Update README.md
Browse files
README.md
CHANGED
|
@@ -36,7 +36,7 @@ This Hugging Face package is designed for immediate use after download, rapid lo
|
|
| 36 |
| Scenario | Description |
|
| 37 |
| :---: | :--- |
|
| 38 |
| Complex structure prediction | Takes a Protenix JSON file and locally generated MSA features as input and outputs predicted structures in CIF format and confidence scores in JSON format. |
|
| 39 |
-
| Full Hugging Face package validation | Uses the included `config/`, `models/`, `scripts/`, `examples/`, and `
|
| 40 |
| Fine-tuning pipeline validation | Starts the fine-tuning workflow using the included weights and `ft_datasets/finetune_subset.txt`. |
|
| 41 |
| Training pipeline setup | Supports single-GPU training once the complete Protenix dataset is provided. |
|
| 42 |
|
|
@@ -87,8 +87,9 @@ pip install onescience[bio] -i http://mirrors.onescience.ai:3141/pypi/simple/ -
|
|
| 87 |
### 2. Download the Model Package and Dataset
|
| 88 |
|
| 89 |
```bash
|
| 90 |
-
hf download
|
| 91 |
-
hf download
|
|
|
|
| 92 |
```
|
| 93 |
|
| 94 |
### Model Weights
|
|
@@ -106,14 +107,14 @@ python scripts/preflight.py --strict-weights --strict-imports
|
|
| 106 |
If the full dataset is already prepared:
|
| 107 |
|
| 108 |
```bash
|
| 109 |
-
export DATA_ROOT_DIR=../
|
| 110 |
python scripts/preflight.py --strict-weights --strict-imports --strict-data
|
| 111 |
```
|
| 112 |
|
| 113 |
### 4. Run Inference
|
| 114 |
|
| 115 |
```bash
|
| 116 |
-
export DATA_ROOT_DIR=../
|
| 117 |
bash scripts/inference_unified_demo.sh
|
| 118 |
```
|
| 119 |
|
|
@@ -128,14 +129,14 @@ output_unified/7r6r/seed_101/predictions/
|
|
| 128 |
Training:
|
| 129 |
|
| 130 |
```bash
|
| 131 |
-
export DATA_ROOT_DIR=../
|
| 132 |
bash scripts/train_demo.sh
|
| 133 |
```
|
| 134 |
|
| 135 |
Fine-tuning:
|
| 136 |
|
| 137 |
```bash
|
| 138 |
-
export DATA_ROOT_DIR=../
|
| 139 |
bash scripts/finetune_demo.sh
|
| 140 |
```
|
| 141 |
|
|
|
|
| 36 |
| Scenario | Description |
|
| 37 |
| :---: | :--- |
|
| 38 |
| Complex structure prediction | Takes a Protenix JSON file and locally generated MSA features as input and outputs predicted structures in CIF format and confidence scores in JSON format. |
|
| 39 |
+
| Full Hugging Face package validation | Uses the included `config/`, `models/`, `scripts/`, `examples/`, and `weights/` directories to run preflight checks and inference directly. |
|
| 40 |
| Fine-tuning pipeline validation | Starts the fine-tuning workflow using the included weights and `ft_datasets/finetune_subset.txt`. |
|
| 41 |
| Training pipeline setup | Supports single-GPU training once the complete Protenix dataset is provided. |
|
| 42 |
|
|
|
|
| 87 |
### 2. Download the Model Package and Dataset
|
| 88 |
|
| 89 |
```bash
|
| 90 |
+
hf download OneScience-Group/protenix --local-dir ./protenix
|
| 91 |
+
hf download OneScience-Group/protenix_dataset --repo-type dataset --local-dir ./protenix_dataset
|
| 92 |
+
cd protenix
|
| 93 |
```
|
| 94 |
|
| 95 |
### Model Weights
|
|
|
|
| 107 |
If the full dataset is already prepared:
|
| 108 |
|
| 109 |
```bash
|
| 110 |
+
export DATA_ROOT_DIR=../protenix_dataset
|
| 111 |
python scripts/preflight.py --strict-weights --strict-imports --strict-data
|
| 112 |
```
|
| 113 |
|
| 114 |
### 4. Run Inference
|
| 115 |
|
| 116 |
```bash
|
| 117 |
+
export DATA_ROOT_DIR=../protenix_dataset
|
| 118 |
bash scripts/inference_unified_demo.sh
|
| 119 |
```
|
| 120 |
|
|
|
|
| 129 |
Training:
|
| 130 |
|
| 131 |
```bash
|
| 132 |
+
export DATA_ROOT_DIR=../protenix_dataset
|
| 133 |
bash scripts/train_demo.sh
|
| 134 |
```
|
| 135 |
|
| 136 |
Fine-tuning:
|
| 137 |
|
| 138 |
```bash
|
| 139 |
+
export DATA_ROOT_DIR=../protenix_dataset
|
| 140 |
bash scripts/finetune_demo.sh
|
| 141 |
```
|
| 142 |
|