diff --git a/models/Dockerfile b/Dockerfile similarity index 100% rename from models/Dockerfile rename to Dockerfile diff --git a/models/DockerfileCUDA b/DockerfileCUDA similarity index 100% rename from models/DockerfileCUDA rename to DockerfileCUDA diff --git a/README.md b/README.md index b97c0796d82b04cd4a07aa5cac192236f7e75700..50fc8695712829ff9a9a1430597644f5b79c374d 100644 --- a/README.md +++ b/README.md @@ -1 +1,1445 @@ -# master_thesis_code \ No newline at end of file +
+ +# Lightning-Hydra-Template + +Python +PyTorch +Lightning +Config: hydra +Code style: black + +A clean and scalable template to kickstart your deep learning project πŸš€βš‘πŸ”₯
+Click on [Use this template](https://github.com/ashleve/lightning-hydra-template/generate) to initialize new repository. + +_Suggestions are always welcome!_ + +
+ +

+ +## πŸ“Œ  Introduction + +This template tries to be as general as possible. It integrates many different MLOps tools. + +> Effective usage of this template requires learning of a couple of technologies: [PyTorch](https://pytorch.org), [PyTorch Lightning](https://www.pytorchlightning.ai) and [Hydra](https://hydra.cc). Knowledge of some experiment logging framework like [Weights&Biases](https://wandb.com), [Neptune](https://neptune.ai) or [MLFlow](https://mlflow.org) is also recommended. + +**Why you should use it:** it allows you to rapidly iterate over new models/datasets and scale your projects from small single experiments to hyperparameter searches on computing clusters, without writing any boilerplate code. To my knowledge, it's one of the most convenient all-in-one technology stack for Deep Learning research. Good starting point for reproducing papers, kaggle competitions or small-team research projects. It's also a collection of best practices for efficient workflow and reproducibility. + +**Why you shouldn't use it:** this template is not fitted to be a production environment, should be used more as a fast experimentation tool. Apart from that, Lightning and Hydra are still evolving and integrate many libraries, which means sometimes things break - for the list of currently known bugs, visit [this page](https://github.com/ashleve/lightning-hydra-template/labels/bug). Also, even though Lightning is very flexible, it's not well suited for every possible deep learning task. See [#Limitations](#limitations) for more. + +### Why PyTorch Lightning? + +[PyTorch Lightning](https://github.com/PyTorchLightning/pytorch-lightning) is a lightweight PyTorch wrapper for high-performance AI research. +It makes your code neatly organized and provides lots of useful features, like ability to run model on CPU, GPU, multi-GPU cluster and TPU. + +### Why Hydra? + +[Hydra](https://github.com/facebookresearch/hydra) is an open-source Python framework that simplifies the development of research and other complex applications. The key feature is the ability to dynamically create a hierarchical configuration by composition and override it through config files and the command line. It allows you to conveniently manage experiments and provides many useful plugins, like [Optuna Sweeper](https://hydra.cc/docs/next/plugins/optuna_sweeper) for hyperparameter search, or [Ray Launcher](https://hydra.cc/docs/next/plugins/ray_launcher) for running jobs on a cluster. + +
+ +## Main Ideas Of This Template + +- **Predefined Structure**: clean and scalable so that work can easily be extended and replicated | [#Project Structure](#project-structure) +- **Rapid Experimentation**: thanks to automating pipeline with config files and hydra command line superpowers | [#Your Superpowers](#your-superpowers) +- **Reproducibility**: obtaining similar results is supported in multiple ways | [#Reproducibility](#reproducibility) +- **Little Boilerplate**: so pipeline can be easily modified | [#How It Works](#how-it-works) +- **Main Configuration**: main config file specifies default training configuration | [#Main Project Configuration](#main-project-configuration) +- **Experiment Configurations**: can be composed out of smaller configs and override chosen hyperparameters | [#Experiment Configuration](#experiment-configuration) +- **Workflow**: comes down to 4 simple steps | [#Workflow](#workflow) +- **Experiment Tracking**: many logging frameworks can be easily integrated, like Tensorboard, MLFlow or W&B | [#Experiment Tracking](#experiment-tracking) +- **Logs**: all logs (checkpoints, data from loggers, hparams, etc.) are stored in a convenient folder structure imposed by Hydra | [#Logs](#logs) +- **Hyperparameter Search**: made easier with Hydra built-in plugins like [Optuna Sweeper](https://hydra.cc/docs/next/plugins/optuna_sweeper) | [#Hyperparameter Search](#hyperparameter-search) +- **Tests**: unit tests and shell/command based tests for speeding up the development | [#Tests](#tests) +- **Best Practices**: a couple of recommended tools, practices and standards for efficient workflow and reproducibility | [#Best Practices](#best-practices) + +
+ +## Project Structure + +The directory structure of new project looks like this: + +``` +β”œβ”€β”€ configs <- Hydra configuration files +β”‚ β”œβ”€β”€ callbacks <- Callbacks configs +β”‚ β”œβ”€β”€ datamodule <- Datamodule configs +β”‚ β”œβ”€β”€ debug <- Debugging configs +β”‚ β”œβ”€β”€ experiment <- Experiment configs +β”‚ β”œβ”€β”€ hparams_search <- Hyperparameter search configs +β”‚ β”œβ”€β”€ local <- Local configs +β”‚ β”œβ”€β”€ log_dir <- Logging directory configs +β”‚ β”œβ”€β”€ logger <- Logger configs +β”‚ β”œβ”€β”€ model <- Model configs +β”‚ β”œβ”€β”€ trainer <- Trainer configs +β”‚ β”‚ +β”‚ β”œβ”€β”€ test.yaml <- Main config for testing +β”‚ └── train.yaml <- Main config for training +β”‚ +β”œβ”€β”€ data <- Project data +β”‚ +β”œβ”€β”€ logs <- Logs generated by Hydra and PyTorch Lightning loggers +β”‚ +β”œβ”€β”€ notebooks <- Jupyter notebooks. Naming convention is a number (for ordering), +β”‚ the creator's initials, and a short `-` delimited description, +β”‚ e.g. `1.0-jqp-initial-data-exploration.ipynb`. +β”‚ +β”œβ”€β”€ scripts <- Shell scripts +β”‚ +β”œβ”€β”€ src <- Source code +β”‚ β”œβ”€β”€ datamodules <- Lightning datamodules +β”‚ β”œβ”€β”€ models <- Lightning models +β”‚ β”œβ”€β”€ utils <- Utility scripts +β”‚ β”œβ”€β”€ vendor <- Third party code that cannot be installed using PIP/Conda +β”‚ β”‚ +β”‚ β”œβ”€β”€ testing_pipeline.py +β”‚ └── training_pipeline.py +β”‚ +β”œβ”€β”€ tests <- Tests of any kind +β”‚ β”œβ”€β”€ helpers <- A couple of testing utilities +β”‚ β”œβ”€β”€ shell <- Shell/command based tests +β”‚ └── unit <- Unit tests +β”‚ +β”œβ”€β”€ test.py <- Run testing +β”œβ”€β”€ train.py <- Run training +β”‚ +β”œβ”€β”€ .env.example <- Template of the file for storing private environment variables +β”œβ”€β”€ .gitignore <- List of files/folders ignored by git +β”œβ”€β”€ .pre-commit-config.yaml <- Configuration of pre-commit hooks for code formatting +β”œβ”€β”€ requirements.txt <- File for installing python dependencies +β”œβ”€β”€ setup.cfg <- Configuration of linters and pytest +└── README.md +``` + +
+ +## πŸš€  Quickstart + +```bash +# clone project +git clone https://github.com/ashleve/lightning-hydra-template +cd lightning-hydra-template + +# [OPTIONAL] create conda environment +conda create -n myenv python=3.8 +conda activate myenv + +# install pytorch according to instructions +# https://pytorch.org/get-started/ + +# install requirements +pip install -r requirements.txt +``` + +Template contains example with MNIST classification.
+When running `python train.py` you should see something like this: + +
+ +![](https://github.com/ashleve/lightning-hydra-template/blob/resources/terminal.png) + +
+ +### ⚑  Your Superpowers + +
+Override any config parameter from command line + +> Hydra allows you to easily overwrite any parameter defined in your config. + +```bash +python train.py trainer.max_epochs=20 model.lr=1e-4 +``` + +> You can also add new parameters with `+` sign. + +```bash +python train.py +model.new_param="uwu" +``` + +
+ +
+Train on CPU, GPU, multi-GPU and TPU + +> PyTorch Lightning makes it easy to train your models on different hardware. + +```bash +# train on CPU +python train.py trainer.gpus=0 + +# train on 1 GPU +python train.py trainer.gpus=1 + +# train on TPU +python train.py +trainer.tpu_cores=8 + +# train with DDP (Distributed Data Parallel) (4 GPUs) +python train.py trainer.gpus=4 +trainer.strategy=ddp + +# train with DDP (Distributed Data Parallel) (8 GPUs, 2 nodes) +python train.py trainer.gpus=4 +trainer.num_nodes=2 +trainer.strategy=ddp +``` + +
+ +
+Train with mixed precision + +```bash +# train with pytorch native automatic mixed precision (AMP) +python train.py trainer.gpus=1 +trainer.precision=16 +``` + +
+ + + +
+Train model with any logger available in PyTorch Lightning, like Weights&Biases or Tensorboard + +> PyTorch Lightning provides convenient integrations with most popular logging frameworks, like Tensorboard, Neptune or simple csv files. Read more [here](#experiment-tracking). Using wandb requires you to [setup account](https://www.wandb.com/) first. After that just complete the config as below.
> **Click [here](https://wandb.ai/hobglob/template-dashboard/) to see example wandb dashboard generated with this template.** + +```bash +# set project and entity names in `configs/logger/wandb` +wandb: + project: "your_project_name" + entity: "your_wandb_team_name" +``` + +```bash +# train model with Weights&Biases (link to wandb dashboard should appear in the terminal) +python train.py logger=wandb +``` + +
+ +
+Train model with chosen experiment config + +> Experiment configurations are placed in [configs/experiment/](configs/experiment/). + +```bash +python train.py experiment=example +``` + +
+ +
+Attach some callbacks to run + +> Callbacks can be used for things such as as model checkpointing, early stopping and [many more](https://pytorch-lightning.readthedocs.io/en/latest/extensions/callbacks.html#built-in-callbacks).
