For the complete documentation index, see llms.txt. Markdown versions of all pages are available by appending .md to any URL (e.g. /get-started.md).

Python class

GenerateResult

`GenerateResult`

class max.pipelines.lib.generate.GenerateResult(sequences, logits=None, log_probs=None)

source

The result of generate().

Parameters:

sequences (list[list[int]])
logits (list[ndarray[tuple[Any, ...], dtype[float32]]] | None)
log_probs (list[LogProbabilities] | None)

log_probs: list[LogProbabilities] | None = None

source

Each row’s log probabilities, as GenerateIterator.log_probs, when requested.

logits: list[ndarray[tuple[Any, ...], dtype[float32]]] | None = None

source

Each row’s next-token logits, as GenerateIterator.logits, when requested.

sequences: list[list[int]]

source

Each row’s prompt followed by its new tokens, in input order, as in Hugging Face generate.