Expand model card
Browse files
README.md
CHANGED
|
@@ -29,14 +29,93 @@ widget:
|
|
| 29 |
|
| 30 |
# boltz2
|
| 31 |
|
| 32 |
-
|
| 33 |
-
It serves the `structure_prediction` task.
|
| 34 |
-
The architecture is `boltz2`.
|
| 35 |
|
| 36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
|
| 38 |
```python
|
| 39 |
from ofoldx import AutoModel
|
| 40 |
|
| 41 |
model = AutoModel.from_pretrained("oteam/boltz2")
|
| 42 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
|
| 30 |
# boltz2
|
| 31 |
|
| 32 |
+
OFoldX `model` artifact for biomolecular structure prediction, using the `boltz2` architecture.
|
|
|
|
|
|
|
| 33 |
|
| 34 |
+
## Disclaimer
|
| 35 |
+
|
| 36 |
+
This model card was generated by the OFoldX team for an OFoldX `model` artifact.
|
| 37 |
+
The upstream model authors did not write this card unless explicitly stated otherwise.
|
| 38 |
+
|
| 39 |
+
OFoldX is pre-alpha research software. Check the source checkpoint, upstream release, and local validation
|
| 40 |
+
before using the artifact for scientific or operational decisions.
|
| 41 |
+
|
| 42 |
+
## Model Details
|
| 43 |
+
|
| 44 |
+
Boltz-2-shaped structure-prediction model converted for the OFoldX runtime.
|
| 45 |
+
|
| 46 |
+
### Model Specification
|
| 47 |
+
|
| 48 |
+
| Field | Value |
|
| 49 |
+
| ----- | ----- |
|
| 50 |
+
| Repository | `oteam/boltz2` |
|
| 51 |
+
| Artifact Kind | `model` |
|
| 52 |
+
| Task | `structure_prediction` |
|
| 53 |
+
| Architecture | `boltz2` |
|
| 54 |
+
| Entrypoint | `ofoldx.models.boltz2.model.Boltz2Model` |
|
| 55 |
+
| Config | `config.json` |
|
| 56 |
+
| Weights | `model.safetensors` |
|
| 57 |
+
|
| 58 |
+
### Links
|
| 59 |
+
|
| 60 |
+
- **Hub repository**: [oteam/boltz2](https://huggingface.co/oteam/boltz2)
|
| 61 |
+
- **Code**: [`ofoldx/models/boltz2/model.py`](https://github.com/OTeam-AI4S/OFoldX/tree/main/ofoldx/models/boltz2/model.py)
|
| 62 |
+
- **Project repository**: [https://github.com/OTeam-AI4S/OFoldX](https://github.com/OTeam-AI4S/OFoldX)
|
| 63 |
+
- **Issues**: [https://github.com/OTeam-AI4S/OFoldX/issues](https://github.com/OTeam-AI4S/OFoldX/issues)
|
| 64 |
+
|
| 65 |
+
## Usage
|
| 66 |
+
|
| 67 |
+
The artifact depends on the [`ofoldx`](https://github.com/OTeam-AI4S/OFoldX) library. Install it with pip:
|
| 68 |
+
|
| 69 |
+
```bash
|
| 70 |
+
pip install ofoldx
|
| 71 |
+
```
|
| 72 |
+
|
| 73 |
+
### Direct Use
|
| 74 |
+
|
| 75 |
+
Load the artifact from `oteam/boltz2` using the OFoldX `Auto*` interface:
|
| 76 |
|
| 77 |
```python
|
| 78 |
from ofoldx import AutoModel
|
| 79 |
|
| 80 |
model = AutoModel.from_pretrained("oteam/boltz2")
|
| 81 |
```
|
| 82 |
+
|
| 83 |
+
When a matching processor is available, load it with `AutoProcessor.from_pretrained(...)` and pass the
|
| 84 |
+
processed batch to the model.
|
| 85 |
+
|
| 86 |
+
### Interface
|
| 87 |
+
|
| 88 |
+
- **Task**: `structure_prediction`
|
| 89 |
+
- **Artifact kind**: `model`
|
| 90 |
+
- **Architecture**: `boltz2`
|
| 91 |
+
- **Runtime files**: `manifest.json`, `config.json`, and `model.safetensors` when present
|
| 92 |
+
|
| 93 |
+
## Training Details
|
| 94 |
+
|
| 95 |
+
The training data and training procedure are inherited from the upstream checkpoint release. They are not
|
| 96 |
+
reproduced inside this OFoldX artifact unless additional source-specific metadata is provided in the files.
|
| 97 |
+
|
| 98 |
+
## Evaluation
|
| 99 |
+
|
| 100 |
+
OFoldX conversion reports and contract tests validate artifact structure and checkpoint loading. Task-level
|
| 101 |
+
scientific evaluation should be checked against the corresponding upstream model release or paper.
|
| 102 |
+
|
| 103 |
+
## Limitations
|
| 104 |
+
|
| 105 |
+
- This artifact is distributed for research use.
|
| 106 |
+
- Inputs must match the model-specific processor and expected biomolecular representation.
|
| 107 |
+
- OFoldX is pre-alpha, so APIs and artifact metadata may still change before a stable release.
|
| 108 |
+
|
| 109 |
+
## Citation
|
| 110 |
+
|
| 111 |
+
Please cite the upstream model or paper associated with the source checkpoint. If OFoldX supports your work,
|
| 112 |
+
please also cite or link the OFoldX project repository.
|
| 113 |
+
|
| 114 |
+
## Contact
|
| 115 |
+
|
| 116 |
+
Please use [OFoldX GitHub issues](https://github.com/OTeam-AI4S/OFoldX/issues) for questions or comments about this model card.
|
| 117 |
+
|
| 118 |
+
## License
|
| 119 |
+
|
| 120 |
+
The OFoldX project license is not yet finalized. The source checkpoint may carry additional upstream license
|
| 121 |
+
terms; review both OFoldX and upstream terms before redistribution or production use.
|