Options to change visualqa model and search tool in open deep research

[suryabdev]

Fix for #565

Added two options to examples/open_deep_research/run.py (--visual-qa-model-id and --search-engine) to allow configuration of the VisualQATool's model and the search tool used. User wanted to used the DuckDuckGoSearchTool since it is free

>  python run.py --model-id "o1" --visual-qa-model-id "gpt-4o-mini" --search-engine "ddgs" "Who is the CEO of huggingface"
Note: Environment variable`HF_TOKEN` is set and is the current active token independently from the token you've just configured.
Caution: you set an authorization for all imports, meaning your agent can decide to import any package it deems necessary. This might raise issues if the
package is not installed in your environment.
╭──────────────────────────────────────────────────────────────────────── New run ─────────────────────────────────────────────────────────────────────────╮
│                                                                                                                                                          │
│ Who is the CEO of huggingface                                                                                                                            │
│                                                                                                                                                          │
╰─ LiteLLMModel - o1 ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────╯
─────────────────────────────────────────────────────────────────────── Initial plan ───────────────────────────────────────────────────────────────────────
Here are the facts I know and the plan of action that I will follow to solve the task:

....

science,” and Delangue is frequently involved in supporting partnerships with major tech firms and open-source organizations.'}]
---
</summary_of_work>

Out: None
[Step 1: Duration 44.11 seconds| Input tokens: 2,662 | Output tokens: 604]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Step 2 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Output message of the LLM: ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Thought: Based on the search_agent's response, the CEO of Hugging Face is Clément Delangue. I will now provide the final answer.

<code>
final_answer("Clément Delangue is the CEO of Hugging Face.")

 ─ Executing parsed code: ─────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
  final_answer("Clément Delangue is the CEO of Hugging Face.")
 ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Final answer: Clément Delangue is the CEO of Hugging Face.
[Step 2: Duration 7.67 seconds| Input tokens: 10,883 | Output tokens: 1,246]
Got this answer: Clément Delangue is the CEO of Hugging Face.

Did not touch examples/open_deep_research/run_gaia.py since that is used for benchmarking