+> Callbacks configurations are placed in [configs/callbacks/](configs/callbacks/). + +```bash +python train.py callbacks=default +``` + +
+ +
+Use different tricks available in Pytorch Lightning + +> PyTorch Lightning provides about [40+ useful trainer flags](https://pytorch-lightning.readthedocs.io/en/latest/common/trainer.html#trainer-flags). + +```yaml +# gradient clipping may be enabled to avoid exploding gradients +python train.py +trainer.gradient_clip_val=0.5 + +# stochastic weight averaging can make your models generalize better +python train.py +trainer.stochastic_weight_avg=true + +# run validation loop 4 times during a training epoch +python train.py +trainer.val_check_interval=0.25 + +# accumulate gradients +python train.py +trainer.accumulate_grad_batches=10 + +# terminate training after 12 hours +python train.py +trainer.max_time="00:12:00:00" +``` + +
+ +
+Easily debug + +> Visit [configs/debug/](configs/debug/) for different debugging configs. + +```bash +# runs 1 epoch in default debugging mode +# changes logging directory to `logs/debugs/...` +# sets level of all command line loggers to 'DEBUG' +# enables extra trainer flags like tracking gradient norm +# enforces debug-friendly configuration +python train.py debug=default + +# runs test epoch without training +python train.py debug=test_only + +# run 1 train, val and test loop, using only 1 batch +python train.py +trainer.fast_dev_run=true + +# raise exception if there are any numerical anomalies in tensors, like NaN or +/-inf +python train.py +trainer.detect_anomaly=true + +# print execution time profiling after training ends +python train.py +trainer.profiler="simple" + +# try overfitting to 1 batch +python train.py +trainer.overfit_batches=1 trainer.max_epochs=20 + +# use only 20% of the data +python train.py +trainer.limit_train_batches=0.2 \ ++trainer.limit_val_batches=0.2 +trainer.limit_test_batches=0.2 + +# log second gradient norm of the model +python train.py +trainer.track_grad_norm=2 +``` + +
+ +
+Resume training from checkpoint + +> Checkpoint can be either path or URL. + +```yaml +python train.py trainer.resume_from_checkpoint="/path/to/ckpt/name.ckpt" +``` + +> ⚠️ Currently loading ckpt in Lightning doesn't resume logger experiment, but it will be supported in future Lightning release. + +
+ +
+Execute evaluation for a given checkpoint + +> Checkpoint can be either path or URL. + +```yaml +python test.py ckpt_path="/path/to/ckpt/name.ckpt" +``` + +
+ +
+Create a sweep over hyperparameters + +```bash +# this will run 6 experiments one after the other, +# each with different combination of batch_size and learning rate +python train.py -m datamodule.batch_size=32,64,128 model.lr=0.001,0.0005 +``` + +> ⚠️ This sweep is not failure resistant (if one job crashes than the whole sweep crashes). + +
+ +
+Create a sweep over hyperparameters with Optuna + +> Using [Optuna Sweeper](https://hydra.cc/docs/next/plugins/optuna_sweeper) plugin doesn't require you to code any boilerplate into your pipeline, everything is defined in a [single config file](configs/hparams_search/mnist_optuna.yaml)! + +```bash +# this will run hyperparameter search defined in `configs/hparams_search/mnist_optuna.yaml` +# over chosen experiment config +python train.py -m hparams_search=mnist_optuna experiment=example_simple +``` + +> ⚠️ Currently this sweep is not failure resistant (if one job crashes than the whole sweep crashes). Might be supported in future Hydra release. + +
+ +
+Execute all experiments from folder + +> Hydra provides special syntax for controlling behavior of multiruns. Learn more [here](https://hydra.cc/docs/next/tutorials/basic/running_your_app/multi-run). The command below executes all experiments from folder [configs/experiment/](configs/experiment/). + +```bash +python train.py -m 'experiment=glob(*)' +``` + +
+ +
+Execute sweep on a remote AWS cluster + +> This should be achievable with simple config using [Ray AWS launcher for Hydra](https://hydra.cc/docs/next/plugins/ray_launcher). Example is not yet implemented in this template. + +
+ + + +
+Use Hydra tab completion + +> Hydra allows you to autocomplete config argument overrides in shell as you write them, by pressing `tab` key. Learn more [here](https://hydra.cc/docs/tutorials/basic/running_your_app/tab_completion). + +
+ +
+Apply pre-commit hooks + +> Apply pre-commit hooks to automatically format your code and configs, perform code analysis and remove output from jupyter notebooks. See [# Best Practices](#best-practices) for more. + +```bash +pre-commit run -a +``` + +
+ +
+ +## ❀️  Contributions + +Have a question? Found a bug? Missing a specific feature? Have an idea for improving documentation? Feel free to file a new issue, discussion or PR with respective title and description. If you already found a solution to your problem, don't hesitate to share it. Suggestions for new best practices are always welcome! + +
+ +## ℹ️  Guide + +### How To Get Started + +- First, you should probably get familiar with [PyTorch Lightning](https://www.pytorchlightning.ai) +- Next, go through [Hydra quick start guide](https://hydra.cc/docs/intro/) and [basic Hydra tutorial](https://hydra.cc/docs/tutorials/basic/your_first_app/simple_cli/) + +
+ +### How It Works + +All PyTorch Lightning modules are dynamically instantiated from module paths specified in config. Example model config: + +```yaml +_target_: src.models.mnist_model.MNISTLitModule +input_size: 784 +lin1_size: 256 +lin2_size: 256 +lin3_size: 256 +output_size: 10 +lr: 0.001 +``` + +Using this config we can instantiate the object with the following line: + +```python +model = hydra.utils.instantiate(config.model) +``` + +This allows you to easily iterate over new models! Every time you create a new one, just specify its module path and parameters in appropriate config file.
+ +Switch between models and datamodules with command line arguments: + +```bash +python train.py model=mnist +``` + +The whole pipeline managing the instantiation logic is placed in [src/training_pipeline.py](src/training_pipeline.py). + +
+ +### Main Project Configuration + +Location: [configs/train.yaml](configs/train.yaml)
+Main project config contains default training configuration.
+It determines how config is composed when simply executing command `python train.py`.
+ +
+Show main project config + +```yaml +# specify here default training configuration +defaults: + - _self_ + - datamodule: mnist.yaml + - model: mnist.yaml + - callbacks: default.yaml + - logger: null # set logger here or use command line (e.g. `python train.py logger=tensorboard`) + - trainer: default.yaml + - log_dir: default.yaml + + # experiment configs allow for version control of specific configurations + # e.g. best hyperparameters for each combination of model and datamodule + - experiment: null + + # debugging config (enable through command line, e.g. `python train.py debug=default) + - debug: null + + # config for hyperparameter optimization + - hparams_search: null + + # optional local config for machine/user specific settings + # it's optional since it doesn't need to exist and is excluded from version control + - optional local: default.yaml + + # enable color logging + - override hydra/hydra_logging: colorlog + - override hydra/job_logging: colorlog + +# path to original working directory +# hydra hijacks working directory by changing it to the new log directory +# https://hydra.cc/docs/next/tutorials/basic/running_your_app/working_directory +original_work_dir: ${hydra:runtime.cwd} + +# path to folder with data +data_dir: ${original_work_dir}/data/ + +# pretty print config at the start of the run using Rich library +print_config: True + +# disable python warnings if they annoy you +ignore_warnings: True + +# set False to skip model training +train: True + +# evaluate on test set, using best model weights achieved during training +# lightning chooses best weights based on the metric specified in checkpoint callback +test: True + +# seed for random number generators in pytorch, numpy and python.random +seed: null + +# default name for the experiment, determines logging folder path +# (you can overwrite this name in experiment configs) +name: "default" +``` + +
+ +
+ +### Experiment Configuration + +Location: [configs/experiment](configs/experiment)
+Experiment configs allow you to overwrite parameters from main project configuration.
+For example, you can use them to version control best hyperparameters for each combination of model and dataset. + +
+Show example experiment config + +```yaml +# to execute this experiment run: +# python train.py experiment=example + +defaults: + - override /datamodule: mnist.yaml + - override /model: mnist.yaml + - override /callbacks: default.yaml + - override /logger: null + - override /trainer: default.yaml + +# all parameters below will be merged with parameters from default configurations set above +# this allows you to overwrite only specified parameters + +# name of the run determines folder name in logs +name: "simple_dense_net" + +seed: 12345 + +trainer: + min_epochs: 10 + max_epochs: 10 + gradient_clip_val: 0.5 + +model: + lin1_size: 128 + lin2_size: 256 + lin3_size: 64 + lr: 0.002 + +datamodule: + batch_size: 64 + +logger: + wandb: + tags: ["mnist", "${name}"] +``` + +
+ +
+ +### Local Configuration + +Location: [configs/local](configs/local)
+Some configurations are user/machine/installation specific (e.g. configuration of local cluster, or harddrive paths on a specific machine). For such scenarios, a file `configs/local/default.yaml` can be created which is automatically loaded but not tracked by Git. + +
+Show example local Slurm cluster config + +```yaml +# @package _global_ + +defaults: + - override /hydra/launcher@_here_: submitit_slurm + +data_dir: /mnt/scratch/data/ + +hydra: + launcher: + timeout_min: 1440 + gpus_per_task: 1 + gres: gpu:1 + job: + env_set: + MY_VAR: /home/user/my/system/path + MY_KEY: asdgjhawi8y23ihsghsueity23ihwd +``` + +
+ +
+ +### Workflow + +1. Write your PyTorch Lightning module (see [models/mnist_module.py](src/models/mnist_module.py) for example) +2. Write your PyTorch Lightning datamodule (see [datamodules/mnist_datamodule.py](src/datamodules/mnist_datamodule.py) for example) +3. Write your experiment config, containing paths to your model and datamodule +4. Run training with chosen experiment config: `python train.py experiment=experiment_name` + +
+ +### Logs + +**Hydra creates new working directory for every executed run.** By default, logs have the following structure: + +``` +β”œβ”€β”€ logs +β”‚ β”œβ”€β”€ experiments # Folder for the logs generated by experiments +β”‚ β”‚ β”œβ”€β”€ runs # Folder for single runs +β”‚ β”‚ β”‚ β”œβ”€β”€ experiment_name # Experiment name +β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ YYYY-MM-DD_HH-MM-SS # Datetime of the run +β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ .hydra # Hydra logs +β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ csv # Csv logs +β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ wandb # Weights&Biases logs +β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ checkpoints # Training checkpoints +β”‚ β”‚ β”‚ β”‚ β”‚ └── ... # Any other thing saved during training +β”‚ β”‚ β”‚ β”‚ └── ... +β”‚ β”‚ β”‚ └── ... +β”‚ β”‚ β”‚ +β”‚ β”‚ └── multiruns # Folder for multiruns +β”‚ β”‚ β”œβ”€β”€ experiment_name # Experiment name +β”‚ β”‚ β”‚ β”œβ”€β”€ YYYY-MM-DD_HH-MM-SS # Datetime of the multirun +β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€1 # Multirun job number +β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€2 +β”‚ β”‚ β”‚ β”‚ └── ... +β”‚ β”‚ β”‚ └── ... +β”‚ β”‚ └── ... +β”‚ β”‚ +β”‚ β”œβ”€β”€ evaluations # Folder for the logs generated during testing +β”‚ β”‚ └── ... +β”‚ β”‚ +β”‚ └── debugs # Folder for the logs generated during debugging +β”‚ └── ... +``` + +You can change this structure by modifying paths in [hydra configuration](configs/log_dir). + +
+ +### Experiment Tracking + +PyTorch Lightning supports many popular logging frameworks:
+**[Weights&Biases](https://www.wandb.com/) Β· [Neptune](https://neptune.ai/) Β· [Comet](https://www.comet.ml/) Β· [MLFlow](https://mlflow.org) Β· [Tensorboard](https://www.tensorflow.org/tensorboard/)** + +These tools help you keep track of hyperparameters and output metrics and allow you to compare and visualize results. To use one of them simply complete its configuration in [configs/logger](configs/logger) and run: + +```bash +python train.py logger=logger_name +``` + +You can use many of them at once (see [configs/logger/many_loggers.yaml](configs/logger/many_loggers.yaml) for example). + +You can also write your own logger. + +Lightning provides convenient method for logging custom metrics from inside LightningModule. Read the docs [here](https://pytorch-lightning.readthedocs.io/en/latest/extensions/logging.html#automatic-logging) or take a look at [MNIST example](src/models/mnist_module.py). + +
+ +### Hyperparameter Search + +Defining hyperparameter optimization is as easy as adding new config file to [configs/hparams_search](configs/hparams_search). + +
+Show example + +```yaml +defaults: + - override /hydra/sweeper: optuna + +# choose metric which will be optimized by Optuna +optimized_metric: "val/acc_best" + +hydra: + # here we define Optuna hyperparameter search + # it optimizes for value returned from function with @hydra.main decorator + # learn more here: https://hydra.cc/docs/next/plugins/optuna_sweeper + sweeper: + _target_: hydra_plugins.hydra_optuna_sweeper.optuna_sweeper.OptunaSweeper + storage: null + study_name: null + n_jobs: 1 + + # 'minimize' or 'maximize' the objective + direction: maximize + + # number of experiments that will be executed + n_trials: 20 + + # choose Optuna hyperparameter sampler + # learn more here: https://optuna.readthedocs.io/en/stable/reference/samplers.html + sampler: + _target_: optuna.samplers.TPESampler + seed: 12345 + consider_prior: true + prior_weight: 1.0 + consider_magic_clip: true + consider_endpoints: false + n_startup_trials: 10 + n_ei_candidates: 24 + multivariate: false + warn_independent_sampling: true + + # define range of hyperparameters + search_space: + datamodule.batch_size: + type: categorical + choices: [32, 64, 128] + model.lr: + type: float + low: 0.0001 + high: 0.2 + model.lin1_size: + type: categorical + choices: [32, 64, 128, 256, 512] + model.lin2_size: + type: categorical + choices: [32, 64, 128, 256, 512] + model.lin3_size: + type: categorical + choices: [32, 64, 128, 256, 512] +``` + +
+ +Next, you can execute it with: `python train.py -m hparams_search=mnist_optuna` + +Using this approach doesn't require you to add any boilerplate into your pipeline, everything is defined in a single config file. + +You can use different optimization frameworks integrated with Hydra, like Optuna, Ax or Nevergrad. + +The `optimization_results.yaml` will be available under `logs/multirun` folder. + +This approach doesn't support advanced technics like prunning - for more sophisticated search, you probably shouldn't use hydra multirun feature and instead write your own optimization pipeline. + +
+ +### Inference + +The following code is an example of loading model from checkpoint and running predictions.
