Llasa-3B

HKUSTAudio/Llasa-3B

Text to Speech4B parametersLicense: cc-by-nc-4.0
Tracked by OpenModelStats since September 6, 2026View on Hugging Face ↗
Downloads 30D
341
Total downloads
36.8K
Gained 7D
+58(increase)
0.2%(up) vs baseline
Gained 30D
—
Likes
531
+0 in 7D
Rank · tracked models
#47,904
665(down 665 places)this week
Parameters
4B
Last source update
May 10, 2025
Momentum
26.5
  • Llasa-3B gained 58 downloads during the last seven tracked days.
  • It moved from #513 to #519 among tracked Text to Speech models this week.

Downloads over time

Daily gain between consecutive observations

Likes over time

Rank among tracked models

Lower is better

Overview

Llasa-3B is a text to speech model published by HKUSTAudio. OpenModelStats has tracked the model since Sep 6, 2026, most recently observing it Oct 6, 2026.

Published
Jan 7, 2025
Last updated
May 10, 2025
Library
—
License
cc-by-nc-4.0
Task
Text to Speech
Parameters
4,009,454,592
Spaces
—
Derivative models
1
Velocity 7D/day
8
safetensorsllamaText-to-Speechtext-to-speechzhenarxiv:2502.04128base_model:meta-llama/Llama-3.2-3B-Instructbase_model:finetune:meta-llama/Llama-3.2-3B-Instructlicense:cc-by-nc-4.0region:us

Publisher

HKUSTAudioView publisher statistics →

Derived from meta-llama/Llama-3.2-3B-Instruct (reported by the source; parent not yet tracked).

Model family

Based on base-model relationships reported by the source's metadata.

Derivative models

Similar models

HKUSTAudio/Llasa-3B — Downloads, Growth & Stats | OpenModelStats