Fish-Speech 1.5

Training compute

2.7×10²¹ FLOP

Published

Nov 24, 2024

Fish-Speech 1.5 is an AI model developed by Fish Audio (United States), first published in November 2024. It works in the speech domain, on tasks such as speech synthesis and text-to-speech (tts).

Training it took an estimated 2.7×10²¹ FLOP of compute (estimation method: comparison with other models). It was trained on roughly 700B datapoints.

Access: Open weights (non-commercial). Its weights are openly available. Epoch AI rates the confidence of this record as likely.

Full record

Organization: Fish Audio
Country of organization: United States
Domain: Speech
Task: Speech synthesis, Text-to-speech (TTS)
Training compute: 2.7×10²¹ FLOP
Compute estimation method: Comparison with other models
Dataset size: 700B
Model accessibility: Open weights (non-commercial)
Open weights: Yes
Epoch confidence: Likely