transformers

26600f6b - [docs] Add a LiteRT page under community integrations (#48540)

[docs] Add a LiteRT page under community integrations (#48540) * [docs] Add a LiteRT page under community integrations Transformers v4 documented LiteRT export through `optimum[exporters-tf]`, which was removed with the TensorFlow support in #40760. This page documents the current route: `litert-torch` converts a model from `torch.export` directly. `litert-torch export_hf` writes a `.litertlm` file for LLMs, and `litert_torch.convert` writes a `.tflite` file for other models. Also adds the toctree entry under "Local deployment" and a redirect from the removed `tflite` page to the new one. * docs(litert): apply review wording suggestions