* fix: handle more parameters

* fix: generate version.txt if not exist

* fix: only add path if exists

* chore: function calling cleanup

* chore: cleanup

---------

Co-authored-by: sangjanai <[email protected]>

* fix: remove jinja parameter

* chore: disable linux arm CI

---------

Co-authored-by: sangjanai <[email protected]>

* fix: check model is loaded before starting

* chore: e2e test

---------

Co-authored-by: sangjanai <[email protected]>

* refactor: remove --pooling flag from model loading

The --pooling flag was removed as the mean pooling functionality not needed in chat models. This fixes the regression

* feat(local-engine): add ctx_len parameter support

Adds support for the ctx_len parameter by appending --ctx-size with its value. Removed outdated parameter mappings from the kParamsMap to reflect current implementation details and ensure consistency.

* feat: add conditional model parameters based on path

When the model path contains both "jan" and "nano" (case-insensitive), automatically add
speculative decoding parameters to adjust generation behavior. This improves
flexibility by enabling environment-specific configurations without manual
parameter tuning. Also includes necessary headers for string manipulation and
fixes whitespace in ctx_len handling.

* chore: remove redundant comment

The comment was redundant as the code's purpose is clear without it, improving readability.

…all local models (#2207)

Co-authored-by: Ethan Garber <[email protected]>
Co-authored-by: Akarshan Biswas <[email protected]>