diff --git a/config/samples/apps_v1alpha1_nemocustomizer.yaml b/config/samples/apps_v1alpha1_nemocustomizer.yaml
index 16beeace3..ff50e19fc 100644
--- a/config/samples/apps_v1alpha1_nemocustomizer.yaml
+++ b/config/samples/apps_v1alpha1_nemocustomizer.yaml
@@ -5,6 +5,7 @@ metadata:
     app.kubernetes.io/name: k8s-nim-operator
     app.kubernetes.io/managed-by: kustomize
   name: nemocustomizer-sample
+  namespace: nemo
 spec:
   wandbSecret:
     name: wandb-secret
@@ -16,7 +17,7 @@ spec:
       user: ncsuser
       secretName: ncs-pg-existing-secret
       passwordKey: password
-    host: <db-service>
+    host: customizer-pg
     port: 5432
     databaseName: ncsdb
   expose:
@@ -30,20 +31,20 @@ spec:
           port: 9009
           protocol: TCP
   image:
-    repository: "nvidian/nemo-llm/customizer"
-    tag: "25.02-rc1"
+    repository: "nvcr.io/nvidian/nemo-llm/customizer"
+    tag: "25.03-rc11"
     pullPolicy: IfNotPresent
     pullSecrets:
       - ngc-secret
   customizerConfig: |
-    namespace: default
-    entity_store_url: http://<entity-store-service>:8000
-    nemo_data_store_url: http://<data-store-service>:3000
-    mlflow_tracking_url: http://<mlflow-tracking-service>:80
+    namespace: nemo
+    entity_store_url: http://nemoentitystore-sample.nemo.svc.cluster.local:8000
+    nemo_data_store_url: http://nemodatastore-sample.nemo.svc.cluster.local:3000
+    mlflow_tracking_url: http://mlflow-tracking.nemo.svc.cluster.local:80
 
     training:
       queue: "default"
-      image: "nvcr.io/nvidian/nemo-llm/customizer:25.02-rc1"
+      image: "nvcr.io/nvidian/nemo-llm/customizer-api:25.03-rc11"
       imagePullSecrets:
         - name: ngc-secret
       pvc:
@@ -95,19 +96,395 @@ spec:
       container_defaults:
           imagePullPolicy: IfNotPresent
 
+    model_download_jobs:
+      image: "nvcr.io/nvidian/nemo-llm/customizer-api:25.03-rc11"
+      imagePullPolicy: "IfNotPresent"
+      imagePullSecrets:
+        - name: ngc-secret
+      ngcAPISecret: ngc-api-secret
+      ngcAPISecretKey: "NGC_API_KEY"
+      securityContext:
+        fsGroup: 1000
+        runAsNonRoot: true
+        runAsUser: 1000
+        runAsGroup: 1000
+      ttlSecondsAfterFinished: 600
+
     models:
-          meta/llama3-8b-instruct:
-            enabled: true
-            max_seq_length: 4096
-            micro_batch_size: 2
-            model_path: llama3-8b-bf16
-            num_parameters: 8000000000
-            precision: bf16
-            training_job_resources:
-            - finetuning_type: lora
-              num_gpus: 2
-              training_type: sft
+      llama-2-7b:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-2-7b:1.0
+        model_path: llama2-7b-bf16
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 4
+            tensor_parallel_size: 4
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 7000000000
+        precision: bf16
+
+      meta/llama3-8b-instruct:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-3-8b-instruct-nemo:1.0
+        model_path: llama3-8b-bf16
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            tensor_parallel_size: 4
+        micro_batch_size: 2
+        max_seq_length: 4096
+        num_parameters: 8000000000
+        precision: bf16
+        # special_tokens:
+        #   begin_of_text: <|begin_of_text|>
+        #   end_of_text: <|end_of_text|>
+        #   role_start: <|start_header_id|>
+        #   role_end: <|end_header_id|>
+        #   label_start: ""
+        #   label_end: ""
+        #   turn_start: ""
+        #   turn_end: <|eot_id|>
+      meta/llama-3.2-1b-embedding:
+        enabled: true
+        model_uri: ngc://nvidian/nemo-llm/llama-3_2-1b-embedding-base:0.0.1
+        model_path: llama32_1b-embedding
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 1
+            num_nodes: 1
+            tensor_parallel_size: 1
+        micro_batch_size: 8
+        max_seq_length: 2048
+        num_parameters: 1000000000
+        precision: bf16-mixed
+
