Shenava — Koochik v1.0 (114M) · CTC Non-Streaming · ONNX (INT8)

Koochik (کوچیک, "small") is the 114M teacher / flagship of the Shenava‑1 family — a FastConformer Hybrid RNNT/CTC model fine‑tuned on clean Persian data. This repository contains the CTC non-streaming INT8 quantized ONNX export for use with sherpa-onnx.

Model Details

Property Value
Parent Model Reza2kn/Shenava-Koochik-v1.0
Architecture FastConformer Hybrid RNNT/CTC → CTC decoder export
Language Persian (Farsi)
Parameters 114M
Vocabulary Size 1025 tokens (BPE + blank)
Tokenizer ve_tok_v4 — SentencePiece BPE‑1024
Sample Rate 16 kHz
Subsampling
Streaming ❌ No (full-context)
Weight Type INT8 (dynamic quantization)
Author Reza2kn
Version 1.0

Files

.
├── README.md
├── model.int8.onnx   # ONNX model (INT8 quantized)
└── tokens.txt        # Token vocabulary (1025 tokens)

Usage

sherpa-onnx-offline \
  --tokens=tokens.txt \
  --encoder=model.int8.onnx \
  --num-threads=4 \
  /path/to/test.wav

Conversion

Exported from a NeMo checkpoint and quantized dynamically with ONNX Runtime.

License

CC-BY-NC 4.0 (inherited from the parent model)

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for mah92/sherpa-onnx-nemo-ctc-fa-shenava-koochik-v1.0-non-streaming-int8-2026-06-26

Finetuned
(5)
this model

Datasets used to train mah92/sherpa-onnx-nemo-ctc-fa-shenava-koochik-v1.0-non-streaming-int8-2026-06-26