4 ms·
Creator here. Thank you for sharing!
by AMICABoard 3y ago
Creator here. Thank you for sharing!
- xcdzvyn 3y agoWhat does cat /dev/llama do? Would I get a kind of LLM stream of consciousness? That's incredible :-)
- AMICABoard 3y agocat /dev/llama is not yet implemented. However we have a module to which you can give a prompt as parameter. It's buggy now. Our goal is to write a proper kernel module to implement three things: 1. a character device 2. 1st backend is a LLMZip ie you write to say /dev/l2ezip, you get a compressed stream out 3. 2nd backend is a LLM, ie you write a prompt to say /dev/llama2, you get a completion back So the 1st backend could be useful for compressed telemetry The second backend could be useful for IoT LLM, or our ambitious plan of responding to telemetry, ie take action, such as control motor speed etc.
- ingenieroariel 3y agoPretty cool! > My research goal is to train models using various hardware telemetry data with the hope that the models learn to interpret sensor inputs and control actuators based on the insights they glean from the sensor inputs. This research direction may open up exciting possibilities in fields such as automation, space, robotics and IoT, where L2E can play a pivotal role in bridging the gap between AI and physical systems. This part was easy to miss but quite interesting, could you expand a bit here? What does L2E stand for?
- throwaway10965 3y agoIt's the name of the project - Llama2Everywhere.
- ingenieroariel 3y agothanks! I am still curious about the sensor inputs part. I am trying to replace the 2.4ghz controller on my electric skateboard to make 0 to 5kmh and braking more pleasant and maybe use gyroscopes to do away with the controller altogether. What would tokens be in that case? Do you create a CAN style representation and feed that to the llm? What kind of throughput do you foresee being possible on which hardware?
- AMICABoard 3y agoI am still thinking how I'll pull this off. But basically it is collect tons of Telemetry converted to in ascii/text. Telemetry in sense sensor streams, both command and responses. Then we'll just train a small model for long enough. Then we will see how it would respond. That's the plan sort of.
- datadrivenangel 3y agoIn this case, I would ask the LLM to suggest an algorithm to minimize acceleration, jerk and snap based on the expected sensor input data, and then just implement that. Probably in memory on whatever runs the board. Straightforward control problem of bringing the board from 5-0kmh smoothly?
- AMICABoard 3y agoLet's see.
- ilc 3y agoBasic control theory will work better than AI here. The mathematical models used in control theory have been used in computing since at least the 50's (Kalman Filters). I suspect you won't have issues with computational power. Figuring out exactly which model to use and how, may take some work. Also understanding control theory will allow you to do things like traction control, etc.
- AMICABoard 3y ago
- AMICABoard 3y agoCorrect :)
- AMICABoard 3y agoThank you. So L2E stands for Llama 2 Everywhere. Thanks again for checking it out.
- pmarreck 3y ago> What does L2E stand for? um...
- leonheld 3y agoThis is hilarious (in a good way), I love it. Thanks for creating it.
- AMICABoard 3y ago:)
- bcjordan 3y agoCan one write installable apps for this OS that Llama can call out to? We need to go deeper!
- AMICABoard 3y agoWe will go the depth. In the coming versions.
- jamal-kumar 3y agoEasily the first practical use I've seen after hearing about cosmo libc. Very cool!
- AMICABoard 3y ago:)
- sbassi 3y agothe iso is 68mb or so, so the model is not there, how does it work? (tried on Oracle VM and didn't work but seems it is not supported)
- AMICABoard 3y agothe model is included
- AMICABoard 3y agotry qemu or boot from pendrive Qemu: qemu-system-x86_64 -display gtk,zoom-to-fit=off -m 512 -accel kvm -vga virtio -cdrom l2eos.iso