eval: add VisualEars golden test set results (3 datasets)

eval: add Common Voice 17 results + consistent normalization, lead with CER

fix: --ignore-requires-python for omnilingual-asr 0.2.0 on Python 3.12

fix: --ignore-requires-python so omnilingual-asr 0.2.0 installs on Python 3.12 (Colab)

fix: pin omnilingual-asr==0.2.0 in install

fix: pin omnilingual-asr==0.2.0 (registers 3b_v2 arch)

notebook: default to repo sample clip (one-click runnable)

Add sample Persian clip for one-click demo

Replace with notebook.ipynb

docs: Open-in-Colab badge + point to notebook.ipynb

Add notebook.ipynb (powers HF Open-in-Colab button)

docs: add Rade AI contact info

notebook: download single fp16 file (6GB)

docs: document fp16 single-file + precise two-path usage

Add single-file fp16 weights (model_fp16.pt)

Update WER/CER with Persian-normalized (hazm) numbers: WER 10.8% / CER 3.4%

Update Colab notebook: install recipe + restart + fp16

Update model card: real benchmarks (fp16/fp32), WER/CER, install recipe

Add config.json (download stats + metadata)

Add pp_00/tp_00/sdp_01.pt

Add pp_00/tp_00/sdp_00.pt

Fix snapshot glob in notebook

Add Colab inference notebook

Add model card

initial commit