+ +
+Show example + +```python +from PIL import Image +from torchvision import transforms + +from src.models.mnist_module import MNISTLitModule + + +def predict(): + """Example of inference with trained model. + It loads trained image classification model from checkpoint. + Then it loads example image and predicts its label. + """ + + # ckpt can be also a URL! + CKPT_PATH = "last.ckpt" + + # load model from checkpoint + # model __init__ parameters will be loaded from ckpt automatically + # you can also pass some parameter explicitly to override it + trained_model = MNISTLitModule.load_from_checkpoint(checkpoint_path=CKPT_PATH) + + # print model hyperparameters + print(trained_model.hparams) + + # switch to evaluation mode + trained_model.eval() + trained_model.freeze() + + # load data + img = Image.open("data/example_img.png").convert("L") # convert to black and white + # img = Image.open("data/example_img.png").convert("RGB") # convert to RGB + + # preprocess + mnist_transforms = transforms.Compose( + [ + transforms.ToTensor(), + transforms.Resize((28, 28)), + transforms.Normalize((0.1307,), (0.3081,)), + ] + ) + img = mnist_transforms(img) + img = img.reshape((1, *img.size())) # reshape to form batch of size 1 + + # inference + output = trained_model(img) + print(output) + + +if __name__ == "__main__": + predict() + +``` + +
+ +
+ +### Tests + +Template comes with example tests implemented with pytest library. To execute them simply run: + +```bash +# run all tests +pytest + +# run tests from specific file +pytest tests/shell/test_basic_commands.py + +# run all tests except the ones marked as slow +pytest -k "not slow" +``` + +To speed up the development, you can once in a while execute tests that run a couple of quick experiments, like training 1 epoch on 25% of data, executing single train/val/test step, etc. Those kind of tests don't check for any specific output - they exist to simply verify that executing some bash commands doesn't end up in throwing exceptions. You can find them implemented in [tests/shell](tests/shell) folder. + +You can easily modify the commands in the scripts for your use case. If 1 epoch is too much for your model, then make it run for a couple of batches instead (by using the right trainer flags). + +
+ +### Callbacks + +The branch [`wandb-callbacks`](https://github.com/ashleve/lightning-hydra-template/tree/wandb-callbacks) contains example callbacks enabling better Weights&Biases integration, which you can use as a reference for writing your own callbacks (see [wandb_callbacks.py](https://github.com/ashleve/lightning-hydra-template/tree/wandb-callbacks/src/callbacks/wandb_callbacks.py)). + +Callbacks which support reproducibility: + +- **WatchModel** +- **UploadCodeAsArtifact** +- **UploadCheckpointsAsArtifact** + +Callbacks which provide examples of logging custom visualisations: + +- **LogConfusionMatrix** +- **LogF1PrecRecHeatmap** +- **LogImagePredictions** + +To try all of the callbacks at once, switch to the right branch: + +```bash +git checkout wandb-callbacks +``` + +And then run the following command: + +```bash +python train.py logger=wandb callbacks=wandb +``` + +To see the result of all the callbacks attached, take a look at [this experiment dashboard](https://wandb.ai/hobglob/template-tests/runs/3rw7q70h). + +
+ +### Multi-GPU Training + +Lightning supports multiple ways of doing distributed training. The most common one is DDP, which spawns separate process for each GPU and averages gradients between them. To learn about other approaches read the [lightning docs](https://pytorch-lightning.readthedocs.io/en/latest/advanced/multi_gpu.html). + +You can run DDP on mnist example with 4 GPUs like this: + +```bash +python train.py trainer.gpus=4 +trainer.strategy=ddp +``` + +⚠️ When using DDP you have to be careful how you write your models - learn more [here](https://pytorch-lightning.readthedocs.io/en/latest/advanced/multi_gpu.html). + +
+ +### Docker + +First you will need to [install Nvidia Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html) to enable GPU support. + +The template Dockerfile is provided on branch [`dockerfiles`](https://github.com/ashleve/lightning-hydra-template/tree/dockerfiles). Copy it to the template root folder. + +To build the container use: + +```bash +docker build -t . +``` + +To mount the project to the container use: + +```bash +docker run -v $(pwd):/workspace/project --gpus all -it --rm +``` + +
+ +### Reproducibility + +What provides reproducibility: + +- Hydra manages your configs +- Hydra manages your logging paths and makes every executed run store its hyperparameters and config overrides in a separate file in logs +- Single seed for random number generators in pytorch, numpy and python.random +- LightningDataModule allows you to encapsulate data split, transformations and default parameters in a single, clean abstraction +- LightningModule separates your research code from engineering code in a clean way +- Experiment tracking frameworks take care of logging metrics and hparams, some can also store results and artifacts in cloud +- Pytorch Lightning takes care of creating training checkpoints +- Example callbacks for wandb show how you can save and upload a snapshot of codebase every time the run is executed, as well as upload ckpts and track model gradients + + +
+ +### Limitations + +- Currently, template doesn't support k-fold cross validation, but it's possible to achieve it with Lightning Loop interface. See the [official example](https://github.com/PyTorchLightning/pytorch-lightning/blob/master/pl_examples/loop_examples/kfold.py). Implementing it requires rewriting the training pipeline. +- Pytorch Lightning might not be the best choice for scalable reinforcement learning, it's probably better to use something like [Ray](https://github.com/ray-project/ray). +- Currently hyperparameter search with Hydra Optuna Plugin doesn't support prunning. +- Hydra changes working directory to new logging folder for every executed run, which might not be compatible with the way some libraries work. + +
+ +## Useful Tricks + +
+Accessing datamodule attributes in model + +1. The simplest way is to pass datamodule attribute directly to model on initialization: + + ```python + # ./src/training_pipeline.py + datamodule = hydra.utils.instantiate(config.datamodule) + model = hydra.utils.instantiate(config.model, some_param=datamodule.some_param) + ``` + + This is not a very robust solution, since it assumes all your datamodules have `some_param` attribute available (otherwise the run will crash). + +2. If you only want to access datamodule config, you can simply pass it as an init parameter: + + ```python + # ./src/training_pipeline.py + model = hydra.utils.instantiate(config.model, dm_conf=config.datamodule, _recursive_=False) + ``` + + Now you can access any datamodule config part like this: + + ```python + # ./src/models/my_model.py + class MyLitModel(LightningModule): + def __init__(self, dm_conf, param1, param2): + super().__init__() + + batch_size = dm_conf.batch_size + ``` + +3. If you need to access the datamodule object attributes, a little hacky solution is to add Omegaconf resolver to your datamodule: + + ```python + # ./src/datamodules/my_datamodule.py + from omegaconf import OmegaConf + + class MyDataModule(LightningDataModule): + def __init__(self, param1, param2): + super().__init__() + + self.param1 = param1 + + resolver_name = "datamodule" + OmegaConf.register_new_resolver( + resolver_name, + lambda name: getattr(self, name), + use_cache=False + ) + ``` + + This way you can reference any datamodule attribute from your config like this: + + ```yaml + # this will return attribute 'param1' from datamodule object + param1: ${datamodule: param1} + ``` + + When later accessing this field, say in your lightning model, it will get automatically resolved based on all resolvers that are registered. Remember not to access this field before datamodule is initialized or it will crash. **You also need to set `resolve=False` in `print_config()` in [train.py](train.py) or it will throw errors:** + + ```python + # ./src/train.py + utils.print_config(config, resolve=False) + ``` + +
+ +
+Automatic activation of virtual environment and tab completion when entering folder + +1. Create a new file called `.autoenv` (this name is excluded from version control in `.gitignore`).
+ You can use it to automatically execute shell commands when entering folder. Add some commands to your `.autoenv` file, like in the example below: + + ```bash + # activate conda environment + conda activate myenv + + # activate hydra tab completion for bash + eval "$(python train.py -sc install=bash)" + ``` + + (these commands will be executed whenever you're openning or switching terminal to folder containing `.autoenv` file) + +2. To setup this automation for bash, execute the following line (it will append your `.bashrc` file): + + ```bash + echo "autoenv() { if [ -x .autoenv ]; then source .autoenv ; echo '.autoenv executed' ; fi } ; cd() { builtin cd \"\$@\" ; autoenv ; } ; autoenv" >> ~/.bashrc + ``` + +3. Lastly add execution previliges to your `.autoenv` file: + + ``` + chmod +x .autoenv + ``` + + (for safety, only `.autoenv` with previligies will be executed) + +**Explanation** + +The mentioned line appends your `.bashrc` file with 2 commands: + +1. `autoenv() { if [ -x .autoenv ]; then source .autoenv ; echo '.autoenv executed' ; fi }` - this declares the `autoenv()` function, which executes `.autoenv` file if it exists in current work dir and has execution previligies +2. `cd() { builtin cd \"\$@\" ; autoenv ; } ; autoenv` - this extends behaviour of `cd` command, to make it execute `autoenv()` function each time you change folder in terminal or open new terminal + +
+ + + +
+ +## Best Practices + +
+Use Miniconda for GPU environments + +Use miniconda for your python environments (it's usually unnecessary to install full anaconda environment, miniconda should be enough). +It makes it easier to install some dependencies, like cudatoolkit for GPU support. It also allows you to acccess your environments globally. + +Example installation: + +```bash +wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh +bash Miniconda3-latest-Linux-x86_64.sh +``` + +Create new conda environment: + +```bash +conda create -n myenv python=3.8 +conda activate myenv +``` + +
+ +
+Use automatic code formatting + +Use pre-commit hooks to standardize code formatting of your project and save mental energy.
+Simply install pre-commit package with: + +```bash +pip install pre-commit +``` + +Next, install hooks from [.pre-commit-config.yaml](.pre-commit-config.yaml): + +```bash +pre-commit install +``` + +After that your code will be automatically reformatted on every new commit.
+Currently template contains configurations of **black** (python code formatting), **isort** (python import sorting), **flake8** (python code analysis), **prettier** (yaml formating) and **nbstripout** (clearing output from jupyter notebooks).
+ +To reformat all files in the project use command: + +```bash +pre-commit run -a +``` + +
+ +
+Set private environment variables in .env file + +System specific variables (e.g. absolute paths to datasets) should not be under version control or it will result in conflict between different users. Your private keys also shouldn't be versioned since you don't want them to be leaked.
+ +Template contains `.env.example` file, which serves as an example. Create a new file called `.env` (this name is excluded from version control in .gitignore). +You should use it for storing environment variables like this: + +``` +MY_VAR=/home/user/my_system_path +``` + +All variables from `.env` are loaded in `train.py` automatically. + +Hydra allows you to reference any env variable in `.yaml` configs like this: + +```yaml +path_to_data: ${oc.env:MY_VAR} +``` + +
+ +
+Name metrics using '/' character + +Depending on which logger you're using, it's often useful to define metric name with `/` character: + +```python +self.log("train/loss", loss) +``` + +This way loggers will treat your metrics as belonging to different sections, which helps to get them organised in UI. + +
+ +
+Use torchmetrics + +Use official [torchmetrics](https://github.com/PytorchLightning/metrics) library to ensure proper calculation of metrics. This is especially important for multi-GPU training! + +For example, instead of calculating accuracy by yourself, you should use the provided `Accuracy` class like this: + +```python +from torchmetrics.classification.accuracy import Accuracy + + +class LitModel(LightningModule): + def __init__(self) + self.train_acc = Accuracy() + self.val_acc = Accuracy() + + def training_step(self, batch, batch_idx): + ... + acc = self.train_acc(predictions, targets) + self.log("train/acc", acc) + ... + + def validation_step(self, batch, batch_idx): + ... + acc = self.val_acc(predictions, targets) + self.log("val/acc", acc) + ... +``` + +Make sure to use different metric instance for each step to ensure proper value reduction over all GPU processes. + +Torchmetrics provides metrics for most use cases, like F1 score or confusion matrix. Read [documentation](https://torchmetrics.readthedocs.io/en/latest/#more-reading) for more. + +
+ +
+Follow PyTorch Lightning style guide + +The style guide is available [here](https://pytorch-lightning.readthedocs.io/en/latest/starter/style_guide.html).
+ +1. Be explicit in your init. Try to define all the relevant defaults so that the user doesn’t have to guess. Provide type hints. This way your module is reusable across projects! + + ```python + class LitModel(LightningModule): + def __init__(self, layer_size: int = 256, lr: float = 0.001): + ``` + +2. Preserve the recommended method order. + + ```python + class LitModel(LightningModule): + + def __init__(): + ... + + def forward(): + ... + + def training_step(): + ... + + def training_step_end(): + ... + + def training_epoch_end(): + ... + + def validation_step(): + ... + + def validation_step_end(): + ... + + def validation_epoch_end(): + ... + + def test_step(): + ... + + def test_step_end(): + ... + + def test_epoch_end(): + ... + + def configure_optimizers(): + ... + + def any_extra_hook(): + ... + ``` + +
+ +
+Version control your data and models with DVC + +Use [DVC](https://dvc.org) to version control big files, like your data or trained ML models.
