MMS models
3 MMS models on Hugging Face, from 315M to 965M parameters. At the precision each one is published in, the smallest needs about 1.4 GB of VRAM (mms-300m-1130-forced-aligner, cheapest live fit: V100) and the largest about 4.3 GB (mms-1b-all, cheapest live fit: V100). The cheapest way to run mms-300m-1130-forced-aligner is $0.088/hr.
MMS models
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| mms-300m-1130-forced-aligner | 315M | F32 | 1.4 GB | V100 | $0.088/hr |
| wav2vec2-large-mms-1b-azerbaijani-common_voice15.0 | 965M | F32 | 4.3 GB | V100 | $0.088/hr |
| mms-1b-all | 965M | F32 | 4.3 GB | V100 | $0.088/hr |
VRAM is for the precision the model is published in, with the same overhead and an 8,192-token context assumed on every page; see the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
More from Meta
- Wav2Vec 2.0 (32 models)
- facebook (3 models)
