Downloads · 30 days
5
23% of all-time downloads
hirotakahiraki/parakeet-turntaking-stage1-mixed-libri-interact-top4-ctc-spk
parakeet-turntaking-stage1-mixed-libri-interact-top4-ctc-spk is a machine learning model from hirotakahiraki. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Speech-LLM checkpoint for ASR alignment, trained on a mix of LibriSpeech and Seamless Interact conversational data.
Downloads · 30 days
5
23% of all-time downloads
All-time downloads
22
Public
Repo size
4.8 GB
Likes
0
Public
Click a slice to open those files.
.pt4.8 GB · 100%
From the Hugging Face model README
Speech-LLM checkpoint for ASR alignment, trained on a mix of LibriSpeech and Seamless Interact conversational data.
| Component | Model | Parameters |
|---|---|---|
| Encoder | nvidia/parakeet-ctc-0.6b (FastConformer CTC) | 600M |
| Projector | Conv1d + Transformer (1024 → 896, stride=2) | trainable |
| LLM | Qwen/Qwen2.5-0.5B | 500M |
projector_and_encoder_top — projector fully trainable,
top-4 encoder layers unfrozen| Dataset | Samples | Description |
|---|---|---|
| LibriSpeech 960h | ~276k | Read English speech (standard ASR benchmark) |
| Seamless Interact | ~2.17M | Conversational English speech |
import torch
checkpoint = torch.load("final.pt", map_location="cpu")
# Load into SpeechLLM model — see the parakeet-turntaking repository for details
final.pt — Full model checkpoint (encoder + projector + LLM state dicts)config.yaml — Training configuration