+To initialize the dvc repository: + +```bash +dvc init +``` + +To start tracking a file or directory, use `dvc add`: + +```bash +dvc add data/MNIST +``` + +DVC stores information about the added file (or a directory) in a special .dvc file named data/MNIST.dvc, a small text file with a human-readable format. This file can be easily versioned like source code with Git, as a placeholder for the original data: + +```bash +git add data/MNIST.dvc data/.gitignore +git commit -m "Add raw data" +``` + +
+ +
+Support installing project as a package + +It allows other people to easily use your modules in their own projects. +Change name of the `src` folder to your project name and add `setup.py` file: + +```python +from setuptools import find_packages, setup + + +setup( + name="src", # change "src" folder name to your project name + version="0.0.0", + description="Describe Your Cool Project", + author="...", + author_email="...", + url="https://github.com/ashleve/lightning-hydra-template", # replace with your own github project link + install_requires=[ + "pytorch>=1.10.0", + "pytorch-lightning>=1.4.0", + "hydra-core>=1.1.0", + ], + packages=find_packages(), +) +``` + +Now your project can be installed from local files: + +```bash +pip install -e . +``` + +Or directly from git repository: + +```bash +pip install git+git://github.com/YourGithubName/your-repo-name.git --upgrade +``` + +So any file can be easily imported into any other file like so: + +```python +from project_name.models.mnist_module import MNISTLitModule +from project_name.datamodules.mnist_datamodule import MNISTDataModule +``` + +
+ + + + + +
+ +## Other Repositories + +
+Inspirations + +This template was inspired by: +[PyTorchLightning/deep-learninig-project-template](https://github.com/PyTorchLightning/deep-learning-project-template), +[drivendata/cookiecutter-data-science](https://github.com/drivendata/cookiecutter-data-science), +[tchaton/lightning-hydra-seed](https://github.com/tchaton/lightning-hydra-seed), +[Erlemar/pytorch_tempest](https://github.com/Erlemar/pytorch_tempest), +[lucmos/nn-template](https://github.com/lucmos/nn-template). + +
+ +
+Useful repositories + +- [pytorch/hydra-torch](https://github.com/pytorch/hydra-torch) - resources for configuring PyTorch classes with Hydra, +- [romesco/hydra-lightning](https://github.com/romesco/hydra-lightning) - resources for configuring PyTorch Lightning classes with Hydra +- [lucmos/nn-template](https://github.com/lucmos/nn-template) - similar template +- [PyTorchLightning/lightning-transformers](https://github.com/PyTorchLightning/lightning-transformers) - official Lightning Transformers repo built with Hydra + +
+ + + +
+ +## License + +This project is licensed under the MIT License. + +``` +MIT License + +Copyright (c) 2021 ashleve + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. +``` + +
+
+
+
+ +**DELETE EVERYTHING ABOVE FOR YOUR PROJECT** + +--- + +
+ +# Your Project Name + +PyTorch +Lightning +Config: Hydra +Template
+[![Paper](http://img.shields.io/badge/paper-arxiv.1001.2234-B31B1B.svg)](https://www.nature.com/articles/nature14539) +[![Conference](http://img.shields.io/badge/AnyConference-year-4b44ce.svg)](https://papers.nips.cc/paper/2020) + +
+ +## Description + +What it does + +## How to run + +Install dependencies + +```bash +# clone project +git clone https://github.com/YourGithubName/your-repo-name +cd your-repo-name + +# [OPTIONAL] create conda environment +conda create -n myenv python=3.8 +conda activate myenv + +# install pytorch according to instructions +# https://pytorch.org/get-started/ + +# install requirements +pip install -r requirements.txt +``` + +Train model with default configuration + +```bash +# train on CPU +python train.py trainer.gpus=0 + +# train on GPU +python train.py trainer.gpus=1 +``` + +Train model with chosen experiment configuration from [configs/experiment/](configs/experiment/) + +```bash +python train.py experiment=experiment_name.yaml +``` + +You can override any parameter from command line like this + +```bash +python train.py trainer.max_epochs=20 datamodule.batch_size=64 +``` diff --git a/annotation-preprocessing/.dockerignore b/annotation-preprocessing/.dockerignore deleted file mode 100644 index 95d4943f33117ceb427b49bb12f424fd5cfd8d49..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/.dockerignore +++ /dev/null @@ -1,3 +0,0 @@ -in/ -out/ - diff --git a/annotation-preprocessing/.env.example b/annotation-preprocessing/.env.example deleted file mode 100644 index 9d675afefda7fef4f36a910919de17752200c10d..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/.env.example +++ /dev/null @@ -1,8 +0,0 @@ -DB_HOST=localhost -DB_USER=someuser -DB_PASSWORD=somepassword -DB_NAME=somedatabase - -IMG_SIZE=75 -ROOT_IN=in - diff --git a/annotation-preprocessing/.gitignore b/annotation-preprocessing/.gitignore deleted file mode 100644 index 6227f60633a22e5dfc79fcb9bd5aec22bab5d7a4..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/.gitignore +++ /dev/null @@ -1,6 +0,0 @@ -*.env -out -in -*.csv -*.jpg -*.sql diff --git a/annotation-preprocessing/0_fetch_from_database.py b/annotation-preprocessing/0_fetch_from_database.py deleted file mode 100644 index 4e2acd9b83ed7571ec15399dd7a65a0d6ec40160..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/0_fetch_from_database.py +++ /dev/null @@ -1,88 +0,0 @@ -import mysql.connector -import pandas as pd -import os -from dotenv import load_dotenv - -BASE_OBJECT_SQL = """ -FROM UniqueGroundTruth - JOIN DetectedObject on DetectedObject.id = UniqueGroundTruth.object_id - JOIN Image on Image.id = DetectedObject.image_id - JOIN FocusStack on FocusStack.id = Image.focus_stack_id - JOIN Scan on Scan.id = FocusStack.scan_id - JOIN Slide on Slide.id = Scan.slide_id - JOIN ObjectType on ObjectType.id = UniqueGroundTruth.object_type_id - WHERE metaclass_id = 1 -- only select eggs; - AND study_id = 31 - ORDER BY UniqueGroundTruth.focus_stack_id -""" - -def fetch_objects_from_datase(db): - cursor = db.cursor() - - cursor.execute("""SELECT - UniqueGroundTruth.focus_stack_id, - UniqueGroundTruth.x_min, - UniqueGroundTruth.y_min, - UniqueGroundTruth.x_max, - UniqueGroundTruth.y_max, - UniqueGroundTruth.object_type_id, - ObjectType.name, - Image.add_date""" + BASE_OBJECT_SQL) - - result = cursor.fetchall() - return result - -def fetch_focus_stacks_from_database(db): - cursor = db.cursor() - - cursor.execute("""SELECT - FocusStack.id as foucs_stack_id, - CONCAT (study_id, "/", uuid, "/", file_name) as file_path, - file_name, - uuid, - study_id, - Image.pos_z, - Image.focus_value, - Image.add_date - FROM FocusStack - JOIN Scan on Scan.id = FocusStack.scan_id - JOIN Slide on Slide.id = Scan.slide_id - JOIN Study on Study .id = Slide.study_id - JOIN Image on Image.focus_stack_id = FocusStack.id - WHERE - FocusStack.id IN( -- get all focus stacks that have objects in them; - SELECT DISTINCT - UniqueGroundTruth.focus_stack_id - """ + BASE_OBJECT_SQL - + """ - ) - ORDER BY FocusStack.id DESC, focus_value, focus_level - """) - result = cursor.fetchall() - return result - - - -if __name__ == "__main__": - load_dotenv() - - db = mysql.connector.connect( - host=os.getenv('DB_HOST'), - user=os.getenv('DB_USER'), - password=os.getenv('DB_PASSWORD'), - database=os.getenv('DB_NAME') - ) - - print("Querring objects...") - df_objects = pd.DataFrame(fetch_objects_from_datase(db)) - print("Querring stacks...") - df_stacks = pd.DataFrame(fetch_focus_stacks_from_database(db)) - - df_objects.columns = ['stack_id', 'x_min', 'y_min', 'x_max', 'y_max', 'object_type_id', 'name', 'add_date'] - df_stacks.columns = ['stack_id', 'file_path', 'file_name', - 'uuid', 'study_id', 'pos_z', 'focus_value', 'add_date'] - - print("Writing objects to file...") - df_objects.to_csv("out/objects.csv") - print("Writing stacks to file...") - df_stacks.to_csv("out/stacks.csv") diff --git a/annotation-preprocessing/1_splitting_into_patches.py b/annotation-preprocessing/1_splitting_into_patches.py deleted file mode 100644 index 51435e6c51a02de2ce96bc07764b133986fb2151..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/1_splitting_into_patches.py +++ /dev/null @@ -1,165 +0,0 @@ -import pandas as pd -from collections import defaultdict -from dotenv import load_dotenv -import os -from PIL import Image, ImageDraw -import math -import json -import random - -class StackEntry: - def __init__(self): - self.images = [] - self.objects = [] - def add_image(self, image): - self.images.append(image) - def add_object(self, object): - self.objects.append(object) - def sort(self): - self.images.sort(key=lambda x: x.focus_value) - -def get_neighbours(img, x, y, dimensions): - neighbour_candidates = [(-1,-1), (0, -1), (1, -1), (-1, 0), (1,0), (-1,1), (0,1), (1,1)] - - width, height = img.size - - neighbours = [] - for x_offset, y_offset in neighbour_candidates: - neighbour_x = x + x_offset * dimensions - neighbour_y = y + y_offset * dimensions - - if neighbour_x >= 0 and neighbour_x + dimensions <= width and neighbour_y >= 0 and neighbour_y + dimensions <= height: - box = [neighbour_x, neighbour_y, neighbour_x + dimensions, neighbour_y + dimensions] - neighbours.append((neighbour_x, neighbour_y, img.crop(box))) - else: - neighbours.append(None) - return neighbours - -def extract_object_tiles(obj, stack_images, in_folder, threshold = 0.25): - x_start = int(obj.x_min / size) * size - x_end = int(math.ceil(obj.x_max / size)) * size - y_start = int(obj.y_min / size) * size - y_end = int(math.ceil(obj.y_max / size)) * size - - tiles = [] - - focus_stack_images = list(map(lambda x: (x, Image.open(os.path.join(in_folder, x.file_path))), stack_images)) - - # Get tiles of the image that contain bounding box of object - for y in range(y_start, y_end, size): - for x in range(x_start, x_end, size): - - if compute_overlap([x, y, x + size, y + size], [obj.x_min, obj.y_min, obj.x_max, obj.y_max]) > size * size * threshold: - stack = [] - for row, img in focus_stack_images: - box = [x, y, x + size, y + size] - crop = img.crop(box) - - neighbours = get_neighbours(img, x, y, size) - stack.append((row, box[:2], crop, neighbours)) - tiles.append(stack) - return tiles - - -def save_tile(original_file_path, out_dir, x : int, y : int, img, overwrite = False): - path, file_name = os.path.split(original_file_path) - name, ext = os.path.splitext(file_name) - - out_path = os.path.join(out_dir, path) - save_to = os.path.join(out_path, f'{name}_{x}_{y}{ext}') - - if not os.path.exists(out_path): - os.makedirs(out_path) - if overwrite or not os.path.exists(save_to): - img.save(save_to) - return os.path.join(path, f'{name}_{x}_{y}{ext}') - -def compute_overlap(rect1, rect2): - dx = min(rect1[2], rect2[2]) - max(rect1[0], rect2[0]) - dy = min(rect1[3], rect2[3]) - max(rect1[1], rect2[1]) - return dx * dy - -def save_obj_tiles(obj, out_folder, in_folder, stack_images): - extracted = extract_object_tiles(obj, stack_images, in_folder) - z_stacks = [] - for z_stack in extracted: - z_stack_images = [] - for row, box, img, neigbours in z_stack: - - neighbours = [] - - image_path = save_tile(row.file_path, out_folder, box[0], box[1], img) - for neighbour in neigbours: - n_path = None - if neighbour: - x, y, n_img = neighbour - n_path = save_tile(row.file_path, out_folder, x, y, n_img) - neighbours.append(n_path) - - z_stack_images.append({ - "focus_value": row["focus_value"], - "image_path": image_path, - "neighbours": neighbours, - "original_filename": row["file_name"], - "scan_uuid": row["uuid"], - "study_id": row["study_id"], - }) - z_stacks.append({ - "best_index": None, - "images" : z_stack_images, - "obj_name": obj["name"], - "stack_id": obj["stack_id"], - }) - - return z_stacks - -def save_stack(stack, out_folder, in_folder): - z_stacks = [] - for obj in stack.objects: - z_stacks.extend(save_obj_tiles(obj, out_folder, in_folder, stack.images)) - return z_stacks - - -if __name__ == "__main__": - load_dotenv() - print("Geting environment variables...") - size = int(os.getenv('IMG_SIZE')) - root_in = os.getenv('ROOT_IN') - - print(f'img_size: ') - print(f'in_folder: {root_in}') - - print("Loading data from csv files...") - objects = pd.read_csv("out/objects.csv", index_col=0) - stacks = pd.read_csv("out/stacks.csv", index_col=0) - - - stacks_dict = defaultdict(lambda: StackEntry()) - - print("Building internal datastructure...") - # adding images to dict - for (index, row) in stacks.iterrows(): - stacks_dict[row.stack_id].add_image(row) - - for values in stacks_dict.values(): - values.sort() - - # adding objects - for (index, row) in objects.iterrows(): - stacks_dict[row.stack_id].add_object(row) - - out_folder = "out" - z_stacks = [] - - print("Generating image tiles and writing them to file...") - for stack in stacks_dict.values(): - z_stacks.extend(save_stack(stack,"out", root_in)) - - # randomize z_stacks - print("Shuffling data...") - random.shuffle(z_stacks) - - print("Writing meta-data for annotation to file...") - with open(os.path.join(out_folder, "data.json"), 'w') as file: - file.write(json.dumps(z_stacks)) - diff --git a/annotation-preprocessing/Dockerfile b/annotation-preprocessing/Dockerfile deleted file mode 100644 index 6cb413a316e4ae0836812fd3e516cfde32f6e324..