Downloads · 30 days
309
23% of all-time downloads
Tonykip/nemotron-3.5-swahili-streaming-asr
nemotron-3.5-swahili-streaming-asr is a automatic speech recognition model from Tonykip. Use it when you need speech turned into text. It is set up for nemo. The card lists the license as other.
A low-latency streaming Swahili ASR model: NVIDIA's cache-aware FastConformer-RNNT (nvidia/nemotron-3.5-asr-streaming-0.6b, 0.6B) fine-tuned to add Swahili, a language the base model does not officially support. Unlik…
Downloads · 30 days
309
23% of all-time downloads
All-time downloads
1.3K
Public
Repo size
8.3 GB
Likes
7
Public
Click a slice to open those files.
.onnx_data3 GB · 53%
From the Hugging Face model README
A low-latency streaming Swahili ASR model: NVIDIA's cache-aware FastConformer-RNNT
(nvidia/nemotron-3.5-asr-streaming-0.6b, 0.6B) fine-tuned to add Swahili, a language the
base model does not officially support. Unlike Whisper (a 30-second-window batch model), this
model transcribes incrementally with sub-100 ms latency — for live captioning, voice agents,
and dictation.
This checkpoint is a weight-average ("soup") of the 5 best fine-tuning checkpoints, each selected on the held-out FLEURS Swahili test set (not the training-time dev metric).
sw_ke test (n=487, identical streaming pipeline, lowercase+punct-strip)| attention context (lookahead) | WER | CER |
|---|---|---|
[56,13] (~480 ms) | 0.298 | 0.106 |
[56,0] (~0 ms, lowest latency) | 0.322 | 0.115 |
base model [56,13] (no Swahili FT) | 1.147 | 0.860 |
The base model is effectively unusable for Swahili (~100% WER, degenerate ⁇ loops);
fine-tuning brings it to ~30% WER / ~11% CER — within the range NVIDIA reports for its
officially supported languages. For reference, offline (non-streaming) Whisper-large-v3 (1.5B)
on the same test set is in a similar WER range but cannot stream.
Good for: streaming/live Swahili transcription, voice agents (with an LLM downstream), live captioning, dictation aids, gisting. Not for: unattended verbatim transcription (legal/medical) without review.
Limitations:
att_context_size for your latency/accuracy budget: more lookahead ([56,13]) is more
accurate; [56,0] is lowest latency.import nemo.collections.asr as nemo_asr
m = nemo_asr.models.ASRModel.restore_from("nemotron-3.5-swahili-streaming-0.6b.nemo")
# Streaming inference: use NeMo's
# examples/asr/asr_cache_aware_streaming/speech_to_text_cache_aware_streaming_infer.py
# with target_lang=sw-KE, att_context_size=[56,13] (or [56,0]), decoder_type=rnnt.
nvidia/nemotron-3.5-asr-streaming-0.6b (FastConformer-RNNT, cache-aware streaming).badrex/swahili-speech-400hr (Afrivoice / DigitalUmuganda,
CC-BY-4.0); speaker-disjoint dev split. A 30 h clean subset is at Tonykip/kenyan-swahili-asr-clean.ONNX exports are in onnx/ for CPU / desktop / edge streaming via ONNX Runtime
(Rust ort crate; parakeet-rs implements the streaming chunk + RNNT greedy loop):
encoder-swahili.int8.onnx (+_data, ~650 MB) — int8 encoder; the macOS/edge build (4× smaller than fp32)encoder-swahili.onnx (+_data, ~2.4 GB) — fp32 masterdecoder_joint-swahili.onnx — RNNT predictor + joint (keep fp32)prompt_kernel.pt + prompt_info.json — the Swahili language prompt (sw-KE:48), applied host-side (the exported encoder is prompt-less)onnx/INTEGRATION.md — cache I/O + prompt + streaming-loop specGraphs are NeMo check_trace-verified (ONNX≈PyTorch). End-to-end streaming WER is validated in the
runtime. GGUF is not applicable (ggml/whisper.cpp is Whisper-only) — ONNX Runtime is the CPU/edge path.
Derivative of nvidia/nemotron-3.5-asr-streaming-0.6b under OpenMDW-1.1 — see the base model
for license terms. Training data: badrex/swahili-speech-400hr (CC-BY-4.0, Afrivoice/DigitalUmuganda).
Please attribute both NVIDIA and the data authors.