Instructions to use open-athena/marinfold-exp199 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use open-athena/marinfold-exp199 with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("open-athena/marinfold-exp199", device_map="auto") - Notebooks
- Google Colab
- Kaggle
MarinFold exp199 checkpoints
Checkpoint artifacts from the MarinFold contacts-v1 AFDB/ESM optimization experiment tracked in Open-Athena/MarinFold#199.
Runs
| Run | Final eval/tokenized/contacts-v1-val/loss |
Levanter checkpoints | Hugging Face export |
|---|---|---|---|
prot-exp199-cv1-s01-m1-p03-aug-us-east1 |
3.0115306377 |
step-72599 | step-72599 |
prot-exp199-cv1-s01-m1-p03-base-us-east5 |
3.0074224472 |
step-72599 | step-72599 |
prot-exp199-cw-cv1-s02-m1-p06-aug |
2.9712009430 |
step-116160, step-145199 | step-145199 |
Formats
The checkpoints/ paths contain the original Levanter OCDBT full
state, including AdamW state for continued training. The hf/
paths contain Qwen3ForCausalLM safetensors exports and the
2,845-token contacts-v1 tokenizer.
The two m1-p03 runs trained for 72,600 steps, making
step-72599 their zero-indexed final checkpoint. The CoreWeave
m1-p06 run trained for 145,200 steps. Its
step-116160 checkpoint marks 80% of training before cooldown and
is retained as a continuation point; step-145199 is the final
checkpoint.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support