Skip to navigation
ReferenceFull Library ReferenceNemo AutomodelNemo AutomodelComponentsDatasetsVlmnemo_automodel.components.datasets.vlm.conversation

nemo_automodel.components.datasets.vlm.conversation

View as Markdown

Raw conversation records shared by VLM builders, tokenization, and packing.

Module Contents

Classes

NameDescription
ConversationMessageA role and its text or structured media content.
ConversationSampleRaw conversation with optional media metadata and token estimate.

Data

MediaValue

API

class nemo_automodel.components.datasets.vlm.conversation.ConversationMessage

Bases: typing.TypedDict

A role and its text or structured media content.

content
str | list[dict[str, MediaValue]]
role
str
class nemo_automodel.components.datasets.vlm.conversation.ConversationSample

Bases: typing.TypedDict

Raw conversation with optional media metadata and token estimate.

_text_tokens
int
conversation
list[ConversationMessage]
mm_inputs_meta
dict[str, MediaValue]
nemo_automodel.components.datasets.vlm.conversation.MediaValue: TypeAlias = str | int | float | bool | None | bytes | Image.Image | torch.Tensor | list['Med...