Skip to main content

gen

Given a path to a pre-trained (.pt) MODEL, it writes an OUTPUT file with sequential data generated by the model auto-regressively—i.e., feeding the output back as input. The generated data can be written in any of the following formats: .txt, .json, or .csv.

Usage

hxmx gen <MODEL> [OUTPUT] [-n <INTEGER>] [-s <INTEGER>] [-d <CHOICE>] [--debug] [--help]

Arguments

NameTypeRequiredDefault
MODELpath — must exist, file
OUTPUTpath — fileout.csv

Options

NameTypeDefaultDescription
--tokens, -ninteger100Number of tokens to generate.
--seed, -sinteger0Random seed. Use 0 for non-deterministic results.
--device, -dchoice (auto|cpu| ...)cpuComputing device. To list all available devices, run hxmx devices.
--debugbooleanFalseDebug mode. Includes traceback when an error is raised.
-h, --helpbooleanFalseOpen documentation in browser.