Nvidia Tesla A30): ![image](https://user-images.githubusercontent.com/45740/230931788-725ebba2-d5d1-46b0-8050-a2dfd7550bd7.png)

> What is the VRAM on your GPU? By default FastChat sets the max memory to 13GiB when using multiple GPUs which can lead to a total of 26 with...

> For me it usually takes around 40GiB. It never worked with 2 GPUs @ 13GiB for me 🤔 what am I doing wrong? With two A30, `nvidia-smi` reports around...

It doesn't work for you ? I used this environment variable successfully.