Continue finetuning Instruct model using LoRA from 0.5B up to 72B.
Mesolitica
company
AI & ML interests
We develop Multimodality, lab from Malaysia
Recent Activity
View all activity
Collections
13
models
271
mesolitica/Malaysian-Audio-Qwen2.5-14B-Instruct
Updated
mesolitica/Qwen2.5-72B-Instruct-FP8
Updated
mesolitica/Malaysian-Audio-Qwen2.5-7B-Instruct
Updated
mesolitica/Malaysian-Qwen2.5-14B-Reasoning-GRPO
Updated
•
22
•
1
mesolitica/Malaysian-Qwen2.5-1.5B-Reasoning-GRPO
Updated
•
4
mesolitica/Malaysian-Podcast-sesame-csm-1b
Updated
•
9
mesolitica/Malaysian-sesame-csm-1b
Updated
•
37
mesolitica/Malaysian-Dia-1.6B
Updated
•
825
mesolitica/Malaysian-Podcast-Dia-1.6B
Updated
•
363
mesolitica/Malaysian-Qwen2.5-1.5B-Reasoning-SFT
Updated
•
124
datasets
202
mesolitica/Transcription-Instructions
Viewer
•
Updated
•
2.63M
•
48
mesolitica/Emilia-Mandarin-Description-Instructions
Viewer
•
Updated
•
184k
•
8
mesolitica/Cantonese-Radio-Description-Instructions
Viewer
•
Updated
•
339k
•
32
mesolitica/CompA-R-Instructions
Viewer
•
Updated
•
199k
•
12
mesolitica/Zeroshot-Audio-Classification-Instructions
Viewer
•
Updated
•
792k
•
28
mesolitica/Malaysian-Multiturn-Chat-Assistant
Viewer
•
Updated
•
150k
•
64
mesolitica/Extra-Emilia
Viewer
•
Updated
•
1.62M
•
208
mesolitica/MusicBench-Instructions
Viewer
•
Updated
•
151k
•
74
mesolitica/Speaker-Diarization-Instructions
Viewer
•
Updated
•
105k
•
70
mesolitica/Animal-Sound-Instructions
Viewer
•
Updated
•
28.3k
•
70