Models
rbio1 is a 3B biology reasoning model from the Chan Zuckerberg Initiative, post-trained from Qwen2.5-3B-Instruct by reinforcement learning (GRPO) with…
An Amharic-language variant of Gemma-3-27B-MeditronFO, from EPFL's LiGHT laboratory and the MeditronFO pipeline.
A medical specialisation of Gemma-3-27B-it, fine-tuned on the Fully Open Meditron corpus (about 601k clinician-vetted medical QA and clinical-vignette…
A medical specialisation of EuroLLM-9B-Instruct, fine-tuned on the Fully Open Meditron corpus (about 601k clinician-vetted medical QA and clinical-vignette…
A medical specialisation of EuroLLM-22B-Instruct, fine-tuned on the Fully Open Meditron corpus (about 601k clinician-vetted medical QA and clinical-vignette…
A medical specialisation of OLMo-2-32B, fine-tuned on the Fully Open Meditron corpus (about 601k clinician-vetted medical QA and clinical-vignette examples)…
A medical specialisation of Apertus-8B-Instruct, fine-tuned on the Fully Open Meditron corpus (about 601k clinician-vetted medical QA and clinical-vignette…
A medical specialisation of Apertus-70B-Instruct, fine-tuned on the Fully Open Meditron corpus (about 601k clinician-vetted medical QA and clinical-vignette…
Baichuan-M3 is a clinical large language model from Baichuan Inc., a 235B mixture-of-experts model built on Qwen3-235B-A22B.
MedGemma 1.5 is Google's 4B multimodal medical model, built on Gemma 3.
Cell2Sentence-Scale (C2S-Scale) model from the van Dijk lab (Yale) and Google, fine-tuned from Gemma 2 27B.
Cell2Sentence-Scale (C2S-Scale) model from the van Dijk lab (Yale) and Google, fine-tuned from Gemma 2 2B.
TxGemma is a family of Google models fine-tuned from Gemma 2 on therapeutic-development tasks from the Therapeutics Data Commons (TDC), covering small…
TxGemma is a family of Google models fine-tuned from Gemma 2 on therapeutic-development tasks from the Therapeutics Data Commons (TDC), covering small…
TxGemma is a family of Google models fine-tuned from Gemma 2 on therapeutic-development tasks from the Therapeutics Data Commons (TDC), covering small…
TxGemma is a family of Google models fine-tuned from Gemma 2 on therapeutic-development tasks from the Therapeutics Data Commons (TDC), covering small…
TxGemma is a family of Google models fine-tuned from Gemma 2 on therapeutic-development tasks from the Therapeutics Data Commons (TDC), covering small…
Biomni-R0-32B-Preview is a biomedical AI agent model from Stanford (Leskovec lab): Qwen3-32B trained by multi-turn reinforcement learning inside the Biomni…
Intern-S2-Preview-397B is the largest Intern-S2 preview from Shanghai AI Laboratory (InternLM): a 397B multimodal model for scientific reasoning and…
Intern-S2-Preview is a 35B multimodal scientific foundation model from Shanghai AI Laboratory (InternLM), trained on hundreds of professional scientific tasks…
Intern-S1-Pro is a trillion-parameter mixture-of-experts multimodal scientific reasoning model from Shanghai AI Laboratory (InternLM): 1T parameters in 512…
Intern-S1-mini is the lightweight Intern-S1 from Shanghai AI Laboratory (InternLM): an 8B Qwen3 language model with a 0.3B InternViT vision encoder, further…
Intern-S1 is an open multimodal scientific reasoning model from Shanghai AI Laboratory (InternLM): a Qwen3-235B mixture-of-experts language model with a 6B…
A lexical baseline that scores a biomedical abstract against a fixed set of therapeutic areas.
A deterministic calculator rather than a learned model: average mass from an amino-acid sequence.
This is a convenient code wrapper to run Lundberg lab SubCell model in inference with your own images
a generalist algorithm for cellular segmentation with human-in-the-loop capabilities
Implementation of DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking
Message Passing Neural Networks for Molecule Property Prediction
The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level.
GEARS is a geometric deep learning model that predicts outcomes of novel multi-gene perturbations
State is a machine learning model that predicts cellular perturbation response across diverse contexts
Deep probabilistic analysis of single-cell and spatial omics data
A unifying representation of single cell expression profiles that quantifies similarity between expression states and generalizes to represent new studies…
UCE is a zero-shot foundation model for single-cell gene expression data
Source code repository mirrored from GitHub (biomap-research/scFoundation).
Source code repository mirrored from GitHub (bowang-lab/scGPT).
Source code repository mirrored from GitHub (dauparas/LigandMPNN).
