Evo 2

30-60 minutesPaperCodeNeed >5GB

Genome modeling and design across all domains of life

Readiness

Readiness: Need >5GB
Smoke: not_tested
Artifacts: not_applicable
Production: need_gt5gb_cuda_evo2_model_runtime
Storage: >5
Strategy: large CUDA SIF plus Evo 2 model cache phase
Next: Defer to large CUDA/model phase; provision Evo 2 model weights and compatible SIF before direct Slurm smoke and runner wiring.

No obvious large model/database dependency identified; validate upstream docs before install. 2026-07-05 triage: official Evo 2 runtime requires CUDA 12.1+, cuDNN 9.3+, Python 3.11/3.12, Torch 2.6/2.7, FlashAttention, and model weights; the 7B model is the smallest practical backend target and the current shared small SIF path is insufficient. Reclassified from Smoke Pending to Need >5GB.