Merge remote-tracking branch 'github/main' into fsiino/clarify-config-yaml-info

Signed-off-by: Frankie Siino <fsiino@nvidia.com>

Author: fsiino-nvidia

docs/tutorials/02-setup.md

@@ -66,9 +66,10 @@ INFO:     Uvicorn running on http://127.0.0.1:62920 (Press CTRL+C to quit)
 
 This means **4 servers are now running**:
 1. **Head server** (coordinates everything)
-2. **Simple weather resource** (provides weather tool)
-3. **OpenAI model server** (connects to GPT-4)
-4. **Simple agent** (orchestrates model + resources)
+2. 3 Gym servers. These 3 servers and their high level config should be printed to terminal!
+   1. **Simple weather resource** (provides weather tool)
+   2. **OpenAI model server** (connects to GPT-4)
+   3. **Simple agent** (orchestrates model + resources)
 
 ## Step 4: Test the Setup
 

responses_api_models/azure_openai_model/README.md

@@ -8,18 +8,18 @@ Set up your `env.yaml` file:
 policy_base_url: https://my.end.point.com/v1/azure
 policy_api_key: <API_KEY>
 policy_model_name: gpt-5-nano
-policy_api_version: 2024-10-21
 ```
 
 ## Usage
 
 ### Running the server
+Set the API version. It usually looks something like "2024-10-21".
 ```bash
 config_paths="responses_api_models/azure_openai_model/configs/azure_openai_model.yaml, \
 resources_servers/equivalence_llm_judge/configs/equivalence_llm_judge.yaml"
 
 ng_run "+config_paths=[${config_paths}]" \
-+policy_model.responses_api_models.azure_openai_model.default_query.api-version=<api_version>
+    +policy_model.responses_api_models.azure_openai_model.default_query.api-version=<api_version>
 ```
 
 ### Collecting Rollouts