Skip to content

feat: add Nemotron video-to-text capability #666

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
caihui-lol wants to merge 2 commits into
base: main
Choose a base branch
from
Open

feat: add Nemotron video-to-text capability #666

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
602f0b5
feat: add Nemotron video-to-text capability
Jul 10, 2026
95ad601
fix: limit Nemotron multimodal capabilities to Omni
Jul 21, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion pkg/apis/ome/v1beta1/model.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -286,7 +286,7 @@ const (
// ModelCapability enum
// TODO: Remove legacy capabilities
//
// +kubebuilder:validation:Enum=TEXT_GENERATION;TEXT_SUMMARIZATION;TEXT_EMBEDDINGS;TEXT_RERANK;CHAT;VISION;EMBEDDING;RERANK;TEXT_TO_TEXT;TEXT_TO_AUDIO;TEXT_TO_IMAGE;TEXT_TO_VIDEO;IMAGE_TEXT_TO_TEXT;IMAGE_TEXT_TO_AUDIO;IMAGE_TEXT_TO_IMAGE;IMAGE_TEXT_TO_VIDEO;VIDEO_TEXT_TO_AUDIO;AUDIO_TO_TEXT;AUDIO_TO_AUDIO;AUDIO_TEXT_TO_TEXT;AUDIO_TRANSLATION
// +kubebuilder:validation:Enum=TEXT_GENERATION;TEXT_SUMMARIZATION;TEXT_EMBEDDINGS;TEXT_RERANK;CHAT;VISION;EMBEDDING;RERANK;TEXT_TO_TEXT;TEXT_TO_AUDIO;TEXT_TO_IMAGE;TEXT_TO_VIDEO;IMAGE_TEXT_TO_TEXT;IMAGE_TEXT_TO_AUDIO;IMAGE_TEXT_TO_IMAGE;IMAGE_TEXT_TO_VIDEO;VIDEO_TEXT_TO_AUDIO;VIDEO_TEXT_TO_TEXT;AUDIO_TO_TEXT;AUDIO_TO_AUDIO;AUDIO_TEXT_TO_TEXT;AUDIO_TRANSLATION
type ModelCapability string

const (
Expand All @@ -310,6 +310,7 @@ const (
ModelCapabilityImageTextToImage ModelCapability = "IMAGE_TEXT_TO_IMAGE"
ModelCapabilityImageTextToVideo ModelCapability = "IMAGE_TEXT_TO_VIDEO"
ModelCapabilityVideoTextToAudio ModelCapability = "VIDEO_TEXT_TO_AUDIO"
ModelCapabilityVideoTextToText ModelCapability = "VIDEO_TEXT_TO_TEXT"
ModelCapabilityAudioToText ModelCapability = "AUDIO_TO_TEXT"
ModelCapabilityAudioToAudio ModelCapability = "AUDIO_TO_AUDIO"
ModelCapabilityAudioTextToText ModelCapability = "AUDIO_TEXT_TO_TEXT"
Expand Down
14 changes: 11 additions & 3 deletions pkg/modelconfig/capability.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@ const (
CapabilityTextToAudio Capability = "TEXT_TO_AUDIO"
CapabilityImageTextToAudio Capability = "IMAGE_TEXT_TO_AUDIO"
CapabilityVideoTextToAudio Capability = "VIDEO_TEXT_TO_AUDIO"
CapabilityVideoTextToText Capability = "VIDEO_TEXT_TO_TEXT"
CapabilityAudioToText Capability = "AUDIO_TO_TEXT"
CapabilityAudioToAudio Capability = "AUDIO_TO_AUDIO"
CapabilityAudioTextToText Capability = "AUDIO_TEXT_TO_TEXT"
Expand Down Expand Up @@ -166,14 +167,21 @@ func diffusionRule(hf HuggingFaceModel) []Capability {
}

func nemotronHNanoRule(hf HuggingFaceModel) []Capability {
if !strings.Contains(strings.ToLower(hf.GetModelType()), "nemotronh_nano") {
modelType := strings.ToLower(hf.GetModelType())
if !strings.Contains(modelType, "nemotronh_nano") {
return nil
}
return []Capability{
capabilities := []Capability{
CapabilityImageTextToText,
CapabilityTextToText,
CapabilityAudioToText,
}
if strings.Contains(modelType, "nemotronh_nano_omni") {
capabilities = append(capabilities,
CapabilityAudioTextToText,
CapabilityVideoTextToText,
)
}
return capabilities
}

func omniRule(hf HuggingFaceModel) []Capability {
Expand Down
21 changes: 19 additions & 2 deletions pkg/modelconfig/capability_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,7 +198,23 @@ func TestClassifyCapabilities_NemotronH_Nano(t *testing.T) {
want := []Capability{
CapabilityImageTextToText,
CapabilityTextToText,
CapabilityAudioToText,
CapabilityAudioTextToText,
CapabilityVideoTextToText,
}
if got := classifyCapabilities(m); !equalCaps(got, want) {
t.Errorf("got %v, want %v", got, want)
}
}

func TestClassifyCapabilities_NemotronH_NanoNonOmni(t *testing.T) {
m := &stubModel{
modelType: "NemotronH_Nano_VL_V2",
architecture: "NemotronH_Nano_VL_V2",
hasVision: true,
}
want := []Capability{
CapabilityImageTextToText,
CapabilityTextToText,
}
if got := classifyCapabilities(m); !equalCaps(got, want) {
t.Errorf("got %v, want %v", got, want)
Expand Down Expand Up @@ -413,7 +429,8 @@ func TestClassifyCapabilities_RegressionSet(t *testing.T) {
want: []Capability{
CapabilityImageTextToText,
CapabilityTextToText,
CapabilityAudioToText,
CapabilityAudioTextToText,
CapabilityVideoTextToText,
},
},
{
Expand Down
1 change: 1 addition & 0 deletions pkg/modelparser/capability.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,7 @@ var capabilityToOME = map[modelconfig.Capability]v1beta1.ModelCapability{
modelconfig.CapabilityTextToAudio: v1beta1.ModelCapabilityTextToAudio,
modelconfig.CapabilityImageTextToAudio: v1beta1.ModelCapabilityImageTextToAudio,
modelconfig.CapabilityVideoTextToAudio: v1beta1.ModelCapabilityVideoTextToAudio,
modelconfig.CapabilityVideoTextToText: v1beta1.ModelCapabilityVideoTextToText,
modelconfig.CapabilityAudioToText: v1beta1.ModelCapabilityAudioToText,
modelconfig.CapabilityAudioToAudio: v1beta1.ModelCapabilityAudioToAudio,
modelconfig.CapabilityAudioTextToText: v1beta1.ModelCapabilityAudioTextToText,
Expand Down
1 change: 1 addition & 0 deletions pkg/modelparser/capability_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ func TestCapabilityToOME(t *testing.T) {
{modelconfig.CapabilityTextToAudio, v1beta1.ModelCapabilityTextToAudio},
{modelconfig.CapabilityImageTextToAudio, v1beta1.ModelCapabilityImageTextToAudio},
{modelconfig.CapabilityVideoTextToAudio, v1beta1.ModelCapabilityVideoTextToAudio},
{modelconfig.CapabilityVideoTextToText, v1beta1.ModelCapabilityVideoTextToText},
{modelconfig.CapabilityAudioToText, v1beta1.ModelCapabilityAudioToText},
{modelconfig.CapabilityAudioToAudio, v1beta1.ModelCapabilityAudioToAudio},
{modelconfig.CapabilityAudioTextToText, v1beta1.ModelCapabilityAudioTextToText},
Expand Down
Loading