What it does
Requests use TypeSafe's systemone shape, and the answers match the Python reference to four decimal places. Weights are about 1.7 GB, cached after the first download.
SDKs & clients
A Node package that loads the open Laya decision model through ONNX, no Python needed at runtime.
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
Requests use TypeSafe's systemone shape, and the answers match the Python reference to four decimal places. Weights are about 1.7 GB, cached after the first download.
Start by planning the questions you want answered about incoming text. For customer messages, list team names like billing or support and define their jobs. You can also create an urgency scale or write yes-or-no questions, such as whether a customer might leave.
A developer can then set up the tool to evaluate text right on your machine. Your data stays private because it never travels to an outside service. The computer will need about two gigabytes of memory, and long messages get cut off if they exceed the length limit.