+      mistral-7b:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/mistral-7b:1.0
+        model_path: mistral-7b
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            tensor_parallel_size: 4
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 7000000000
+        precision: bf16
+
+      mistralai/mistral-7b-instruct-v0.3:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/mistral-7b-instruct-v0.3
+        model_path: mistral-7b-v0.3
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            tensor_parallel_size: 4
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 7000000000
+        precision: bf16
+        # tokenizer:
+        #   special_tokens:
+        #     bos_token: "<s>"
+        #     eos_token: "</s>"
+        #     user_start_token: "[INST]"
+        #     user_end_token: "[/INST]"
+        #     tool_call_token: "[TOOL_CALLS]"
+        #     available_tool_start_token: "[AVAILABLE_TOOLS]"
+        #     available_tool_end_token: "[/AVAILABLE_TOOLS]"
+        #     tool_result_start_token: "[TOOL_RESULTS]"
+        #     tool_result_end_token: "[/TOOL_RESULTS]"
+        #   sentencepiece_legacy: true
+        # special_tokens:
+        #   begin_of_text: "<s>"
+        #   end_of_text: "</s>"
+        #   user_turn_start: "[INST]"
+        #   user_turn_end: "[/INST]"
+        #   system_turn_start: ""
+        #   system_turn_end: "\n\n"
+        #   turn_start: ""
+        #   turn_end: "</s>"
+        #   include_roles: false
+        #   supports_system_role: false
+
+      llama-2-13b:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-2-13b:1.0
+        model_path: llama2-13b-bf16
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 4
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 4
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            num_nodes: 2
+            tensor_parallel_size: 8
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 7000000000
+        precision: bf16
+
+      gpt-43b-002:
+        enabled: false
+        model_uri: ngc://gku9emylpays/nemomodel/gpt_43b_002:1.0
+        model_path: gpt-43b-002
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 43000000000
+        precision: bf16
+        sft_prompt_template: 'User: {prompt}\n\nAssistant:{completion}'
+
+      llama-2-70b:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-2-70b:1.0
+        model_path: llama2-70b-bf16
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+        max_seq_length: 4096
+        num_parameters: 70000000000
+        micro_batch_size: 1
+        precision: bf16
+
+      meta/llama3-70b-instruct:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-3-70b-instruct-nemo:1.0
+        model_path: llama3-70b-bf16
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+        max_seq_length: 4096
+        num_parameters: 70000000000
+        micro_batch_size: 1
+        precision: bf16
+        # special_tokens:
+        #   begin_of_text: <|begin_of_text|>
+        #   end_of_text: <|end_of_text|>
+        #   role_start: <|start_header_id|>
+        #   role_end: <|end_header_id|>
+        #   label_start: ""
+        #   label_end: ""
+        #   turn_start: ""
+        #   turn_end: <|eot_id|>
+
+      meta/llama-3_1-8b-instruct:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-3_1-8b-instruct-nemo:1.0
+        model_path: llama-3_1-8b-instruct
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            num_nodes: 1
+            tensor_parallel_size: 4
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 8000000000
+        precision: bf16
+        tokenizer_path: /home/customizer/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-instruct/snapshots/c4a54320a52ed5f88b7a2f84496903ea4ff07b45
+        # special_tokens:
+        #   begin_of_text: <|begin_of_text|>
+        #   end_of_text: <|finetune_right_pad_id|>
+        #   role_start: <|start_header_id|>
+        #   role_end: <|end_header_id|>
+        #   label_start: ""
+        #   label_end: ""
+        #   turn_start: ""
