YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Celerity 906M — 4k — ad0p2-ild
Celerity 906M checkpoint converted from Cerebras CS format to Hugging Face format.
- Sequence length: 4k
- Attention-dropout variant: ad0p2-ild
- Source checkpoint: checkpoint_29117
- Source runtime experiments: cbcore 2.6.0
- Training rerun: corrected no-repeat 906M CommonPile experiment
- Source experiment root: cp_906_msl4k_norepeat
- Converted with strict checkpoint-key matching.
This revision replaces the earlier 906M/4K model trained from the undersized repeating HDF5 pool. Previous Hugging Face revisions remain available in the repository history.
The model uses custom Celerity Hugging Face modeling code and should be loaded
with trust_remote_code=True.
- Downloads last month
- 31
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support