Skip to content

Replace the deprecated backend with model_type tags for non LLM nims #513

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

Merged
visheshtanksale merged 3 commits into from
May 30, 2025
Merged

Replace the deprecated backend with model_type tags for non LLM nims #513

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
91b6793
Replace the deprecated backend with model_type tags for non LLM nims
visheshtanksale May 30, 2025
ebe068b
Replace the deprecated backend with model_type tags for non LLM nims
visheshtanksale May 30, 2025
0104166
Addressing review comments
visheshtanksale May 30, 2025
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
2 changes: 1 addition & 1 deletion api/apps/v1alpha1/nimcache_types.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -127,7 +127,7 @@ type ModelSpec struct {
Profiles []string `json:"profiles,omitempty"`
// Precision is the precision for model quantization
Precision string `json:"precision,omitempty"`
// Engine is the backend engine (tensort_llm, vllm)
// Engine is the backend engine (tensorrt_llm, vllm)
Engine string `json:"engine,omitempty"`
// TensorParallelism is the minimum GPUs required for the model computations
TensorParallelism string `json:"tensorParallelism,omitempty"`
Expand Down
2 changes: 1 addition & 1 deletion bundle/manifests/apps.nvidia.com_nimcaches.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -344,7 +344,7 @@ spec:
GPUs
type: boolean
engine:
description: Engine is the backend engine (tensort_llm,
description: Engine is the backend engine (tensorrt_llm,
vllm)
type: string
gpus:
Expand Down
2 changes: 1 addition & 1 deletion config/crd/bases/apps.nvidia.com_nimcaches.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -344,7 +344,7 @@ spec:
GPUs
type: boolean
engine:
description: Engine is the backend engine (tensort_llm,
description: Engine is the backend engine (tensorrt_llm,
vllm)
type: string
gpus:
Expand Down
2 changes: 1 addition & 1 deletion deployments/helm/k8s-nim-operator/crds/apps.nvidia.com_nimcaches.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -344,7 +344,7 @@ spec:
GPUs
type: boolean
engine:
description: Engine is the backend engine (tensort_llm,
description: Engine is the backend engine (tensorrt_llm,
vllm)
type: string
gpus:
Expand Down
12 changes: 12 additions & 0 deletions internal/nimparser/v2/nimparser.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,10 +80,22 @@ func (manifest NIMManifest) MatchProfiles(modelSpec appsv1alpha1.ModelSpec, disc

// Determine backend type
backend := profile.Tags["llm_engine"]

// Use "model_type" if "llm_engine" is empty for non LLM models
if backend == "" {
backend = profile.Tags["model_type"]
}

// Fallback to deprecated "backend" tag for non LLM Models
if backend == "" {
backend = profile.Tags["backend"]
}

// modespec.Engine value can be "tensorrt_llm". If backend tag is "triton", convert it to "tensorrt".
if backend == "triton" {
backend = "tensorrt"
}

if modelSpec.Engine != "" && !strings.Contains(backend, strings.TrimSuffix(modelSpec.Engine, "_llm")) {
continue
}
Expand Down
Loading