added example for making an inference call with minimal client

Author: tamirse

CHANGELOG.rst

@@ -1,6 +1,9 @@
 Changelog
 =========
 
+* Added example for calling the inference endpoint with a minimal client
+* Added missing doc generation for inference examples
+
 v1.10.0 (2025-04-17)
 -------------------
 

docs/source/examples/containers/index.rst

@@ -7,10 +7,13 @@ This section contains examples demonstrating how to work with containers in Data
    :maxdepth: 1
    :caption: Contents:
 
-   compute_resources
    deployments
+   compute_resources
    environment_variables
    registry_credentials
    secrets
    sglang
-   scaling 
+   scaling 
+   inference_async
+   inference_sync
+   inference_minimal

docs/source/examples/containers/inference_async.rst

@@ -0,0 +1,8 @@
+Calling the inference endpoint in async mode
+============================================
+
+This example demonstrates how to call the inference endpoint in async mode.
+
+.. literalinclude:: ../../../../examples/containers/calling_the_inference_endpoint_in_async_mode.py
+   :language: python
+   :caption: Calling the inference endpoint in async mode

docs/source/examples/containers/inference_minimal.rst

@@ -0,0 +1,8 @@
+Calling the inference endpoint using a minimal client
+=====================================================
+
+This example demonstrates how to call the inference endpoint using a minimal client that only uses only an inference key (no client credentials).
+
+.. literalinclude:: ../../../../examples/containers/calling_the_endpoint_with_inference_key.py
+   :language: python
+   :caption: Calling the inference endpoint using a minimal client

docs/source/examples/containers/sglang.rst

@@ -5,4 +5,4 @@ This example demonstrates how to deploy and manage SGLang applications in DataCr
 
 .. literalinclude:: ../../../../examples/containers/sglang_deployment_example.py
    :language: python
-   :caption: SGLang Deployment 
+   :caption: SGLang Deployment Example

examples/containers/calling_the_endpoint_with_inference_key.py

@@ -0,0 +1,27 @@
+import os
+from datacrunch.InferenceClient import InferenceClient
+
+# Get inference key and endpoint base url from environment variables
+DATACRUNCH_INFERENCE_KEY = os.environ.get('DATACRUNCH_INFERENCE_KEY')
+DATACRUNCH_ENDPOINT_BASE_URL = os.environ.get('DATACRUNCH_ENDPOINT_BASE_URL')
+
+# Create an inference client that uses only the inference key, without client credentials
+inference_client = InferenceClient(
+    inference_key=DATACRUNCH_INFERENCE_KEY,
+    endpoint_base_url=DATACRUNCH_ENDPOINT_BASE_URL
+)
+
+# Make a synchronous request to the endpoint.
+# This example demonstrates calling a SGLang deployment which serves LLMs using an OpenAI-compatible API format
+data = {
+    "model": "deepseek-ai/deepseek-llm-7b-chat",
+    "prompt": "Is consciousness fundamentally computational, or is there something more to subjective experience that cannot be reduced to information processing?",
+    "max_tokens": 128,
+    "temperature": 0.7,
+    "top_p": 0.9
+}
+
+response = inference_client.run_sync(data=data, path='v1/completions')
+
+# Print the response
+print(response.output())