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/Dockerfile +++ /dev/null @@ -1,13 +0,0 @@ -FROM python:3.7 - -WORKDIR /usr/src/app - -RUN apt-get update -RUN apt-get install libgl1 -y - -COPY requirements.txt ./ -RUN pip install --no-cache-dir -r requirements.txt - -COPY *.py ./ - -CMD sh -c "python 0_fetch_from_database.py && python 1_splitting_into_patches.py" diff --git a/annotation-preprocessing/README.md b/annotation-preprocessing/README.md deleted file mode 100644 index e4e0569dfdf889d16f4e8b113e7969658f29b197..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/README.md +++ /dev/null @@ -1,48 +0,0 @@ -# Annotation Preprocessing - -This directory contains code to extract image metadata from the database. In the first step the metadata is converted to csv files. The second step then loads the metadata and creates small tiles out of all images that contain eggs. The corresponding information about these patches is stored in a json file which cna be read by [Focus Annotator -](https://github.com/13hannes11/focus_annotator). - - -## Environment Variables - -To run the preprocessing you need to create a `.env` file or set the corresponding environmental variables directly. The content of the file should be all of the following: - -For step 0, fetching data from the database you need: - -``` -DB_HOST= -DB_USER= -DB_PASSWORD= -DB_NAME= -``` - -For step 1, cropping and extracting images with eggs you need: - -``` -IMG_SIZE=75 -ROOT_IN="in" -``` - -The actual code can be either run in a docker-container, for that you can run `docker-compose up` inside the this directory. Make sure you edit the mount in the docker-compose to your directories: - -```yaml - volumes: - - :/usr/src/app/out:z - - :/usr/src/app/in:z -``` - - -Alternatively, you can manually run these python steps: - -``` -python 0_fetch_from_database.py -``` - -and - -``` -python 1_splitting_into_patches.py -``` - - diff --git a/annotation-preprocessing/docker-compose.yml b/annotation-preprocessing/docker-compose.yml deleted file mode 100644 index ea6907e8299ef5fb1f6384b8bfed1854ae086b61..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/docker-compose.yml +++ /dev/null @@ -1,10 +0,0 @@ -version: "3" # optional since v1.27.0 -services: - preprocess: - build: . - volumes: - - ./out/:/usr/src/app/out:z - - ./in/:/usr/src/app/in:z - env_file: - - .env - network_mode: host # use host networking; you can also just link container networks directly diff --git a/annotation-preprocessing/out/.gitignore b/annotation-preprocessing/out/.gitignore deleted file mode 100644 index e69de29bb2d1d6434b8b29ae775ad8c2e48c5391..0000000000000000000000000000000000000000 diff --git a/annotation-preprocessing/requirements.txt b/annotation-preprocessing/requirements.txt deleted file mode 100644 index e2bcac8a1dcfff9af87bcb81d4c7c87dd0eaea12..0000000000000000000000000000000000000000 --- a/annotation-preprocessing/requirements.txt +++ /dev/null @@ -1,6 +0,0 @@ -opencv-python -numpy -pillow -mysql-connector-python -pandas -python-dotenv diff --git a/models/configs/callbacks/default.yaml b/configs/callbacks/default.yaml similarity index 100% rename from models/configs/callbacks/default.yaml rename to configs/callbacks/default.yaml diff --git a/models/configs/callbacks/none.yaml b/configs/callbacks/none.yaml similarity index 100% rename from models/configs/callbacks/none.yaml rename to configs/callbacks/none.yaml diff --git a/models/configs/datamodule/focus.yaml b/configs/datamodule/focus.yaml similarity index 100% rename from models/configs/datamodule/focus.yaml rename to configs/datamodule/focus.yaml diff --git a/models/configs/datamodule/mnist.yaml b/configs/datamodule/mnist.yaml similarity index 100% rename from models/configs/datamodule/mnist.yaml rename to configs/datamodule/mnist.yaml diff --git a/models/configs/debug/default.yaml b/configs/debug/default.yaml similarity index 100% rename from models/configs/debug/default.yaml rename to configs/debug/default.yaml diff --git a/models/configs/debug/limit_batches.yaml b/configs/debug/limit_batches.yaml similarity index 100% rename from models/configs/debug/limit_batches.yaml rename to configs/debug/limit_batches.yaml diff --git a/models/configs/debug/overfit.yaml b/configs/debug/overfit.yaml similarity index 100% rename from models/configs/debug/overfit.yaml rename to configs/debug/overfit.yaml diff --git a/models/configs/debug/profiler.yaml b/configs/debug/profiler.yaml similarity index 100% rename from models/configs/debug/profiler.yaml rename to configs/debug/profiler.yaml diff --git a/models/configs/debug/step.yaml b/configs/debug/step.yaml similarity index 100% rename from models/configs/debug/step.yaml rename to configs/debug/step.yaml diff --git a/models/configs/debug/test_only.yaml b/configs/debug/test_only.yaml similarity index 100% rename from models/configs/debug/test_only.yaml rename to configs/debug/test_only.yaml diff --git a/models/configs/experiment/example.yaml b/configs/experiment/example.yaml similarity index 100% rename from models/configs/experiment/example.yaml rename to configs/experiment/example.yaml diff --git a/models/configs/hparams_search/mnist_optuna.yaml b/configs/hparams_search/mnist_optuna.yaml similarity index 100% rename from models/configs/hparams_search/mnist_optuna.yaml rename to configs/hparams_search/mnist_optuna.yaml diff --git a/models/configs/local/.gitkeep b/configs/local/.gitkeep similarity index 100% rename from models/configs/local/.gitkeep rename to configs/local/.gitkeep diff --git a/models/configs/log_dir/debug.yaml b/configs/log_dir/debug.yaml similarity index 100% rename from models/configs/log_dir/debug.yaml rename to configs/log_dir/debug.yaml diff --git a/models/configs/log_dir/default.yaml b/configs/log_dir/default.yaml similarity index 100% rename from models/configs/log_dir/default.yaml rename to configs/log_dir/default.yaml diff --git a/models/configs/log_dir/evaluation.yaml b/configs/log_dir/evaluation.yaml similarity index 100% rename from models/configs/log_dir/evaluation.yaml rename to configs/log_dir/evaluation.yaml diff --git a/models/configs/logger/comet.yaml b/configs/logger/comet.yaml similarity index 100% rename from models/configs/logger/comet.yaml rename to configs/logger/comet.yaml diff --git a/models/configs/logger/csv.yaml b/configs/logger/csv.yaml similarity index 100% rename from models/configs/logger/csv.yaml rename to configs/logger/csv.yaml diff --git a/models/configs/logger/many_loggers.yaml b/configs/logger/many_loggers.yaml similarity index 100% rename from models/configs/logger/many_loggers.yaml rename to configs/logger/many_loggers.yaml diff --git a/models/configs/logger/mlflow.yaml b/configs/logger/mlflow.yaml similarity index 100% rename from models/configs/logger/mlflow.yaml rename to configs/logger/mlflow.yaml diff --git a/models/configs/logger/neptune.yaml b/configs/logger/neptune.yaml similarity index 100% rename from models/configs/logger/neptune.yaml rename to configs/logger/neptune.yaml diff --git a/models/configs/logger/tensorboard.yaml b/configs/logger/tensorboard.yaml similarity index 100% rename from models/configs/logger/tensorboard.yaml rename to configs/logger/tensorboard.yaml diff --git a/models/configs/logger/wandb.yaml b/configs/logger/wandb.yaml similarity index 100% rename from models/configs/logger/wandb.yaml rename to configs/logger/wandb.yaml diff --git a/models/configs/model/focus.yaml b/configs/model/focus.yaml similarity index 100% rename from models/configs/model/focus.yaml rename to configs/model/focus.yaml diff --git a/models/configs/model/mnist.yaml b/configs/model/mnist.yaml similarity index 100% rename from models/configs/model/mnist.yaml rename to configs/model/mnist.yaml diff --git a/models/configs/test.yaml b/configs/test.yaml similarity index 100% rename from models/configs/test.yaml rename to configs/test.yaml diff --git a/models/configs/train.yaml b/configs/train.yaml similarity index 100% rename from models/configs/train.yaml rename to configs/train.yaml diff --git a/models/configs/trainer/ddp.yaml b/configs/trainer/ddp.yaml similarity index 100% rename from models/configs/trainer/ddp.yaml rename to configs/trainer/ddp.yaml diff --git a/models/configs/trainer/default.yaml b/configs/trainer/default.yaml similarity index 100% rename from models/configs/trainer/default.yaml rename to configs/trainer/default.yaml diff --git a/models/configs/trainer/long.yaml b/configs/trainer/long.yaml similarity index 100% rename from models/configs/trainer/long.yaml rename to configs/trainer/long.yaml diff --git a/data-preprocessing/.env.example b/data-preprocessing/.env.example deleted file mode 100644 index 8dfa85b475ae69a849b3a20050daabfac8654dc4..0000000000000000000000000000000000000000 --- a/data-preprocessing/.env.example +++ /dev/null @@ -1,3 +0,0 @@ -DATA_FILE=data.json -OUT_FILE=metadata.csv - diff --git a/data-preprocessing/extract_annotations.py b/data-preprocessing/extract_annotations.py deleted file mode 100644 index 33fb5036bee67e3689bc267e41f9416a74b74abf..0000000000000000000000000000000000000000 --- a/data-preprocessing/extract_annotations.py +++ /dev/null @@ -1,45 +0,0 @@ -import json -import os -from itertools import chain -from dotenv import load_dotenv -import pandas as pd - -def to_relative_focus(stack): - best_index = stack["best_index"] - images = stack["images"] - - best_value = images[best_index]["focus_value"] - for i in range(len(images)): - images[i]["focus_value"] = images[i]["focus_value"] - best_value - return stack - - -def flatten_stack(stack): - images = stack["images"] - - def f(image): - del image["neighbours"] - image["stack_id"] = stack["stack_id"] - image["obj_name"] = stack["obj_name"] - return image - - images = list(map(f, images)) - return images - - -if __name__ == "__main__": - load_dotenv() - data_file = os.getenv('DATA_FILE') - out_file = os.getenv('OUT_FILE') - - with open(data_file) as f: - content = json.load(f) - - annotated = filter(lambda x: x["best_index"], content) - relative_focus = map(to_relative_focus, annotated) - flattened = chain(*map(flatten_stack,relative_focus)) - - dataframe = pd.DataFrame(flattened) - dataframe.to_csv(out_file) - - \ No newline at end of file diff --git a/data-preprocessing/requirements.txt b/data-preprocessing/requirements.txt deleted file mode 100644 index 3e338bfa253ebc1d00943c5b4413d66ffc145f98..0000000000000000000000000000000000000000 --- a/data-preprocessing/requirements.txt +++ /dev/null @@ -1 +0,0 @@ -python-dotenv \ No newline at end of file diff --git a/models/docker-compose.cuda.yml b/docker-compose.cuda.yml similarity index 100% rename from models/docker-compose.cuda.yml rename to docker-compose.cuda.yml diff --git a/models/docker-compose.yml b/docker-compose.yml similarity index 100% rename from models/docker-compose.yml rename to docker-compose.yml diff --git a/focus_annotator b/focus_annotator deleted file mode 160000 index fe8ee5b5cbaf9271668fbf003c0a3ccac3fdb65b..0000000000000000000000000000000000000000 --- a/focus_annotator +++ /dev/null @@ -1 +0,0 @@ -Subproject commit fe8ee5b5cbaf9271668fbf003c0a3ccac3fdb65b diff --git a/models/.dockerignore b/models/.dockerignore deleted file mode 100644 index 903ae1bbf18bbcf74b36a90fcad63aef313693d8..0000000000000000000000000000000000000000 --- a/models/.dockerignore +++ /dev/null @@ -1,4 +0,0 @@ -*.env -data -logs -configs diff --git a/models/.env.example b/models/.env.example deleted file mode 100644 index dc8d1cd79e8df7089dca7ece769020d52b7b5c41..0000000000000000000000000000000000000000 --- a/models/.env.example +++ /dev/null @@ -1,5 +0,0 @@ -# this is example of the file that can be used for storing private and user specific environment variables, like keys or system paths -# create a file named .env (by default .env will be excluded from version control) -# the variables declared in .env are loaded in train.py automatically -# hydra allows you to reference variables in .yaml configs with special syntax: ${oc.env:MY_VAR} - diff --git a/models/.gitignore b/models/.gitignore deleted file mode 100644 index 68466635f41c3ece6488c793c8508e3a1f841fc2..0000000000000000000000000000000000000000 --- a/models/.gitignore +++ /dev/null @@ -1,148 +0,0 @@ -# Byte-compiled / optimized / DLL files -__pycache__/ -*.py[cod] -*$py.class - -# C extensions -*.so - -# Distribution / packaging -.Python -build/ -develop-eggs/ -dist/ -downloads/ -eggs/ -.eggs/ -lib/ -lib64/ -parts/ -sdist/ -var/ -wheels/ -pip-wheel-metadata/ -share/python-wheels/ -*.egg-info/ -.installed.cfg -*.egg -MANIFEST - -# PyInstaller -# Usually these files are written by a python script from a template -# before PyInstaller builds the exe, so as to inject date/other infos into it. -*.manifest -*.spec - -# Installer logs -pip-log.txt -pip-delete-this-directory.txt - -# Unit test / coverage reports -htmlcov/ -.tox/ -.nox/ -.coverage -.coverage.