Downloads · 30 days
5
23% of all-time downloads
FluA_Pro_t33_650M_esm2
sihuapeng/FluA_Pro_t33_650M_esm2
FluA_Pro_t33_650M_esm2 is a machine learning model from sihuapeng. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
The FluAPro is a protein model trained on a masked language modeling objective, resulting from the unsupervised fine-tuning of the ESM-2 (https://huggingface.co/facebook/esm2t33650MUR50D) protein language model. Its f…
All-time downloads
22
Public
Parameters
652M
2.6 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.3 GB · 100%
At a glance
- License
- mit
- Model type
- esm
- Access
- Public
- Created
- Apr 28, 2025
- Updated
- May 10, 2025
- SHA
- de824872
Model card
From the Hugging Face model README
FluA_Pro
The FluA_Pro is a protein model trained on a masked language modeling objective, resulting from the unsupervised fine-tuning of the ESM-2 (https://huggingface.co/facebook/esm2_t33_650M_UR50D) protein language model. Its fine-tuning dataset is sourced from Influenza A protein sequences (11 types of influenza A proteins from different species) in the NCBI Virus database (as of January 12, 2025). The original dataset comprises 1,737,586 protein sequences, reduced to 388,385 sequences after removing duplicates and those with 100% identity. The FluA_Pro model is suitable for downstream applications related to Influenza Aviruses. For detailed information on the model’s training data, please refer to the accompanying paper. Multiple FluA_Pro checkpoints are available on the Hub, with varying model sizes. Generally, larger models offer higher accuracy but require more memory and training time.
| Checkpoint name | Num layers | Num parameters |
|---|---|---|
| .FluA_t36_3B_esm2 | 36 | 3B |
| FluA_t33_650M_esm2 | 33 | 650M |
| FluA_t30_150M_esm2 | 30 | 150M |
| FluA_t12_35M_esm2 | 12 | 35M |
| FluA_t6_8M_esm2 | 6 | 8M |
Results on Epoch 20:
Training Metrics:
Training Loss: 0.0856
Training Perplexity: 1.0894
Evaluation Metrics:
Evaluation Loss: 0.0888
Evaluation Perplexity: 1.0929
Note: New data.
- Type
- esm
- License
- mit
- Created
- Apr 28, 2025
- Updated
- May 10, 2025