fix: provider CI health checks

.agents/recipes/docs-and-references/recipe.md

@@ -4,7 +4,7 @@ description: Audit documentation freshness - docstrings vs signatures, broken li
 trigger: schedule
 tool: claude-code
 timeout_minutes: 20
-max_turns: 30
+max_turns: 50
 permissions:
   contents: write
 ---

.github/workflows/agentic-ci-issue-triage.yml

@@ -55,18 +55,25 @@ jobs:
           echo "Claude CLI version: $(claude --version 2>&1 || true)"
 
           if [ -n "$ANTHROPIC_BASE_URL" ] && [ -n "$ANTHROPIC_API_KEY" ]; then
-            HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" \
-              --max-time 30 \
-              -X POST "${ANTHROPIC_BASE_URL}/v1/messages" \
-              -H "Content-Type: application/json" \
-              -H "x-api-key: ${ANTHROPIC_API_KEY}" \
-              -H "anthropic-version: 2023-06-01" \
-              -d "{\"model\":\"${AGENTIC_CI_MODEL}\",\"max_tokens\":5,\"messages\":[{\"role\":\"user\",\"content\":\"hi\"}]}")
-            if [ "$HTTP_CODE" -lt 200 ] || [ "$HTTP_CODE" -ge 300 ]; then
-              echo "::error::API pre-flight failed with HTTP ${HTTP_CODE}"
-              exit 1
-            fi
-            echo "API pre-flight passed (HTTP ${HTTP_CODE})"
+            for ATTEMPT in 1 2 3; do
+              HTTP_CODE=$(curl -s -o /dev/null -w "%{http_code}" \
+                --max-time 30 \
+                -X POST "${ANTHROPIC_BASE_URL}/v1/messages" \
+                -H "Content-Type: application/json" \
+                -H "x-api-key: ${ANTHROPIC_API_KEY}" \
+                -H "anthropic-version: 2023-06-01" \
+                -d "{\"model\":\"${AGENTIC_CI_MODEL}\",\"max_tokens\":5,\"messages\":[{\"role\":\"user\",\"content\":\"hi\"}]}")
+              if [ "$HTTP_CODE" -ge 200 ] && [ "$HTTP_CODE" -lt 300 ]; then
+                echo "API pre-flight passed (HTTP ${HTTP_CODE})"
+                break
+              fi
+              if [ "$ATTEMPT" = "3" ]; then
+                echo "::error::API pre-flight failed with HTTP ${HTTP_CODE}"
+                exit 1
+              fi
+              echo "API pre-flight failed with HTTP ${HTTP_CODE}; retrying (${ATTEMPT}/3)"
+              sleep $((ATTEMPT * 10))
+            done
           fi
 
       - name: Run issue triage recipe

.github/workflows/build-notebooks.yml

@@ -87,7 +87,11 @@ jobs:
             echo "No previous successful run found, proceeding without cache"
           fi
       - name: Convert and execute notebooks
-        run: make convert-execute-notebooks ${{ inputs.use_cache && 'USE_CACHE=1' || '' }}
+        run: |
+          if [ "$GITHUB_EVENT_NAME" = "schedule" ]; then
+            export DATA_DESIGNER_FLUX_2_PRO_CREATE_NUM_RECORDS=2
+          fi
+          make convert-execute-notebooks ${{ inputs.use_cache && 'USE_CACHE=1' || '' }}
       - name: Upload notebooks as artifacts
         uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
         with:

docs/concepts/models/default-model-settings.md

@@ -46,7 +46,7 @@ The following model configurations are automatically available when `NVIDIA_API_
 | `nvidia-text` | `nvidia/nemotron-3-nano-30b-a3b` | General text generation | `temperature=1.0, top_p=1.0` |
 | `nvidia-reasoning` | `nvidia/nemotron-3-super-120b-a12b` | Reasoning and analysis tasks | `temperature=1.0, top_p=0.95, extra_body={"reasoning_effort": "medium"}` |
 | `nvidia-vision` | `nvidia/nemotron-3-nano-omni-30b-a3b-reasoning` | Vision and image understanding | `temperature=0.60, top_p=0.95` |
-| `nvidia-embedding` | `nvidia/llama-3.2-nv-embedqa-1b-v2` | Text embeddings | `encoding_format="float", extra_body={"input_type": "query"}` |
+| `nvidia-embedding` | `nvidia/llama-nemotron-embed-1b-v2` | Text embeddings | `encoding_format="float", extra_body={"input_type": "query"}` |
 
 
 ### OpenAI Models

docs/notebook_source/4-providing-images-as-context.py

@@ -76,6 +76,10 @@
 
 # %%
 config_builder = dd.DataDesignerConfigBuilder()
+for model_config in config_builder.model_configs:
+    if model_config.alias == "nvidia-vision":
+        model_config.skip_health_check = True
+        break
 
 # %% [markdown]
 # ### 🌱 Seed Dataset Creation

docs/notebook_source/5-generating-images.py

@@ -38,6 +38,8 @@
 #
 
 # %%
+import os
+
 from IPython.display import Image as IPImage
 from IPython.display import display
 
@@ -272,7 +274,10 @@
 #
 
 # %%
-results = data_designer.create(config_builder, num_records=2, dataset_name="tutorial-5-images")
+create_num_records = 2
+if MODEL_ID == "black-forest-labs/flux.2-pro":
+    create_num_records = int(os.environ.get("DATA_DESIGNER_FLUX_2_PRO_CREATE_NUM_RECORDS") or create_num_records)
+results = data_designer.create(config_builder, num_records=create_num_records, dataset_name="tutorial-5-images")
 