* -.cache -nosetests.xml -coverage.xml -*.cover -*.py,cover -.hypothesis/ -.pytest_cache/ - -# Translations -*.mo -*.pot - -# Django stuff: -*.log -local_settings.py -db.sqlite3 -db.sqlite3-journal - -# Flask stuff: -instance/ -.webassets-cache - -# Scrapy stuff: -.scrapy - -# Sphinx documentation -docs/_build/ - -# PyBuilder -target/ - -# Jupyter Notebook -.ipynb_checkpoints - -# IPython -profile_default/ -ipython_config.py - -# pyenv -.python-version - -# pipenv -# According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control. -# However, in case of collaboration, if having platform-specific dependencies or dependencies -# having no cross-platform support, pipenv may install dependencies that don't work, or not -# install all needed dependencies. -#Pipfile.lock - -# PEP 582; used by e.g. github.com/David-OConnor/pyflow -__pypackages__/ - -# Celery stuff -celerybeat-schedule -celerybeat.pid - -# SageMath parsed files -*.sage.py - -# Environments -.venv -env/ -venv/ -ENV/ -env.bak/ -venv.bak/ - -# Spyder project settings -.spyderproject -.spyproject - -# Rope project settings -.ropeproject - -# mkdocs documentation -/site - -# mypy -.mypy_cache/ -.dmypy.json -dmypy.json - -# Pyre type checker -.pyre/ - -### VisualStudioCode -.vscode/* -!.vscode/settings.json -!.vscode/tasks.json -!.vscode/launch.json -!.vscode/extensions.json -*.code-workspace -**/.vscode - -# JetBrains -.idea/ - -# Lightning-Hydra-Template -configs/local/default.yaml -data/ -logs/ -wandb/ -.env -.autoenv diff --git a/models/.pre-commit-config.yaml b/models/.pre-commit-config.yaml deleted file mode 100644 index 9173e1e1c4f599ecf53588ce0d98f3a1e13b5c7e..0000000000000000000000000000000000000000 --- a/models/.pre-commit-config.yaml +++ /dev/null @@ -1,44 +0,0 @@ -repos: - - repo: https://github.com/pre-commit/pre-commit-hooks - rev: v4.1.0 - hooks: - # list of supported hooks: https://pre-commit.com/hooks.html - - id: trailing-whitespace - - id: end-of-file-fixer - - id: check-yaml - - id: check-added-large-files - - id: debug-statements - - id: detect-private-key - - # python code formatting - - repo: https://github.com/psf/black - rev: 22.1.0 - hooks: - - id: black - args: [--line-length, "99"] - - # python import sorting - - repo: https://github.com/PyCQA/isort - rev: 5.10.1 - hooks: - - id: isort - args: ["--profile", "black", "--filter-files"] - - # yaml formatting - - repo: https://github.com/pre-commit/mirrors-prettier - rev: v2.5.1 - hooks: - - id: prettier - types: [yaml] - - # python code analysis - - repo: https://github.com/PyCQA/flake8 - rev: 4.0.1 - hooks: - - id: flake8 - - # jupyter notebook cell output clearing - - repo: https://github.com/kynan/nbstripout - rev: 0.5.0 - hooks: - - id: nbstripout diff --git a/models/README.md b/models/README.md deleted file mode 100644 index 50fc8695712829ff9a9a1430597644f5b79c374d..0000000000000000000000000000000000000000 --- a/models/README.md +++ /dev/null @@ -1,1445 +0,0 @@ -
- -# Lightning-Hydra-Template - -Python -PyTorch -Lightning -Config: hydra -Code style: black - -A clean and scalable template to kickstart your deep learning project πŸš€βš‘πŸ”₯
-Click on [Use this template](https://github.com/ashleve/lightning-hydra-template/generate) to initialize new repository. - -_Suggestions are always welcome!_ - -
- -

- -## πŸ“Œ  Introduction - -This template tries to be as general as possible. It integrates many different MLOps tools. - -> Effective usage of this template requires learning of a couple of technologies: [PyTorch](https://pytorch.org), [PyTorch Lightning](https://www.pytorchlightning.ai) and [Hydra](https://hydra.cc). Knowledge of some experiment logging framework like [Weights&Biases](https://wandb.com), [Neptune](https://neptune.ai) or [MLFlow](https://mlflow.org) is also recommended. - -**Why you should use it:** it allows you to rapidly iterate over new models/datasets and scale your projects from small single experiments to hyperparameter searches on computing clusters, without writing any boilerplate code. To my knowledge, it's one of the most convenient all-in-one technology stack for Deep Learning research. Good starting point for reproducing papers, kaggle competitions or small-team research projects. It's also a collection of best practices for efficient workflow and reproducibility. - -**Why you shouldn't use it:** this template is not fitted to be a production environment, should be used more as a fast experimentation tool. Apart from that, Lightning and Hydra are still evolving and integrate many libraries, which means sometimes things break - for the list of currently known bugs, visit [this page](https://github.com/ashleve/lightning-hydra-template/labels/bug). Also, even though Lightning is very flexible, it's not well suited for every possible deep learning task. See [#Limitations](#limitations) for more. - -### Why PyTorch Lightning? - -[PyTorch Lightning](https://github.com/PyTorchLightning/pytorch-lightning) is a lightweight PyTorch wrapper for high-performance AI research. -It makes your code neatly organized and provides lots of useful features, like ability to run model on CPU, GPU, multi-GPU cluster and TPU. - -### Why Hydra? - -[Hydra](https://github.com/facebookresearch/hydra) is an open-source Python framework that simplifies the development of research and other complex applications. The key feature is the ability to dynamically create a hierarchical configuration by composition and override it through config files and the command line. It allows you to conveniently manage experiments and provides many useful plugins, like [Optuna Sweeper](https://hydra.cc/docs/next/plugins/optuna_sweeper) for hyperparameter search, or [Ray Launcher](https://hydra.cc/docs/next/plugins/ray_launcher) for running jobs on a cluster. - -
- -## Main Ideas Of This Template - -- **Predefined Structure**: clean and scalable so that work can easily be extended and replicated | [#Project Structure](#project-structure) -- **Rapid Experimentation**: thanks to automating pipeline with config files and hydra command line superpowers | [#Your Superpowers](#your-superpowers) -- **Reproducibility**: obtaining similar results is supported in multiple ways | [#Reproducibility](#reproducibility) -- **Little Boilerplate**: so pipeline can be easily modified | [#How It Works](#how-it-works) -- **Main Configuration**: main config file specifies default training configuration | [#Main Project Configuration](#main-project-configuration) -- **Experiment Configurations**: can be composed out of smaller configs and override chosen hyperparameters | [#Experiment Configuration](#experiment-configuration) -- **Workflow**: comes down to 4 simple steps | [#Workflow](#workflow) -- **Experiment Tracking**: many logging frameworks can be easily integrated, like Tensorboard, MLFlow or W&B | [#Experiment Tracking](#experiment-tracking) -- **Logs**: all logs (checkpoints, data from loggers, hparams, etc.) are stored in a convenient folder structure imposed by Hydra | [#Logs](#logs) -- **Hyperparameter Search**: made easier with Hydra built-in plugins like [Optuna Sweeper](https://hydra.cc/docs/next/plugins/optuna_sweeper) | [#Hyperparameter Search](#hyperparameter-search) -- **Tests**: unit tests and shell/command based tests for speeding up the development | [#Tests](#tests) -- **Best Practices**: a couple of recommended tools, practices and standards for efficient workflow and reproducibility | [#Best Practices](#best-practices) - -
- -## Project Structure - -The directory structure of new project looks like this: - -``` -β”œβ”€β”€ configs <- Hydra configuration files -β”‚ β”œβ”€β”€ callbacks <- Callbacks configs -β”‚ β”œβ”€β”€ datamodule <- Datamodule configs -β”‚ β”œβ”€β”€ debug <- Debugging configs -β”‚ β”œβ”€β”€ experiment <- Experiment configs -β”‚ β”œβ”€β”€ hparams_search <- Hyperparameter search configs -β”‚ β”œβ”€β”€ local <- Local configs -β”‚ β”œβ”€β”€ log_dir <- Logging directory configs -β”‚ β”œβ”€β”€ logger <- Logger configs -β”‚ β”œβ”€β”€ model <- Model configs -β”‚ β”œβ”€β”€ trainer <- Trainer configs -β”‚ β”‚ -β”‚ β”œβ”€β”€ test.yaml <- Main config for testing -β”‚ └── train.yaml <- Main config for training -β”‚ -β”œβ”€β”€ data <- Project data -β”‚ -β”œβ”€β”€ logs <- Logs generated by Hydra and PyTorch Lightning loggers -β”‚ -β”œβ”€β”€ notebooks <- Jupyter notebooks. Naming convention is a number (for ordering), -β”‚ the creator's initials, and a short `-` delimited description, -β”‚ e.g. `1.0-jqp-initial-data-exploration.ipynb`. -β”‚ -β”œβ”€β”€ scripts <- Shell scripts -β”‚ -β”œβ”€β”€ src <- Source code -β”‚ β”œβ”€β”€ datamodules <- Lightning datamodules -β”‚ β”œβ”€β”€ models <- Lightning models -β”‚ β”œβ”€β”€ utils <- Utility scripts -β”‚ β”œβ”€β”€ vendor <- Third party code that cannot be installed using PIP/Conda -β”‚ β”‚ -β”‚ β”œβ”€β”€ testing_pipeline.py -β”‚ └── training_pipeline.py -β”‚ -β”œβ”€β”€ tests <- Tests of any kind -β”‚ β”œβ”€β”€ helpers <- A couple of testing utilities -β”‚ β”œβ”€β”€ shell <- Shell/command based tests -β”‚ └── unit <- Unit tests -β”‚ -β”œβ”€β”€ test.py <- Run testing -β”œβ”€β”€ train.py <- Run training -β”‚ -β”œβ”€β”€ .env.example <- Template of the file for storing private environment variables -β”œβ”€β”€ .gitignore <- List of files/folders ignored by git -β”œβ”€β”€ .pre-commit-config.yaml <- Configuration of pre-commit hooks for code formatting -β”œβ”€β”€ requirements.txt <- File for installing python dependencies -β”œβ”€β”€ setup.cfg <- Configuration of linters and pytest -└── README.md -``` - -
- -## πŸš€  Quickstart - -```bash -# clone project -git clone https://github.com/ashleve/lightning-hydra-template -cd lightning-hydra-template - -# [OPTIONAL] create conda environment -conda create -n myenv python=3.8 -conda activate myenv - -# install pytorch according to instructions -# https://pytorch.org/get-started/ - -# install requirements -pip install -r requirements.txt -``` - -Template contains example with MNIST classification.
-When running `python train.py` you should see something like this: - -
- -![](https://github.com/ashleve/lightning-hydra-template/blob/resources/terminal.png) - -
- -### ⚑  Your Superpowers - -
-Override any config parameter from command line - -> Hydra allows you to easily overwrite any parameter defined in your config. - -```bash -python train.py trainer.max_epochs=20 model.lr=1e-4 -``` - -> You can also add new parameters with `+` sign. - -```bash -python train.py +model.new_param="uwu" -``` - -
- -
-Train on CPU, GPU, multi-GPU and TPU - -> PyTorch Lightning makes it easy to train your models on different hardware. - -```bash -# train on CPU -python train.py trainer.gpus=0 - -# train on 1 GPU -python train.py trainer.gpus=1 - -# train on TPU -python train.py +trainer.tpu_cores=8 - -# train with DDP (Distributed Data Parallel) (4 GPUs) -python train.py trainer.gpus=4 +trainer.strategy=ddp - -# train with DDP (Distributed Data Parallel) (8 GPUs, 2 nodes) -python train.py trainer.gpus=4 +trainer.num_nodes=2 +trainer.strategy=ddp -``` - -
- -
-Train with mixed precision - -```bash -# train with pytorch native automatic mixed precision (AMP) -python train.py trainer.gpus=1 +trainer.precision=16 -``` - -
- - - -
-Train model with any logger available in PyTorch Lightning, like Weights&Biases or Tensorboard - -> PyTorch Lightning provides convenient integrations with most popular logging frameworks, like Tensorboard, Neptune or simple csv files. Read more [here](#experiment-tracking). Using wandb requires you to [setup account](https://www.wandb.com/) first. After that just complete the config as below.
> **Click [here](https://wandb.ai/hobglob/template-dashboard/) to see example wandb dashboard generated with this template.** - -```bash -# set project and entity names in `configs/logger/wandb` -wandb: - project: "your_project_name" - entity: "your_wandb_team_name" -``` - -```bash -# train model with Weights&Biases (link to wandb dashboard should appear in the terminal) -python train.py logger=wandb -``` - -
- -
-Train model with chosen experiment config - -> Experiment configurations are placed in [configs/experiment/](configs/experiment/). - -```bash -python train.py experiment=example -``` - -
- -
-Attach some callbacks to run - -> Callbacks can be used for things such as as model checkpointing, early stopping and [many more](https://pytorch-lightning.readthedocs.io/en/latest/extensions/callbacks.html#built-in-callbacks).
