Layla uses an emerging technology, allowing "ChatGPT-like" chatbots to run on your phone locally without connection to the internet. So it requires quite a lot of resources.
The expected response time for Layla Full on a high end phone such as the S23 Ultra with 12GB RAM is around 1-2 words a second. If you have a high end phone, and you are experiencing slow response times, try one or all of these methods below:
Close all background apps
Turn off "battery optimisation" for Layla: Navigate to Settings > Apps & notifications > Layla > Advanced > Battery > Battery optimization, and then choose "Don't optimize."
Turn on "high performance mode" or "gaming mode" in your phone settings if you have it
Switch to the "Q2" base model. It has roughly the same intelligence as Full Layla but with a performance boost
RAM is only one part of the equation: if you have enough, any more does not help. The rest is up to your CPU.
Here's a rough estimate on the response times you can expect:
12nm CPU chipset: Tiny model does 1-2 words per second, any other model pretty unusable
6nm CPU chipset: Lite model does 1-2 words per second, Full model may do a few words per minute
4nm CPU chipset: Full model does 1-2 words per second. Lite model is almost as fast as cloud AI providers
For more information on which phones Layla supports, you can check here: https://www.layla-network.ai/contact
Commentaires