ESMFold is a state-of-the-art end-to-end protein folding model based on an ESM-2 backbone. It does not require any lookup or MSA step, and therefore does not require any external databases to be present in order to make predictions.
Model Card
By AI at Meta, published under mit, revision 75a3841ee059.
ESMFold is a state-of-the-art end-to-end protein folding model based on an ESM-2 backbone. It does not require any lookup or MSA step, and therefore does not require any external databases to be present in order to make predictions. As a result, inference time is very significantly faster than AlphaFold2. For details on the model architecture and training, please refer to the accompanying paper. If you're interested in using ESMFold in practice, please check out the associated tutorial notebook.
Read AI at Meta's full model card
ESMFold
ESMFold is a state-of-the-art end-to-end protein folding model based on an ESM-2 backbone. It does not require any lookup or MSA step, and therefore does not require any external databases to be present in order to make predictions. As a result, inference time is very significantly faster than AlphaFold2. For details on the model architecture and training, please refer to the accompanying paper.
If you're interested in using ESMFold in practice, please check out the associated tutorial notebook.
Configuration
- Architecture
- EsmForProteinFolding
- Context length (tokens)
- 1,026
- Layers
- 36
- Hidden size
- 2,560
- Feed-forward size
- 10,240
- Attention heads
- 40
- Vocabulary size
- 33
- Stored precision
- float32
- Model type
- esm
Identity and Version
- Repository
- facebook/esmfold_v1
- Publisher
- AI at Meta
- Task
- Not stated by the source
- Modality
- Other
- Library
- transformers
- Parameters
- Not stated by the source
- Languages
- esm
- Revision
- 75a3841ee059df2bf4d56688166c8fb459ddd97a
- First published
- 2022-11-01
- Last updated
- 2023-03-22
Files and Weights
7 files, 8.4 GB in total. The weights are 1 file totalling 8.4 GB in bin.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| pytorch_model.bin | Weights | 8.4 GB | 2ee07356b125 |
| config.json | Configuration | 2.1 KB | — |
| special_tokens_map.json | Configuration | 121 B | — |
| README.md | Documentation | 696 B | — |
| .gitattributes | Repository | 1.4 KB | — |
| tokenizer_config.json | Tokenizer | 40 B | — |
| vocab.txt | Tokenizer | 72 B | — |
License and Download
- License
- mit
- Access
- Open weights, no gate
- Download size
- 8.4 GB
Released by AI at Meta through its official repository on Hugging Face. Read the license.
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 8.4 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About esmfold_v1
Can I use esmfold_v1 commercially?
Yes. esmfold_v1 is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.
What is esmfold_v1's context length?
1,026 tokens, from the maximum position embeddings in its published configuration.