Downloads · 30 days
20
23% of all-time downloads
dys-asr/parakeet-rnnt-0.6b-all-syn-chunk-cutout-fade-bpedrop
parakeet-rnnt-0.6b-all-syn-chunk-cutout-fade-bpedrop is a automatic speech recognition model from dys-asr. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as other.
The ten-epoch recipe behind dys-asr/parakeet-rnnt-0.6b-all-syn-chunk-cutout, rerun with both FadeOutIn and BPE-dropout at p = 0.1. The two single-factor arms exist so this one can be attributed.
Downloads · 30 days
20
23% of all-time downloads
All-time downloads
87
Public
Parameters
618M
2.5 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors2.5 GB · 100%
From the Hugging Face model README
The ten-epoch recipe behind
dys-asr/parakeet-rnnt-0.6b-all-syn-chunk-cutout,
rerun with both FadeOutIn and BPE-dropout at p = 0.1. The two single-factor
arms exist so this one can be attributed.
On the held-out dev split, against the run it varies:
| run | best epoch | dev CER (4,000-utterance subset) | dev CER (full split) |
|---|---|---|---|
| baseline, no change | 9 | 6.06% | 6.25% |
| this model | 10 | 6.13% | 6.12% |
Treat both columns with suspicion. The subset and the full split disagree about the ordering across this family of runs, the spread is under 0.2 CER points, and there is one seed behind each row. The full split is 17,492 utterances against the subset's 4,000 and is the more trustworthy of the two, but neither is a significance test.
Every run in this family peaked at epoch 9 or 10 with no sign of overfitting, so ten epochs may simply be short.
FadeOutIn at sixteen segments of 2%, and BPE-dropout at 0.1; the sibling cards describe each. On this corpus the combination scored between the two single arms rather than below both, which is the one thing these four runs say about whether they compose. One seed, so not much.
The same corpus as
dys-asr/parakeet-rnnt-0.6b-all-syn-chunk-cutout,
unchanged, so the two are directly comparable.
| source | records |
|---|---|
| SAPC-1 train and dev | 250,014 |
| SAPC-2 train | 153,500 |
| synthetic CosyVoice3 speech | 55,988 |
| chunks force-aligned out of over-length recordings | 8,933 |
| total before filtering | 468,435 |
463,177 records survive the 0.5-45 s and 200-label-token filters. Nothing comes from outside the challenge corpora, so this is a constrained-track model.
The held-out dev split is genuinely held out. sapc2_dev_heldout.jsonl is
17,582 clips and its intersection with the training manifests is empty; that was
checked by comparing file paths, not assumed.
Ten epochs on sixteen GH200s, effective batch 32 at per-device 2, AdamW at 1e-4 with a tri-stage schedule (10% warmup, 40% hold), weight decay 0.01, layerdrop 0.05, gradient clip 1.0, bf16, seed 42, 144,750 optimizer steps. Augmentation is online speed perturbation over 0.8-1.2, SpecAugment (5% of the time axis in spans of 10 frames, 40% of the mel axis in spans of 27 bins) and SpecCutout (two 20x20 rectangles).
import soundfile as sf
import torch
from transformers import AutoProcessor, ParakeetForRNNT
model_id = "dys-asr/parakeet-rnnt-0.6b-all-syn-chunk-cutout-fade-bpedrop"
device = "cuda" if torch.cuda.is_available() else "cpu"
processor = AutoProcessor.from_pretrained(model_id)
model = ParakeetForRNNT.from_pretrained(model_id).to(device).eval()
audio, rate = sf.read("utterance.wav", dtype="float32")
inputs = processor(audio, sampling_rate=rate, return_tensors="pt").to(device)
with torch.inference_mode():
generated = model.generate(**inputs)
print(processor.batch_decode(generated, skip_special_tokens=True)[0])
Requires transformers>=5.9. Audio must be 16 kHz mono.
A competition entry for the Speech Accessibility Project Challenge, and research on atypical-speech recognition.
transformers>=5.9. Not loadable by the 4.x line.The Speech Accessibility Project corpora are governed by their own data use agreement and are not redistributed here; reproducing this training set requires authorized access. Fun-CosyVoice3's terms apply to the synthesis component.