| --- |
| license: apache-2.0 |
| language: |
| - fa |
| library_name: nemo |
| pipeline_tag: automatic-speech-recognition |
| tags: |
| - nemo |
| - automatic-speech-recognition |
| - speech |
| - persian |
| - farsi |
| - fastconformer |
| - ctc |
| - on-device |
| - shenava |
| - shenava-1 |
| - visualears |
| - liteasr |
| - compression |
| - low-rank |
| - dhh |
| base_model: Reza2kn/Shenava-Koochik-v1.0 |
| base_model_relation: quantized |
| datasets: |
| - Reza2kn/visualears-golden-6669 |
| - Reza2kn/fleurs-fa-benchmark |
| --- |
| |
| # 🪶🎙️ Shenava Koochik Lite v1.0 |
|
|
| A LITEASR-compressed encoder for [Shenava Koochik v1.0](https://huggingface.co/Reza2kn/Shenava-Koochik-v1.0). Post-training low-rank factorization reduces the encoder from 108.9M to 85.4M parameters (21.6%) without retraining. |
|
|
| This repository is **not a standalone ASR checkpoint**. It contains a replacement encoder state dict and must be loaded on top of the base `.nemo` model; the decoder, CTC head, and tokenizer still come from Koochik. |
|
|
| ## ✨ At a glance | معرفی سریع |
|
|
| | | English | فارسی | |
| |---|---|---| |
| | 🪶 Role | Compressed Koochik encoder | encoder فشردهشدهٔ کوچیک | |
| | 📉 Reduction | 108.9M → 85.4M encoder parameters | کاهش ۲۱٫۶ درصدی پارامترهای encoder | |
| | 🧪 Method | Post-training LITEASR low-rank factorization | فشردهسازی low-rank بدون آموزش مجدد | |
| | 🧩 Requirement | Base Koochik `.nemo` is required | فایل NeMo مدل اصلی الزامی است | |
| | ⚠️ Scope | Not a standalone checkpoint | checkpoint مستقل نیست | |
|
|
| - Canonical repository: [`Reza2kn/Shenava-Koochik-Lite-v1.0`](https://huggingface.co/Reza2kn/Shenava-Koochik-Lite-v1.0) |
| - PersianML mirror: [`PersianML/Shenava-Koochik-Lite-v1.0`](https://huggingface.co/PersianML/Shenava-Koochik-Lite-v1.0) |
|
|
| ## 📦 Files |
|
|
| - `koochik_lite099_enc.pt`: compressed FP32 encoder state dict. |
| - `koochik_lite099_kmap.json`: retained rank for each factorized layer. |
| - `load_koochik_lite.py`: reconstructs the low-rank modules and loads the state dict into the base model. |
|
|
| ## 🚀 Load |
|
|
| ```python |
| from huggingface_hub import hf_hub_download, snapshot_download |
| from load_koochik_lite import load_koochik_lite |
| |
| base = hf_hub_download( |
| "Reza2kn/Shenava-Koochik-v1.0", |
| "shenava-koochik-v1.0.nemo", |
| ) |
| repo = snapshot_download("Reza2kn/Shenava-Koochik-Lite-v1.0") |
| model = load_koochik_lite( |
| base, |
| f"{repo}/koochik_lite099_enc.pt", |
| f"{repo}/koochik_lite099_kmap.json", |
| ) |
| print(model.transcribe(["speech.wav"])[0].text) |
| ``` |
|
|
| ## 📊 Published trade-off |
|
|
| The release evaluated both greedy decoding and an optional Vosk-guided hotword beam. Lower is better. |
|
|
| | Decode | golden-6669 keyword-band WER | FLEURS keyword-band WER | golden-6669 overall WER | FLEURS overall WER | |
| |---|---:|---:|---:|---:| |
| | Full Koochik, greedy | 8.0 | 13.1 | 4.64 | 5.36 | |
| | Koochik Lite, greedy | 12.5 | 18.0 | 6.92 | 7.23 | |
| | Koochik Lite + Vosk guide | 6.4 | 11.7 | 5.30 | 5.31 | |
|
|
| Compression alone reduces quality; the Vosk-guided result requires a separate Vosk first pass plus hotword-aware `pyctcdecode` beam search. Do not compare the guided row to a greedy-only deployment as though they used the same runtime. |
|
|
| ## 🇮🇷 خلاصهٔ فارسی |
|
|
| این مخزن یک مدل کامل و مستقل نیست؛ فقط encoder فشردهشده را نگه میدارد و برای اجرا به فایل NeMo مدل اصلی نیاز دارد. نسخهٔ greedy سبکتر است ولی افت دقت دارد؛ ردیف Vosk-guided به یک مرحلهٔ جداگانهٔ Vosk و beam search نیاز دارد. |
|
|
| ## 🌌 Explore Shenava-1 |
|
|
| [🧠 Full Koochik](https://huggingface.co/Reza2kn/Shenava-Koochik-v1.0) · **🪶 Koochik Lite** · [⚖️ Rizeh 32M](https://huggingface.co/Reza2kn/Shenava-Rizeh-v1.0) · [🐣 Rizeh-Pizeh 6.9M](https://huggingface.co/Reza2kn/Shenava-Rizeh-Pizeh-v1.0) |
|
|
| Apache-2.0. Compression method: [LITEASR](https://arxiv.org/abs/2502.20583). |
|
|