+        #   turn_end: <|eot_id|>
+      # for nim release 1.2 and later
+
+      meta/llama-3.1-8b-instruct:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-3_1-8b-instruct-nemo:1.0
+        model_path: llama-3_1-8b-instruct
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            num_nodes: 1
+            tensor_parallel_size: 4
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 8000000000
+        precision: bf16
+        tokenizer_path: /home/customizer/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3-8B-instruct/snapshots/c4a54320a52ed5f88b7a2f84496903ea4ff07b45
+        # special_tokens:
+        #   begin_of_text: <|begin_of_text|>
+        #   end_of_text: <|finetune_right_pad_id|>
+        #   role_start: <|start_header_id|>
+        #   role_end: <|end_header_id|>
+        #   label_start: ""
+        #   label_end: ""
+        #   turn_start: ""
+        #   turn_end: <|eot_id|>
+
+      meta/llama-3.1-70b-instruct:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/llama-3_1-70b-instruct-nemo:1.0
+        model_path: llama-3_1-70b-instruct
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+        micro_batch_size: 1
+        max_seq_length: 4096
+        num_parameters: 70000000000
+        precision: bf16
+        # special_tokens:
+        #   begin_of_text: <|begin_of_text|>
+        #   end_of_text: <|finetune_right_pad_id|>
+        #   role_start: <|start_header_id|>
+        #   role_end: <|end_header_id|>
+        #   label_start: ""
+        #   label_end: ""
+        #   turn_start: ""
+        #   turn_end: <|eot_id|>
+
+      gpt8b-4k:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/nemotron-3-8b-base-4k:1.0
+        model_path: nemotron_3_8b_base_4k
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 4
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+          - training_type: sft
+            finetuning_type: all_weights
+            num_gpus: 8
+            tensor_parallel_size: 4
+        max_seq_length: 4096
+        num_parameters: 8000000000
+        precision: bf16
+        micro_batch_size: 1
+        multinode_dataset_thresholds:
+          - num_nodes: 2
+            threshold: 80000
+
+      gemma-7b:
+        enabled: false
+        model_uri: ngc://nvidia/nemo/gemma_7b_base:1.1
+        model_path: "gemma-7b"
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 2
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 2
+        micro_batch_size: 4
+        max_seq_length: 8192
+        num_parameters: 7000000000
+        precision: bf16
+
+      codellama-70b:
+        enabled: false
+        model_path: "codellama_70b_bf16"
+        training_options:
+          - training_type: sft
+            finetuning_type: p_tuning
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 4
+            num_nodes: 2
+            tensor_parallel_size: 8
+        micro_batch_size: 1
+        max_seq_length: 16384
+        num_parameters: 70000000000
+        precision: bf16
+
+      meta/llama3-405b-instruct:
+        enabled: false
+        max_seq_length: 2048
+        num_parameters: 405000000000
+        precision: bf16
+        micro_batch_size: 1
+        model_path: llama-3_1-405b-instruct
+        training_options:
+          - training_type: sft
+            finetuning_type: lora
+            num_gpus: 8
+            num_nodes: 3
+            tensor_parallel_size: 4
+        # pipeline_parallel_size: 6
+        # use_sequence_parallel: false
+        # special_tokens:
+        #   begin_of_text: <|begin_of_text|>
+        #   end_of_text: <|end_of_text|>
+        #   role_start: <|start_header_id|>
+        #   role_end: <|end_header_id|>
+        #   label_start: ""
+        #   label_end: ""
+        #   turn_start: ""
+        #   turn_end: <|eot_id|>
 
     nemo_data_store_tools:
-      image: nvcr.io/nvidian/nemo-llm/nds-v2-huggingface-cli:278bb5cd0fde8ad46842daf6d2471134624ae891
+      image: nvcr.io/nvidian/nemo-llm/nds-v2-huggingface-cli:25.03-rc11
       imagePullSecret: ngc-secret