-> Callbacks configurations are placed in [configs/callbacks/](configs/callbacks/). - -```bash -python train.py callbacks=default -``` - -
- -
-Use different tricks available in Pytorch Lightning - -> PyTorch Lightning provides about [40+ useful trainer flags](https://pytorch-lightning.readthedocs.io/en/latest/common/trainer.html#trainer-flags). - -```yaml -# gradient clipping may be enabled to avoid exploding gradients -python train.py +trainer.gradient_clip_val=0.5 - -# stochastic weight averaging can make your models generalize better -python train.py +trainer.stochastic_weight_avg=true - -# run validation loop 4 times during a training epoch -python train.py +trainer.val_check_interval=0.25 - -# accumulate gradients -python train.py +trainer.accumulate_grad_batches=10 - -# terminate training after 12 hours -python train.py +trainer.max_time="00:12:00:00" -``` - -
- -
-Easily debug - -> Visit [configs/debug/](configs/debug/) for different debugging configs. - -```bash -# runs 1 epoch in default debugging mode -# changes logging directory to `logs/debugs/...` -# sets level of all command line loggers to 'DEBUG' -# enables extra trainer flags like tracking gradient norm -# enforces debug-friendly configuration -python train.py debug=default - -# runs test epoch without training -python train.py debug=test_only - -# run 1 train, val and test loop, using only 1 batch -python train.py +trainer.fast_dev_run=true - -# raise exception if there are any numerical anomalies in tensors, like NaN or +/-inf -python train.py +trainer.detect_anomaly=true - -# print execution time profiling after training ends -python train.py +trainer.profiler="simple" - -# try overfitting to 1 batch -python train.py +trainer.overfit_batches=1 trainer.max_epochs=20 - -# use only 20% of the data -python train.py +trainer.limit_train_batches=0.2 \ -+trainer.limit_val_batches=0.2 +trainer.limit_test_batches=0.2 - -# log second gradient norm of the model -python train.py +trainer.track_grad_norm=2 -``` - -
- -
-Resume training from checkpoint - -> Checkpoint can be either path or URL. - -```yaml -python train.py trainer.resume_from_checkpoint="/path/to/ckpt/name.ckpt" -``` - -> ⚠️ Currently loading ckpt in Lightning doesn't resume logger experiment, but it will be supported in future Lightning release. - -
- -
-Execute evaluation for a given checkpoint - -> Checkpoint can be either path or URL. - -```yaml -python test.py ckpt_path="/path/to/ckpt/name.ckpt" -``` - -
- -
-Create a sweep over hyperparameters - -```bash -# this will run 6 experiments one after the other, -# each with different combination of batch_size and learning rate -python train.py -m datamodule.batch_size=32,64,128 model.lr=0.001,0.0005 -``` - -> ⚠️ This sweep is not failure resistant (if one job crashes than the whole sweep crashes). - -
- -
-Create a sweep over hyperparameters with Optuna - -> Using [Optuna Sweeper](https://hydra.cc/docs/next/plugins/optuna_sweeper) plugin doesn't require you to code any boilerplate into your pipeline, everything is defined in a [single config file](configs/hparams_search/mnist_optuna.yaml)! - -```bash -# this will run hyperparameter search defined in `configs/hparams_search/mnist_optuna.yaml` -# over chosen experiment config -python train.py -m hparams_search=mnist_optuna experiment=example_simple -``` - -> ⚠️ Currently this sweep is not failure resistant (if one job crashes than the whole sweep crashes). Might be supported in future Hydra release. - -
- -
-Execute all experiments from folder - -> Hydra provides special syntax for controlling behavior of multiruns. Learn more [here](https://hydra.cc/docs/next/tutorials/basic/running_your_app/multi-run). The command below executes all experiments from folder [configs/experiment/](configs/experiment/). - -```bash -python train.py -m 'experiment=glob(*)' -``` - -
- -
-Execute sweep on a remote AWS cluster - -> This should be achievable with simple config using [Ray AWS launcher for Hydra](https://hydra.cc/docs/next/plugins/ray_launcher). Example is not yet implemented in this template. - -
- - - -
-Use Hydra tab completion - -> Hydra allows you to autocomplete config argument overrides in shell as you write them, by pressing `tab` key. Learn more [here](https://hydra.cc/docs/tutorials/basic/running_your_app/tab_completion). - -
- -
-Apply pre-commit hooks - -> Apply pre-commit hooks to automatically format your code and configs, perform code analysis and remove output from jupyter notebooks. See [# Best Practices](#best-practices) for more. - -```bash -pre-commit run -a -``` - -
- -
- -## ❀️  Contributions - -Have a question? Found a bug? Missing a specific feature? Have an idea for improving documentation? Feel free to file a new issue, discussion or PR with respective title and description. If you already found a solution to your problem, don't hesitate to share it. Suggestions for new best practices are always welcome! - -
- -## ℹ️  Guide - -### How To Get Started - -- First, you should probably get familiar with [PyTorch Lightning](https://www.pytorchlightning.ai) -- Next, go through [Hydra quick start guide](https://hydra.cc/docs/intro/) and [basic Hydra tutorial](https://hydra.cc/docs/tutorials/basic/your_first_app/simple_cli/) - -
- -### How It Works - -All PyTorch Lightning modules are dynamically instantiated from module paths specified in config. Example model config: - -```yaml -_target_: src.models.mnist_model.MNISTLitModule -input_size: 784 -lin1_size: 256 -lin2_size: 256 -lin3_size: 256 -output_size: 10 -lr: 0.001 -``` - -Using this config we can instantiate the object with the following line: - -```python -model = hydra.utils.instantiate(config.model) -``` - -This allows you to easily iterate over new models! Every time you create a new one, just specify its module path and parameters in appropriate config file.
- -Switch between models and datamodules with command line arguments: - -```bash -python train.py model=mnist -``` - -The whole pipeline managing the instantiation logic is placed in [src/training_pipeline.py](src/training_pipeline.py). - -
- -### Main Project Configuration - -Location: [configs/train.yaml](configs/train.yaml)
-Main project config contains default training configuration.
-It determines how config is composed when simply executing command `python train.py`.
- -
-Show main project config - -```yaml -# specify here default training configuration -defaults: - - _self_ - - datamodule: mnist.yaml - - model: mnist.yaml - - callbacks: default.yaml - - logger: null # set logger here or use command line (e.g. `python train.py logger=tensorboard`) - - trainer: default.yaml - - log_dir: default.yaml - - # experiment configs allow for version control of specific configurations - # e.g. best hyperparameters for each combination of model and datamodule - - experiment: null - - # debugging config (enable through command line, e.g. `python train.py debug=default) - - debug: null - - # config for hyperparameter optimization - - hparams_search: null - - # optional local config for machine/user specific settings - # it's optional since it doesn't need to exist and is excluded from version control - - optional local: default.yaml - - # enable color logging - - override hydra/hydra_logging: colorlog - - override hydra/job_logging: colorlog - -# path to original working directory -# hydra hijacks working directory by changing it to the new log directory -# https://hydra.cc/docs/next/tutorials/basic/running_your_app/working_directory -original_work_dir: ${hydra:runtime.cwd} - -# path to folder with data -data_dir: ${original_work_dir}/data/ - -# pretty print config at the start of the run using Rich library -print_config: True - -# disable python warnings if they annoy you -ignore_warnings: True - -# set False to skip model training -train: True - -# evaluate on test set, using best model weights achieved during training -# lightning chooses best weights based on the metric specified in checkpoint callback -test: True - -# seed for random number generators in pytorch, numpy and python.random -seed: null - -# default name for the experiment, determines logging folder path -# (you can overwrite this name in experiment configs) -name: "default" -``` - -
- -
- -### Experiment Configuration - -Location: [configs/experiment](configs/experiment)
-Experiment configs allow you to overwrite parameters from main project configuration.
-For example, you can use them to version control best hyperparameters for each combination of model and dataset. - -
-Show example experiment config - -```yaml -# to execute this experiment run: -# python train.py experiment=example - -defaults: - - override /datamodule: mnist.yaml - - override /model: mnist.yaml - - override /callbacks: default.yaml - - override /logger: null - - override /trainer: default.yaml - -# all parameters below will be merged with parameters from default configurations set above -# this allows you to overwrite only specified parameters - -# name of the run determines folder name in logs -name: "simple_dense_net" - -seed: 12345 - -trainer: - min_epochs: 10 - max_epochs: 10 - gradient_clip_val: 0.5 - -model: - lin1_size: 128 - lin2_size: 256 - lin3_size: 64 - lr: 0.002 - -datamodule: - batch_size: 64 - -logger: - wandb: - tags: ["mnist", "${name}"] -``` - -
- -
- -### Local Configuration - -Location: [configs/local](configs/local)
-Some configurations are user/machine/installation specific (e.g. configuration of local cluster, or harddrive paths on a specific machine). For such scenarios, a file `configs/local/default.yaml` can be created which is automatically loaded but not tracked by Git. - -
-Show example local Slurm cluster config - -```yaml -# @package _global_ - -defaults: - - override /hydra/launcher@_here_: submitit_slurm - -data_dir: /mnt/scratch/data/ - -hydra: - launcher: - timeout_min: 1440 - gpus_per_task: 1 - gres: gpu:1 - job: - env_set: - MY_VAR: /home/user/my/system/path - MY_KEY: asdgjhawi8y23ihsghsueity23ihwd -``` - -
- -
- -### Workflow - -1. Write your PyTorch Lightning module (see [models/mnist_module.py](src/models/mnist_module.py) for example) -2. Write your PyTorch Lightning datamodule (see [datamodules/mnist_datamodule.py](src/datamodules/mnist_datamodule.py) for example) -3. Write your experiment config, containing paths to your model and datamodule -4. Run training with chosen experiment config: `python train.py experiment=experiment_name` - -
- -### Logs - -**Hydra creates new working directory for every executed run.** By default, logs have the following structure: - -``` -β”œβ”€β”€ logs -β”‚ β”œβ”€β”€ experiments # Folder for the logs generated by experiments -β”‚ β”‚ β”œβ”€β”€ runs # Folder for single runs -β”‚ β”‚ β”‚ β”œβ”€β”€ experiment_name # Experiment name -β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ YYYY-MM-DD_HH-MM-SS # Datetime of the run -β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ .hydra # Hydra logs -β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ csv # Csv logs -β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ wandb # Weights&Biases logs -β”‚ β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€ checkpoints # Training checkpoints -β”‚ β”‚ β”‚ β”‚ β”‚ └── ... # Any other thing saved during training -β”‚ β”‚ β”‚ β”‚ └── ... -β”‚ β”‚ β”‚ └── ... -β”‚ β”‚ β”‚ -β”‚ β”‚ └── multiruns # Folder for multiruns -β”‚ β”‚ β”œβ”€β”€ experiment_name # Experiment name -β”‚ β”‚ β”‚ β”œβ”€β”€ YYYY-MM-DD_HH-MM-SS # Datetime of the multirun -β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€1 # Multirun job number -β”‚ β”‚ β”‚ β”‚ β”œβ”€β”€2 -β”‚ β”‚ β”‚ β”‚ └── ... -β”‚ β”‚ β”‚ └── ... -β”‚ β”‚ └── ... -β”‚ β”‚ -β”‚ β”œβ”€β”€ evaluations # Folder for the logs generated during testing -β”‚ β”‚ └── ... -β”‚ β”‚ -β”‚ └── debugs # Folder for the logs generated during debugging -β”‚ └── ... -``` - -You can change this structure by modifying paths in [hydra configuration](configs/log_dir). - -
- -### Experiment Tracking - -PyTorch Lightning supports many popular logging frameworks:
-**[Weights&Biases](https://www.wandb.com/) Β· [Neptune](https://neptune.ai/) Β· [Comet](https://www.comet.ml/) Β· [MLFlow](https://mlflow.org) Β· [Tensorboard](https://www.tensorflow.org/tensorboard/)** - -These tools help you keep track of hyperparameters and output metrics and allow you to compare and visualize results. To use one of them simply complete its configuration in [configs/logger](configs/logger) and run: - -```bash -python train.py logger=logger_name -``` - -You can use many of them at once (see [configs/logger/many_loggers.yaml](configs/logger/many_loggers.yaml) for example). - -You can also write your own logger. - -Lightning provides convenient method for logging custom metrics from inside LightningModule. Read the docs [here](https://pytorch-lightning.readthedocs.io/en/latest/extensions/logging.html#automatic-logging) or take a look at [MNIST example](src/models/mnist_module.py). - -
- -### Hyperparameter Search - -Defining hyperparameter optimization is as easy as adding new config file to [configs/hparams_search](configs/hparams_search). - -
-Show example - -```yaml -defaults: - - override /hydra/sweeper: optuna - -# choose metric which will be optimized by Optuna -optimized_metric: "val/acc_best" - -hydra: - # here we define Optuna hyperparameter search - # it optimizes for value returned from function with @hydra.main decorator - # learn more here: https://hydra.cc/docs/next/plugins/optuna_sweeper - sweeper: - _target_: hydra_plugins.hydra_optuna_sweeper.optuna_sweeper.OptunaSweeper - storage: null - study_name: null - n_jobs: 1 - - # 'minimize' or 'maximize' the objective - direction: maximize - - # number of experiments that will be executed - n_trials: 20 - - # choose Optuna hyperparameter sampler - # learn more here: https://optuna.readthedocs.io/en/stable/reference/samplers.html - sampler: - _target_: optuna.samplers.TPESampler - seed: 12345 - consider_prior: true - prior_weight: 1.0 - consider_magic_clip: true - consider_endpoints: false - n_startup_trials: 10 - n_ei_candidates: 24 - multivariate: false - warn_independent_sampling: true - - # define range of hyperparameters - search_space: - datamodule.batch_size: - type: categorical - choices: [32, 64, 128] - model.lr: - type: float - low: 0.0001 - high: 0.2 - model.lin1_size: - type: categorical - choices: [32, 64, 128, 256, 512] - model.lin2_size: - type: categorical - choices: [32, 64, 128, 256, 512] - model.lin3_size: - type: categorical - choices: [32, 64, 128, 256, 512] -``` - -
- -Next, you can execute it with: `python train.py -m hparams_search=mnist_optuna` - -Using this approach doesn't require you to add any boilerplate into your pipeline, everything is defined in a single config file. - -You can use different optimization frameworks integrated with Hydra, like Optuna, Ax or Nevergrad. - -The `optimization_results.yaml` will be available under `logs/multirun` folder. - -This approach doesn't support advanced technics like prunning - for more sophisticated search, you probably shouldn't use hydra multirun feature and instead write your own optimization pipeline. - -
- -### Inference - -The following code is an example of loading model from checkpoint and running predictions.
