Downloads · 30 days
8
9% of all-time downloads
parakeet-tdt-0.6b-v3-fp16-onnx
Masterx/parakeet-tdt-0.6b-v3-fp16-onnx
parakeet-tdt-0.6b-v3-fp16-onnx is a machine learning model from Masterx. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for nemo. The card lists the license as cc-by-4.0.
fp16 conversion of istupakov/parakeet-tdt-0.6b-v3-onnx (same graph layout: encoder-model + fused decoderjoint-model, .fp16 suffix), made with onnxruntime's transformers.float16 converter (keepiotypes=True, path-based…
All-time downloads
90
Public
Repo size
1.3 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.onnx1.3 GB · 100%
At a glance
- Library
- nemo
- License
- cc-by-4.0
- Model type
- nemo-conformer-tdt
- Access
- Public
- Created
- Jul 11, 2026
- Updated
- Jul 11, 2026
- SHA
- 93883c8c
Base models
Model card
From the Hugging Face model README
parakeet-tdt-0.6b-v3 — fp16 ONNX
fp16 conversion of istupakov/parakeet-tdt-0.6b-v3-onnx
(same graph layout: encoder-model + fused decoder_joint-model, .fp16 suffix), made with
onnxruntime's transformers.float16 converter (keep_io_types=True, path-based shape inference
for the >2GB encoder). I/O stays fp32, weights/compute fp16.
Verified vs the fp32 export: random-input parity cos ≈ 1.0, and END-TO-END transcripts are byte-identical on DirectML while decoding ~2× faster (66 s clip: 231 ms vs 459 ms warm, RTX 3080 Ti; encoder 92 ms vs 300 ms). Half the download/VRAM of fp32.
Made for WinSTT.
- Library
- nemo
- Type
- nemo-conformer-tdt
- License
- cc-by-4.0
- Created
- Jul 11, 2026
- Updated
- Jul 11, 2026