File size: 6,250 Bytes
7248c75 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 | #!/usr/bin/env python3
"""
Script to upload PhaseNet-TF model to Hugging Face Hub
"""
import os
import json
from pathlib import Path
from huggingface_hub import HfApi, create_repo, upload_file
from huggingface_hub import hf_hub_download
import torch
import yaml
def create_model_card():
"""Create a comprehensive model card for PhaseNet-TF"""
return """---
language: en
tags:
- seismic
- earthquake
- phase-picking
- deep-learning
- pytorch
license: mit
datasets:
- PS_Alaska
metrics:
- f1-score
- precision
- recall
---
# PhaseNet-TF: Advanced Seismic Arrival Time Detection
## Model Description
PhaseNet-TF is an advanced deep learning model for automatic seismic phase picking (P-wave, S-wave, and PS-wave detection) using spectrogram-based image segmentation approaches. The model leverages DeepLabV3Plus architecture to detect seismic arrivals with high accuracy, especially for weak and noisy signals from ocean-bottom seismometers and weak phases such as slab interface refracted PS and SP waves. This Alaska version is specifically trained on the PS_Alaska dataset for P and S phases. For more details, please refer to the paper and the [PhaseNet-TF](https://github.com/swei-seismo/PhaseNet-TF) repository.
## Model Architecture
- **Backbone**: DeepLabV3Plus with ResNet34 encoder
- **Input**: 3-component seismic waveforms converted to 6-channel spectrograms (real + imaginary)
- **Output**: Probability maps for P, S, PS phases and noise
- **Sampling Rate**: 40 Hz (dt_s = 0.025s)
- **Window Length**: 4800 points (120 seconds)
- **Spectrogram Size**: 64 × 4800 (frequency × time)
- **Input Channels**: 6 (3 real + 3 imaginary spectrogram channels)
- **Output Classes**: 4 (noise, P, S, PS)
## Load the checkpoint
checkpoint = torch.load("alaska_iter2.ckpt", map_location="cpu")
## Citation
If you use this model in your research, please cite:
```bibtex
@article{jie2025background,
title={Background Seismicity and Aftershocks of the 2020-2021 Large Earthquakes at the Alaska Peninsula Revealed by a Deep-learning-based Catalog},
author={Jie, Yaqi and Wei, Songqiao Shawn and Zhu, Weiqiang and Freymueller, Jeffrey Todd and Elliott, Julie},
journal={Authorea Preprints},
year={2025},
publisher={Authorea}
}
```
## License
This model is licensed under the MIT License.
"""
def create_config_json(model_path):
"""Create config.json with model metadata"""
config = {
"model_type": "phasenet-tf",
"architecture": "DeepLabV3Plus with ResNet34 encoder",
"input_channels": 6, # 3-component real + 3-component imaginary spectrograms
"output_classes": 4, # noise, P, S, PS
"sampling_rate": 40, # 1/0.025 = 40 Hz
"window_length": 4800, # 120 seconds at 40 Hz
"phases": ["P", "S", "PS"],
"framework": "pytorch",
"license": "mit",
"tags": ["seismic", "earthquake", "phase-picking", "deep-learning", "deeplabv3plus"]
}
return config
def upload_model_to_hf(
checkpoint_path: str,
config_path: str = None,
repo_name: str = "PhaseNet-TF_Alaska",
username: str = None,
token: str = None
):
"""Upload model to Hugging Face Hub"""
# Initialize API
if token:
api = HfApi(token=token)
else:
api = HfApi()
# Get username if not provided
if username is None:
try:
username = api.whoami()["name"]
print(f"Using logged-in username: {username}")
except Exception as e:
print(f"Error getting username: {e}")
print("Please provide username with --username or login with huggingface-cli login")
return
# Create repository
repo_id = f"{username}/{repo_name}"
try:
if token:
create_repo(repo_id, token=token, exist_ok=True)
else:
create_repo(repo_id, exist_ok=True)
print(f"Repository {repo_id} created/accessed successfully")
except Exception as e:
print(f"Error creating repository: {e}")
return
# Upload checkpoint
print("Uploading model checkpoint...")
upload_file(
path_or_fileobj=checkpoint_path,
path_in_repo="pytorch_model.bin",
repo_id=repo_id,
token=token
)
# Upload config
if config_path and os.path.exists(config_path):
print("Uploading config file...")
upload_file(
path_or_fileobj=config_path,
path_in_repo="config.yaml",
repo_id=repo_id,
token=token
)
# Create and upload config.json
config_json = create_config_json(checkpoint_path)
config_json_path = "config.json"
with open(config_json_path, 'w') as f:
json.dump(config_json, f, indent=2)
upload_file(
path_or_fileobj=config_json_path,
path_in_repo="config.json",
repo_id=repo_id,
token=token
)
# Create and upload README.md
model_card = create_model_card()
readme_path = "README.md"
with open(readme_path, 'w') as f:
f.write(model_card)
upload_file(
path_or_fileobj=readme_path,
path_in_repo="README.md",
repo_id=repo_id,
token=token
)
# Clean up temporary files
os.remove(config_json_path)
os.remove(readme_path)
print(f"Model uploaded successfully to https://huggingface.co/{repo_id}")
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser(description="Upload PhaseNet-TF model to Hugging Face")
parser.add_argument("--checkpoint", required=True, help="Path to model checkpoint (.ckpt)")
parser.add_argument("--config", help="Path to config file (.yaml)")
parser.add_argument("--repo-name", default="PhaseNet-TF_Alaska", help="Repository name")
parser.add_argument("--username", help="Hugging Face username (optional if already logged in)")
parser.add_argument("--token", help="Hugging Face token (optional if already logged in)")
args = parser.parse_args()
upload_model_to_hf(
checkpoint_path=args.checkpoint,
config_path=args.config,
repo_name=args.repo_name,
username=args.username,
token=args.token
) |