fair-esm

library2.0.0pypypiunverified

FAIR ESM provides pretrained transformer language models for proteins, including ESM-2 and ESM-1b. Version 2.0.0 adds new models and enhancements. The library is actively maintained by Meta AI.

ai-ml llm-agents

pip install fair-esm

github ↗package ↗

INSTALL

IMPORT

SIG · FAIR-ESM

fair-esm

ai-mlpythonv2.0.0

Install

1.5s avg

Import

—

Disk

16MB

Pass rate

10/ 10

Env Coverage10 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v2.0.0 · pip install

no network on importno background threads

Install × environment matrix

Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux

musl

py 3.10–3.95 runs

installs and imports cleanly · install 0.0s · import 0.000s · 18.4MB

glibc

py 3.10–3.95 runs

installs and imports cleanly · install 1.5s · import 0.000s · 19MB

16MB installed

● package 16MB

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

esm2_t48_15B_UR50D

✓ from esm.pretrained import esm2_t48_15B_UR50D

✗ from esm.pretrained import esm2_t48_15B_UR50D

Load an ESM-2 model, tokenize sequences, and extract representations.

import torch from esm.pretrained import esm2_t33_650M_UR50D from esm import Alphabet model, alphabet = esm2_t33_650M_UR50D() batch_converter = alphabet.get_batch_converter() model.eval() data = [ ("protein1", "MKTVRQERLKSIVRILERSKEPVSGAQLAEELSVSRQVIVQDIAYLRSLGYNIVATPRGYVLAGG"), ] batch_labels, batch_strs, batch_tokens = batch_converter(data) with torch.no_grad(): results = model(batch_tokens, repr_layers=[33], return_contacts=True) token_representations = results["representations"][33] print(token_representations.shape)

Debug

Known issues

breakingIn v2.0.0, model loading through torch.hub is deprecated; use esm.pretrained instead.

fix

Use from esm.pretrained import ... instead of torch.hub.load('facebookresearch/esm', ...)

affects: >=2.0.0

breakingThe 'esm.pretrained' module no longer provides the old ESM-1b model alias 'esm1b_t33_650M_UR50S' directly; use 'esm1b_t33_650M_UR50S()' remains but check model list.

fix

Refer to the model zoo documentation for updated model names.

affects: >=2.0.0

gotchaGPU memory is very high (e.g., ESM-2 15B requires ~300GB); smaller models like 650M are recommended for most users.

fix

Use esm2_t33_650M_UR50D (650M params) for typical usage.

affects: all

gotchaThe 'return_contacts' flag in forward() can cause OOM; only use if needed.

fix

Set return_contacts=False (default) unless you need attention map.

affects: all

deprecatedThe 'esm.model' submodule is being reorganized; direct class imports may break in future versions.

fix

Use esm.pretrained for model loading and esm for Alphabet.

affects: <2.0.0

Upgrade

Version history

2.0.0latest on PyPI · released Nov 1, 2022

Audit

Dependencies

No dependency data recorded yet.

Agent activity

47 hits · last 30 days

node

OpenAI (training)

Resources

githubhttps://github.com/facebookresearch/esm ↗

packagehttps://pypi.org/project/fair-esm/ ↗