 # %%
 dataset = results.load_dataset()

docs/notebook_source/6-editing-images-with-image-context.py

@@ -39,6 +39,7 @@
 
 # %%
 import base64
+import os
 from pathlib import Path
 
 from IPython.display import Image as IPImage
@@ -244,7 +245,10 @@ def display_before_after(row, index: int, base_path: Path | None = None) -> None
 #
 
 # %%
-results = data_designer.create(config_builder, num_records=5, dataset_name="tutorial-6-edited-images")
+create_num_records = 5
+if MODEL_ID == "black-forest-labs/flux.2-pro":
+    create_num_records = int(os.environ.get("DATA_DESIGNER_FLUX_2_PRO_CREATE_NUM_RECORDS") or create_num_records)
+results = data_designer.create(config_builder, num_records=create_num_records, dataset_name="tutorial-6-edited-images")
 
 # %%
 dataset = results.load_dataset()

fern/versions/latest/pages/concepts/models/default-model-settings.mdx

@@ -49,7 +49,7 @@ The following model configurations are automatically available when `NVIDIA_API_
 | `nvidia-text` | `nvidia/nemotron-3-nano-30b-a3b` | General text generation | `temperature=1.0, top_p=1.0` |
 | `nvidia-reasoning` | `nvidia/nemotron-3-super-120b-a12b` | Reasoning and analysis tasks | `temperature=1.0, top_p=0.95, extra_body={"reasoning_effort": "medium"}` |
 | `nvidia-vision` | `nvidia/nemotron-3-nano-omni-30b-a3b-reasoning` | Omni multimodal understanding for image, audio, and video inputs | `temperature=0.60, top_p=0.95` |
-| `nvidia-embedding` | `nvidia/llama-3.2-nv-embedqa-1b-v2` | Text embeddings | `encoding_format="float", extra_body={"input_type": "query"}` |
+| `nvidia-embedding` | `nvidia/llama-nemotron-embed-1b-v2` | Text embeddings | `encoding_format="float", extra_body={"input_type": "query"}` |
 
 
 ### OpenAI Models

packages/data-designer-config/src/data_designer/config/utils/constants.py

@@ -360,7 +360,7 @@ class NordColor(Enum):
             "inference_parameters": NEMOTRON_3_NANO_OMNI_30B_A3B_REASONING_INFERENCE_PARAMS,
         },
         "embedding": {
-            "model": "nvidia/llama-3.2-nv-embedqa-1b-v2",
+            "model": "nvidia/llama-nemotron-embed-1b-v2",
             "inference_parameters": DEFAULT_EMBEDDING_INFERENCE_PARAMS | {"extra_body": {"input_type": "query"}},
         },
     },

packages/data-designer-config/tests/config/test_default_model_settings.py

@@ -65,7 +65,7 @@ def test_get_builtin_model_configs():
     assert builtin_model_configs[2].model == "nvidia/nemotron-3-nano-omni-30b-a3b-reasoning"
     assert builtin_model_configs[2].provider == "nvidia"
     assert builtin_model_configs[3].alias == "nvidia-embedding"
-    assert builtin_model_configs[3].model == "nvidia/llama-3.2-nv-embedqa-1b-v2"
+    assert builtin_model_configs[3].model == "nvidia/llama-nemotron-embed-1b-v2"
     assert builtin_model_configs[3].provider == "nvidia"
     assert builtin_model_configs[4].alias == "openai-text"
     assert builtin_model_configs[4].model == "gpt-4.1"

scripts/health_checks.py

@@ -31,11 +31,21 @@
 )
 from data_designer.interface import DataDesigner
 
+MAX_ATTEMPTS = 3
 PROVIDER_API_KEY_ENV_VARS = {
     NVIDIA_PROVIDER_NAME: NVIDIA_API_KEY_ENV_VAR_NAME,
     OPENAI_PROVIDER_NAME: OPENAI_API_KEY_ENV_VAR_NAME,
     OPENROUTER_PROVIDER_NAME: OPENROUTER_API_KEY_ENV_VAR_NAME,
 }
+RETRYABLE_ERROR_NAMES = {
+    "ModelAPIConnectionError",
+    "ModelAPIError",
+    "ModelInternalServerError",
+    "ModelRequestAdmissionTimeoutError",
+    "ModelRateLimitError",
+    "ModelTimeoutError",
+    "TimeoutError",
+}
 
 
 def _get_provider(provider_name: str) -> dd.ModelProvider:
@@ -88,8 +98,15 @@ def _check_model(provider_name: str, model_type: str) -> None:
     model_config = _get_model_config(provider_name, model_type)
     config_builder = _build_check_config(model_config, model_type)
 
-    with TemporaryDirectory(prefix="data-designer-health-check-") as temp_dir:
-        DataDesigner(artifact_path=Path(temp_dir), model_providers=[provider]).check_models(config_builder)
+    for attempt in range(1, MAX_ATTEMPTS + 1):
+        try:
+            with TemporaryDirectory(prefix="data-designer-health-check-") as temp_dir:
+                DataDesigner(artifact_path=Path(temp_dir), model_providers=[provider]).check_models(config_builder)
+            return
+        except Exception as exc:
+            if type(exc).__name__ not in RETRYABLE_ERROR_NAMES or attempt == MAX_ATTEMPTS:
+                raise
+            print(f"RETRY {provider_name}/{model_type} (attempt {attempt + 1}/{MAX_ATTEMPTS})")
 
 
 def main() -> int: