Jeremy Tregunna

@SirSamShaw Pull from the master branch in your Podfile instead, this fixed the issue for me.

@bettio so approximately how much memory and flash would you need to port atomvm to another board? I was thinking about doing that with one of mine this weekend until...

@teamon I've also run into this problem in my umbrella app, and have retried using master. I get the same error from master. Output: ``` The command '/bin/sh -c mix...

I manually copy all the apps as well. I don't recall how I worked around this issue, but I since have. Not helpful I know, so my apologies about that....

One possible solution for you, that works well for me... is define a schema to represent your function's arguments, and pass that into the system message, get json formatted response...

I have been able to get function calling (openai style) to work with dolphin-mistral:8x7b-v2.7-q8_0, probably works with others too but probably not them all. Here's the prompt I send: ```...

> CUDA_VISIBLE_DEVICES should work. We do have a defect related to memory prediction calculations in this case tracked via #1514 > > If you're seeing it load onto unexpected GPUs...

> @jeremytregunna it sounds like there might be an ordering/enumeration bug where we're not consistent with other tools. If I had to guess, I'd speculate this is some tools/libraries using...

> @jeremytregunna looking back on that screen shot you posted above, I think the problem may be a result of how you have your cards plugged into your PCI slots....

> Can you try setting `CUDA_DEVICE_ORDER` as well. Options are `FASTEST_FIRST` or `PCI_BUS_ID` It looks like you can also specify device UUIDs for the visible device setting which might help....