Search results

Why is the model size in Evo2 set to 40.3 B parameters? Is it because of the 9.3 T tokens of training data?

Status: Open.

#24 In togethercomputer/stripedhyena;

· endNone opened on Jun 5, 2025

Any thoughts on a GRPO version?

Status: Open.

#23 In togethercomputer/stripedhyena;

· sydneyrenee opened on Feb 11, 2025

gradient checkpointing is not implement

Status: Open.

#22 In togethercomputer/stripedhyena;

· xiyang-aads-lilly opened on Oct 15, 2024

evo and stripedhyena crash the server when doing a simple inference

Status: Open.

#21 In togethercomputer/stripedhyena;

· sun-qibo opened on Aug 28, 2024

flash attention not compatible?

Status: Open.

#20 In togethercomputer/stripedhyena;

· oxPJ opened on Aug 25, 2024

Limiting attention radius and extracting embeddings

Status: Open.

#19 In togethercomputer/stripedhyena;

· george-henderson opened on May 15, 2024

tokenizer.model file in addition to tokenizer.json

Status: Open.

#14 In togethercomputer/stripedhyena;

· kenneth-miura opened on Feb 22, 2024

Apple Silicon support

Status: Open.

#5 In togethercomputer/stripedhyena;

· amrohendawi opened on Jan 20, 2024

Pretrain

Status: Open.

#1 In togethercomputer/stripedhyena;

· mietekrmd opened on Dec 8, 2023