fix: unpinned dependencies cause slow resolution and silent version downgrade

[Jaimin-Kundal]

Problem

While setting up the project locally, running pip install -r requirements.txt
triggered a dependency backtracking warning:

INFO: pip is still looking at multiple versions of gradio to determine which version is compatible with other requirements. This could take a while.
Downloading gradio-6.6.0-py3-none-any.whl.metadata (16 kB)
Collecting gradio-client==2.1.0 (from gradio->-r requirements.txt (line 8))
Downloading gradio_client-2.1.0-py3-none-any.whl.metadata (7.1 kB)
Collecting gradio (from -r requirements.txt (line 8))
Downloading gradio-6.5.1-py3-none-any.whl.metadata (16 kB)
Collecting gradio-client==2.0.3 (from gradio->-r requirements.txt (line 8))
Downloading gradio_client-2.0.3-py3-none-any.whl.metadata (7.1 kB)
Collecting gradio (from -r requirements.txt (line 8))
Downloading gradio-6.5.0-py3-none-any.whl.metadata (16 kB)
Downloading gradio-6.4.0-py3-none-any.whl.metadata (16 kB)
Downloading gradio-6.3.0-py3-none-any.whl.metadata (16 kB)
INFO: This is taking longer than usual. You might need to provide the dependency resolver with stricter constraints to reduce runtime. See https://pip.pypa.io/warnings/backtracking for guidance. If you want to abort this run, press Ctrl + C.

pip backtracked through 90+ versions of gradio before settling on
gradio==4.44.1, which is 2 major versions behind the current 6.x release.
This happened silently with no error a new contributor would have no idea
they're running outdated code.

Root cause: huggingface_hub==0.22.2 is pinned to April 2024. Modern
gradio requires a newer version, so pip is forced to keep downgrading gradio
until it finds a compatible version. All other packages are completely unpinned,
making installs unpredictable across different machines and times.

Result on my machine:

Package	Actually Installed	Latest Available
gradio	4.44.1	6.14.0
huggingface-hub	0.22.2 (old pin)	0.30+
openai	2.36.0 (unpinned)	varies

Proposed Fix

Pin all direct dependencies to their verified compatible versions so every
contributor gets a fast, reproducible install with no backtracking.

I've verified a working pinned set locally pip check returns
No broken requirements found with zero conflicts.

Happy to open a PR with the updated requirements.txt if this looks good.

Environment

• Python 3.13
• pip latest
• OS: Linux (Kali)

[DavidH-1]

we do not use issues we use Jira please raise a valid Jira ticket

[Jaimin-Kundal]

@DavidH-1 Apologies for the process miss! I've raised the corresponding Jira ticket here: AI-217 also signed the CLA and accepted the contributor invitation. The fix is in PR #116 when you get a chance to review. Also Added Jira Ticket reference there .