Fix column header validation issue

Author: dksimpson

articles/ai-foundry/concepts/models-featured.md

@@ -27,9 +27,9 @@ To perform inferencing with the models, some models such as [Nixtla's TimeGEN-1]
 
 The Jamba family models are AI21's production-grade Mamba-based large language model (LLM) which uses AI21's hybrid Mamba-Transformer architecture. It's an instruction-tuned version of AI21's hybrid structured state space model (SSM) transformer Jamba model. The Jamba family models are built for reliable commercial use with respect to quality and performance.
 
-| Model  | Type | Capabilities | 
-| ------ | ---- | --- | 
-| [AI21-Jamba-1.5-Mini](https://ai.azure.com/explore/models/AI21-Jamba-1.5-Mini/version/1/registry/azureml-ai21) | [chat-completion](../model-inference/how-to/use-chat-completions.md?context=/azure/ai-foundry/context/context) | - **Input:** text (262,144 tokens) <br /> - **Output:**  text (4,096 tokens) <br /> - **Tool calling:** Yes <br /> - **Response formats:** Text, JSON, structured outputs | | 
+| Model  | Type | Capabilities |
+| ------ | ---- | --- |
+| [AI21-Jamba-1.5-Mini](https://ai.azure.com/explore/models/AI21-Jamba-1.5-Mini/version/1/registry/azureml-ai21) | [chat-completion](../model-inference/how-to/use-chat-completions.md?context=/azure/ai-foundry/context/context) | - **Input:** text (262,144 tokens) <br /> - **Output:**  text (4,096 tokens) <br /> - **Tool calling:** Yes <br /> - **Response formats:** Text, JSON, structured outputs |
 | [AI21-Jamba-1.5-Large](https://ai.azure.com/explore/models/AI21-Jamba-1.5-Large/version/1/registry/azureml-ai21) | [chat-completion](../model-inference/how-to/use-chat-completions.md?context=/azure/ai-foundry/context/context) | - **Input:** text (262,144 tokens) <br /> - **Output:**  text (4,096 tokens) <br /> - **Tool calling:** Yes <br /> - **Response formats:** Text, JSON, structured outputs |