ESM (CPU)
Embed protein sequences on CPU.
Results
Loading starting example…
Method & evidence
Generate mean embeddings with ESMC 300M or 600M on the CPU.
Scope & limitations
- The public CPU lane intentionally excludes ESMC-6B, ESM3 generation, and ESMFold2.
Interpreting results
- Embedding coordinates are model features, not calibrated biochemical measurements.
- Compare vectors only when checkpoint and pooling are identical.
References
Input requirements & outputs
Input checks
- Canonical amino acids plus X
- At most 8 rows and 4,096 total residues
- At most 2,046 residues per sequence
Expected results
- Annotated sequence Dataset — Embedding vectors, dimensions, and model details for each source row.
- Downloadable results — Complete embeddings and model settings.
