Downloads · 30 days
0
minimind-o-lfm-moss-smart60k
MingSafeR/minimind-o-lfm-moss-smart60k
minimind-o-lfm-moss-smart60k is a text-to-speech model from MingSafeR. Use it when you need text read aloud.
Final export recovered from the verified completed training checkpoint.
Access
Public
Updated Sep 10, 2026
Repo size
12.5 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt4.3 GB · 100%
At a glance
- Task
- Text-to-Speech
- Access
- Public
- Created
- Sep 10, 2026
- Updated
- Sep 10, 2026
- SHA
- 6ec5a247
Model card
From the Hugging Face model README
MiniMind-O LFM + MOSS Smart60K
Final export recovered from the verified completed training checkpoint.
- Thinker:
LiquidAI/LFM2.5-350M@9e6c6ccf47cd318696e137d381a7ded8fe4df09f - Talker:
OpenMOSS-Team/MOSS-TTS-Nano-100M@44502f80dbf9743528fa921cc544d662c685ebec - PRE:
MingSafeR/miniloop-o-smart60k-moss16rvq - T2A: 20K
- A2A: 15K
- I2T: 25K
- Total: 60K
- GPU used for training: A100 40GB
- Final optimizer step: 1875
- Final training rows: 60000
joint_core.pt is the exact verified checkpoint/latest/core.pt from step 1875.
Frozen SenseVoice and SigLIP2 weights are not embedded in joint_core.pt.
- Task
- Text-to-Speech
- Created
- Sep 10, 2026
- Updated
- Sep 10, 2026