Add Model column to collect-results summary table

[Copilot]

When collecting results across all experiments (not filtered by experiment name), the summary table lacked model identification, making it impossible to distinguish which model produced each result.

Changes

• Added Model column as the first column in utils/summarize.py summary table
• Updated sort order to group by model first: model → hardware → framework → precision → TP → EP → concurrency
• Handles missing data gracefully with 'unknown' default for backward compatibility

Example Output

| Model | Hardware | Framework | Precision | TP | EP | DP Attention | Conc | TTFT (ms) | ... |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: | ... |
| deepseek-ai/DeepSeek-R1-Distill-Llama-70B | H200 | SGLANG | FP8 | 8 | 8 | false | 4 | 13.9000 | ... |
| meta-llama/Llama-3.3-70b-Instruct | B200 | TRT | FP4 | 8 | 1 | false | 4 | 11.5000 | ... |
| meta-llama/Llama-3.3-70b-Instruct | H200 | VLLM | FP8 | 8 | 1 | false | 4 | 30.5000 | ... |

Model data already captured by process_result.py from benchmark results.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Add model name in some capacity to summarize results table</issue_title>
<issue_description>In InferenceMAX/InferenceMAX#167, we realized that the changes made to collect results warrant a need to display model in the table in the case that we are not collecting by experiment name but rather all results at once.

</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes InferenceMAX/InferenceMAX#193

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[Copilot]

Pull Request Overview

This PR adds a Model column to the results summary table in utils/summarize.py to distinguish which model produced each result when collecting results across all experiments. Previously, when not filtering by experiment name, users couldn't identify the model associated with each benchmark result.

Key changes:

• Added Model as the first column in the summary table
• Updated sort order to group results by model first, then hardware, framework, precision, TP, EP, and concurrency
• Handles missing model data with 'unknown' default for backward compatibility

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[functionstackx]

lgtm