Trainable, memory-efficient, and GPU-friendly PyTorch reproduction of AlphaFold 2
Channel-agnostic image encoding model CA-MAE with a ViT-S/16 encoder backbone designed for microscopy image featurization.
MoLFormer is a class of models pretrained on SMILES string representations of up to 1.1B molecules from ZINC and PubChem.
State Embedding model checkpoints are meant to be used with the Arc State
Geneformer Geneformer is a foundational transformer model pretrained on a large-scale corpus of human single cell transcriptomes to enable context-aware…
The code for using the CodonFM model checkpoint is available in the official Github
The code for using the CodonFM model checkpoint is available in the official Github
Pre-trained model on non-coding RNA (ncRNA) using a masked language modeling (MLM)
Pre-trained model on non-coding RNA (ncRNA) using a masked language modeling (MLM)
Pre-trained model on non-coding RNA (ncRNA) using a masked language modeling (MLM)
Pre-trained model on non-coding RNA (ncRNA) using a masked language modeling (MLM)
Evo 2 is a state-of-the-art DNA language model trained autoregressively on trillions of DNA
Evo 2 is a state-of-the-art DNA language model trained autoregressively on trillions of DNA
Evo 2 is a state-of-the-art DNA language model trained autoregressively on trillions of DNA
Using Caduceus To use the pre-trained model for masked language modeling, use the following snippet: python from transformers import AutoModelForMaskedLM,
Using Caduceus To use the pre-trained model for masked language modeling, use the following snippet: python from transformers import AutoModelForMaskedLM,
This is the official pre-trained model introduced in DNABERT-2: Efficient Foundation Model and Benchmark For Multi-Species
The Nucleotide Transformers are a collection of foundational language models that were pre-trained on DNA sequences from whole-genomes.
The Nucleotide Transformers are a collection of foundational language models that were pre-trained on DNA sequences from whole-genomes.
The Nucleotide Transformers are a collection of foundational language models that were pre-trained on DNA sequences from whole-genomes.
Chai-1 is a multi-modal foundation model for molecular structure prediction that performs at the state-of-the-art across a variety of benchmarks.
Boltz-1 is an open-source model which predicts the 3D structure of proteins, rna, dna and small molecules; it handles modified residues, covalent ligands and…
ESMFold is a state-of-the-art end-to-end protein folding model based on an ESM-2 backbone.
esm3-sm-open-v1 is trained on 2.78 billion natural proteins.
This set of model weights was released with the GitHub-compatible esm package format.
Mirror of the base ProGen2-base model (with slightly modified configuration and forward pass) by Nijkamp, et
Mirror of the base ProGen2-small model (with slightly modified configuration and forward pass) by Nijkamp, et
AMPLIFY is an efficient, state-of-the-art protein language model pre-trained using masked language modeling on UniRef100, OAS, and SCOP (UR100P).
AMPLIFY is an efficient, state-of-the-art protein language model pre-trained using masked language modeling on UniRef100, OAS, and SCOP (UR100P).
An encoder-only, half-precision version of the ProtT5-XL-UniRef50 model.
Pretrained model on protein sequences using a masked language modeling (MLM) objective.
ESM-2 is a state-of-the-art protein model trained on a masked language modelling objective.
ESM-2 is a state-of-the-art protein model trained on a masked language modelling objective.
ESM-2 is a state-of-the-art protein model trained on a masked language modelling objective.
An Evolutionary-scale Model (ESM) for protein function prediction from amino acid sequences using the Gene Ontology (GO).
Model description - Intended uses and limitations - How to use - Limitations and bias - Training - Tokenization and model pretraining - Training corpora and…
Model artifact mirrored from the HuggingFace Hub.
SMAFormer: Synergistic Multi-Attention Transformer for Medical Image
Model artifact mirrored from the HuggingFace Hub.
Model artifact mirrored from the HuggingFace Hub.
This model performs both segmentation and classification on chest radiographs (X-rays).
This repository contains a fine-tuned Vision Transformer (ViT) model for classifying chest X-ray images, utilizing the CheXpert dataset.
Model artifact mirrored from the HuggingFace Hub.
Model Card for ChemBERTa-10M-MTR Model Details Model Description More information needed - Developed by: DeepChem - Shared by [Optional]:
This repository contains 82 Clan-level protein language model checkpoints generated for the study “Does domain-specific unsupervised fine-tuning improve…
This repository contains 171 Clan-level protein language model checkpoints generated for the study “Does domain-specific unsupervised fine-tuning improve…
This repository contains 254 Pfam-level protein language model checkpoints generated for the study “Does domain-specific unsupervised fine-tuning improve…
This repository contains 356 Pfam-level protein language model checkpoints generated for the study “Does domain-specific unsupervised fine-tuning improve…