Skip to main content

← All model recipe tasks

eagle_vlm model family

Tasks

Supported architectures and task heads

The Hugging Face values below are copied from checkpoint metadata at the recorded revision. The TRTMC task contract comes from the exact E2E recipe. Architecture identity selects or describes a source graph; it does not imply that TRTMC reproduces every Hugging Face head with the same model_type. For example, an encoder recipe may consume the base model and intentionally return hidden states instead of the checkpoint's pretraining or classification logits.

HF model_typeHF architecture / pipeline classTRTMC task contractExact recipe profiles
llama_nemotron_vlLlamaNemotronVLModel
Source: config.architectures
Metadata: config.json at 582e3bf72aee
embedding
nemotron-embed-vl-1b-v2
nemotron-embed-vl-1b-v2-tp4
llama_nemotron_vl_rerankLlamaNemotronVLForSequenceClassification
Source: config.architectures
Metadata: config.json at 9c20c4aedf9e
reranking
nemotron-rerank-vl-1b-v2
nemotron-rerank-vl-1b-v2-tp4

Declared recipes

Each row comes from a model-owned E2E manifest. It declares a test recipe; it is not a live pass receipt for every hardware target.

RecipeExact Hugging Face checkpointTaskBuild configurationDeclared E2E casesManifest
nemotron-embed-vl-1b-v2nvidia/llama-nemotron-embed-vl-1b-v2fp16; single devicenemotron-embed-vl-1b-v2tests/e2e/models/eagle_vlm/manifests/nemotron-embed-vl-1b-v2.json
nemotron-embed-vl-1b-v2-tp4nvidia/llama-nemotron-embed-vl-1b-v2family default; TP4nemotron-embed-vl-1b-v2-tp4tests/e2e/models/eagle_vlm/manifests/nemotron-embed-vl-1b-v2-tp4.json
nemotron-rerank-vl-1b-v2nvidia/llama-nemotron-rerank-vl-1b-v2fp16; single devicenemotron-rerank-vl-1b-v2tests/e2e/models/eagle_vlm/manifests/nemotron-rerank-vl-1b-v2.json
nemotron-rerank-vl-1b-v2-tp4nvidia/llama-nemotron-rerank-vl-1b-v2family default; TP4nemotron-rerank-vl-1b-v2-tp4tests/e2e/models/eagle_vlm/manifests/nemotron-rerank-vl-1b-v2-tp4.json

Family-owned configuration

This family does not declare a family-owned --set namespace. Use the explicit CLI options shown below and the shared configuration namespaces documented in Configure Runtime Behavior.

Family-specific CLI contracts

Inputs and options below are filtered by the declared E2E task and the methods and configuration fields used by that family's native runtime implementation. The global CLI parser accepts a wider union of flags; flags absent here are not declared for this family.

trtmc embed

Return an embedding for text.

Declared recipes: nemotron-embed-vl-1b-v2, nemotron-embed-vl-1b-v2-tp4

trtmc embed <bundle.bundle> --prompt "<text>"
Supported input or optionRequirementRuntime behavior
--prompt <TEXT>RequiredText input passed to the embedding runtime.
Code basis

Runtime provider: eagle_vlm

  • include/trtmc/pipeline.h
  • src/cli/main.cpp
  • tests/e2e/models/eagle_vlm/manifests/nemotron-embed-vl-1b-v2-tp4.json
  • tests/e2e/models/eagle_vlm/manifests/nemotron-embed-vl-1b-v2.json

trtmc rerank

Score one document against a query.

Declared recipes: nemotron-rerank-vl-1b-v2, nemotron-rerank-vl-1b-v2-tp4

trtmc rerank <bundle.bundle> --prompt "<query>" --document "<document>"
Supported input or optionRequirementRuntime behavior
--prompt <TEXT>RequiredQuery text.
--document <TEXT>RequiredCandidate document text.
Code basis

Runtime provider: eagle_vlm

  • include/trtmc/pipeline.h
  • src/cli/main.cpp
  • tests/e2e/models/eagle_vlm/manifests/nemotron-rerank-vl-1b-v2-tp4.json
  • tests/e2e/models/eagle_vlm/manifests/nemotron-rerank-vl-1b-v2.json

Bundle lifecycle commands

These commands apply to every declared recipe in this family; they do not add task inputs or model capabilities.

trtmc build

Build one exact checkpoint into a TensorRT-Model-Connect bundle.

trtmc build <hf-id> --output <bundle.bundle>

trtmc inspect

Inspect bundle metadata, runtime identity, and packaged sections.

trtmc inspect <bundle.bundle>

See the CLI Reference for all options and limitations.