- -
-Show example - -```python -from PIL import Image -from torchvision import transforms - -from src.models.mnist_module import MNISTLitModule - - -def predict(): - """Example of inference with trained model. - It loads trained image classification model from checkpoint. - Then it loads example image and predicts its label. - """ - - # ckpt can be also a URL! - CKPT_PATH = "last.ckpt" - - # load model from checkpoint - # model __init__ parameters will be loaded from ckpt automatically - # you can also pass some parameter explicitly to override it - trained_model = MNISTLitModule.load_from_checkpoint(checkpoint_path=CKPT_PATH) - - # print model hyperparameters - print(trained_model.hparams) - - # switch to evaluation mode - trained_model.eval() - trained_model.freeze() - - # load data - img = Image.open("data/example_img.png").convert("L") # convert to black and white - # img = Image.open("data/example_img.png").convert("RGB") # convert to RGB - - # preprocess - mnist_transforms = transforms.Compose( - [ - transforms.ToTensor(), - transforms.Resize((28, 28)), - transforms.Normalize((0.1307,), (0.3081,)), - ] - ) - img = mnist_transforms(img) - img = img.reshape((1, *img.size())) # reshape to form batch of size 1 - - # inference - output = trained_model(img) - print(output) - - -if __name__ == "__main__": - predict() - -``` - -
- -
- -### Tests - -Template comes with example tests implemented with pytest library. To execute them simply run: - -```bash -# run all tests -pytest - -# run tests from specific file -pytest tests/shell/test_basic_commands.py - -# run all tests except the ones marked as slow -pytest -k "not slow" -``` - -To speed up the development, you can once in a while execute tests that run a couple of quick experiments, like training 1 epoch on 25% of data, executing single train/val/test step, etc. Those kind of tests don't check for any specific output - they exist to simply verify that executing some bash commands doesn't end up in throwing exceptions. You can find them implemented in [tests/shell](tests/shell) folder. - -You can easily modify the commands in the scripts for your use case. If 1 epoch is too much for your model, then make it run for a couple of batches instead (by using the right trainer flags). - -
- -### Callbacks - -The branch [`wandb-callbacks`](https://github.com/ashleve/lightning-hydra-template/tree/wandb-callbacks) contains example callbacks enabling better Weights&Biases integration, which you can use as a reference for writing your own callbacks (see [wandb_callbacks.py](https://github.com/ashleve/lightning-hydra-template/tree/wandb-callbacks/src/callbacks/wandb_callbacks.py)). - -Callbacks which support reproducibility: - -- **WatchModel** -- **UploadCodeAsArtifact** -- **UploadCheckpointsAsArtifact** - -Callbacks which provide examples of logging custom visualisations: - -- **LogConfusionMatrix** -- **LogF1PrecRecHeatmap** -- **LogImagePredictions** - -To try all of the callbacks at once, switch to the right branch: - -```bash -git checkout wandb-callbacks -``` - -And then run the following command: - -```bash -python train.py logger=wandb callbacks=wandb -``` - -To see the result of all the callbacks attached, take a look at [this experiment dashboard](https://wandb.ai/hobglob/template-tests/runs/3rw7q70h). - -
- -### Multi-GPU Training - -Lightning supports multiple ways of doing distributed training. The most common one is DDP, which spawns separate process for each GPU and averages gradients between them. To learn about other approaches read the [lightning docs](https://pytorch-lightning.readthedocs.io/en/latest/advanced/multi_gpu.html). - -You can run DDP on mnist example with 4 GPUs like this: - -```bash -python train.py trainer.gpus=4 +trainer.strategy=ddp -``` - -⚠️ When using DDP you have to be careful how you write your models - learn more [here](https://pytorch-lightning.readthedocs.io/en/latest/advanced/multi_gpu.html). - -
- -### Docker - -First you will need to [install Nvidia Container Toolkit](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html) to enable GPU support. - -The template Dockerfile is provided on branch [`dockerfiles`](https://github.com/ashleve/lightning-hydra-template/tree/dockerfiles). Copy it to the template root folder. - -To build the container use: - -```bash -docker build -t . -``` - -To mount the project to the container use: - -```bash -docker run -v $(pwd):/workspace/project --gpus all -it --rm -``` - -
- -### Reproducibility - -What provides reproducibility: - -- Hydra manages your configs -- Hydra manages your logging paths and makes every executed run store its hyperparameters and config overrides in a separate file in logs -- Single seed for random number generators in pytorch, numpy and python.random -- LightningDataModule allows you to encapsulate data split, transformations and default parameters in a single, clean abstraction -- LightningModule separates your research code from engineering code in a clean way -- Experiment tracking frameworks take care of logging metrics and hparams, some can also store results and artifacts in cloud -- Pytorch Lightning takes care of creating training checkpoints -- Example callbacks for wandb show how you can save and upload a snapshot of codebase every time the run is executed, as well as upload ckpts and track model gradients - - -
- -### Limitations - -- Currently, template doesn't support k-fold cross validation, but it's possible to achieve it with Lightning Loop interface. See the [official example](https://github.com/PyTorchLightning/pytorch-lightning/blob/master/pl_examples/loop_examples/kfold.py). Implementing it requires rewriting the training pipeline. -- Pytorch Lightning might not be the best choice for scalable reinforcement learning, it's probably better to use something like [Ray](https://github.com/ray-project/ray). -- Currently hyperparameter search with Hydra Optuna Plugin doesn't support prunning. -- Hydra changes working directory to new logging folder for every executed run, which might not be compatible with the way some libraries work. - -
- -## Useful Tricks - -
-Accessing datamodule attributes in model - -1. The simplest way is to pass datamodule attribute directly to model on initialization: - - ```python - # ./src/training_pipeline.py - datamodule = hydra.utils.instantiate(config.datamodule) - model = hydra.utils.instantiate(config.model, some_param=datamodule.some_param) - ``` - - This is not a very robust solution, since it assumes all your datamodules have `some_param` attribute available (otherwise the run will crash). - -2. If you only want to access datamodule config, you can simply pass it as an init parameter: - - ```python - # ./src/training_pipeline.py - model = hydra.utils.instantiate(config.model, dm_conf=config.datamodule, _recursive_=False) - ``` - - Now you can access any datamodule config part like this: - - ```python - # ./src/models/my_model.py - class MyLitModel(LightningModule): - def __init__(self, dm_conf, param1, param2): - super().__init__() - - batch_size = dm_conf.batch_size - ``` - -3. If you need to access the datamodule object attributes, a little hacky solution is to add Omegaconf resolver to your datamodule: - - ```python - # ./src/datamodules/my_datamodule.py - from omegaconf import OmegaConf - - class MyDataModule(LightningDataModule): - def __init__(self, param1, param2): - super().__init__() - - self.param1 = param1 - - resolver_name = "datamodule" - OmegaConf.register_new_resolver( - resolver_name, - lambda name: getattr(self, name), - use_cache=False - ) - ``` - - This way you can reference any datamodule attribute from your config like this: - - ```yaml - # this will return attribute 'param1' from datamodule object - param1: ${datamodule: param1} - ``` - - When later accessing this field, say in your lightning model, it will get automatically resolved based on all resolvers that are registered. Remember not to access this field before datamodule is initialized or it will crash. **You also need to set `resolve=False` in `print_config()` in [train.py](train.py) or it will throw errors:** - - ```python - # ./src/train.py - utils.print_config(config, resolve=False) - ``` - -
- -
-Automatic activation of virtual environment and tab completion when entering folder - -1. Create a new file called `.autoenv` (this name is excluded from version control in `.gitignore`).
- You can use it to automatically execute shell commands when entering folder. Add some commands to your `.autoenv` file, like in the example below: - - ```bash - # activate conda environment - conda activate myenv - - # activate hydra tab completion for bash - eval "$(python train.py -sc install=bash)" - ``` - - (these commands will be executed whenever you're openning or switching terminal to folder containing `.autoenv` file) - -2. To setup this automation for bash, execute the following line (it will append your `.bashrc` file): - - ```bash - echo "autoenv() { if [ -x .autoenv ]; then source .autoenv ; echo '.autoenv executed' ; fi } ; cd() { builtin cd \"\$@\" ; autoenv ; } ; autoenv" >> ~/.bashrc - ``` - -3. Lastly add execution previliges to your `.autoenv` file: - - ``` - chmod +x .autoenv - ``` - - (for safety, only `.autoenv` with previligies will be executed) - -**Explanation** - -The mentioned line appends your `.bashrc` file with 2 commands: - -1. `autoenv() { if [ -x .autoenv ]; then source .autoenv ; echo '.autoenv executed' ; fi }` - this declares the `autoenv()` function, which executes `.autoenv` file if it exists in current work dir and has execution previligies -2. `cd() { builtin cd \"\$@\" ; autoenv ; } ; autoenv` - this extends behaviour of `cd` command, to make it execute `autoenv()` function each time you change folder in terminal or open new terminal - -
- - - -
- -## Best Practices - -
-Use Miniconda for GPU environments - -Use miniconda for your python environments (it's usually unnecessary to install full anaconda environment, miniconda should be enough). -It makes it easier to install some dependencies, like cudatoolkit for GPU support. It also allows you to acccess your environments globally. - -Example installation: - -```bash -wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -bash Miniconda3-latest-Linux-x86_64.sh -``` - -Create new conda environment: - -```bash -conda create -n myenv python=3.8 -conda activate myenv -``` - -
- -
-Use automatic code formatting - -Use pre-commit hooks to standardize code formatting of your project and save mental energy.
-Simply install pre-commit package with: - -```bash -pip install pre-commit -``` - -Next, install hooks from [.pre-commit-config.yaml](.pre-commit-config.yaml): - -```bash -pre-commit install -``` - -After that your code will be automatically reformatted on every new commit.
-Currently template contains configurations of **black** (python code formatting), **isort** (python import sorting), **flake8** (python code analysis), **prettier** (yaml formating) and **nbstripout** (clearing output from jupyter notebooks).
- -To reformat all files in the project use command: - -```bash -pre-commit run -a -``` - -
- -
-Set private environment variables in .env file - -System specific variables (e.g. absolute paths to datasets) should not be under version control or it will result in conflict between different users. Your private keys also shouldn't be versioned since you don't want them to be leaked.
- -Template contains `.env.example` file, which serves as an example. Create a new file called `.env` (this name is excluded from version control in .gitignore). -You should use it for storing environment variables like this: - -``` -MY_VAR=/home/user/my_system_path -``` - -All variables from `.env` are loaded in `train.py` automatically. - -Hydra allows you to reference any env variable in `.yaml` configs like this: - -```yaml -path_to_data: ${oc.env:MY_VAR} -``` - -
- -
-Name metrics using '/' character - -Depending on which logger you're using, it's often useful to define metric name with `/` character: - -```python -self.log("train/loss", loss) -``` - -This way loggers will treat your metrics as belonging to different sections, which helps to get them organised in UI. - -
- -
-Use torchmetrics - -Use official [torchmetrics](https://github.com/PytorchLightning/metrics) library to ensure proper calculation of metrics. This is especially important for multi-GPU training! - -For example, instead of calculating accuracy by yourself, you should use the provided `Accuracy` class like this: - -```python -from torchmetrics.classification.accuracy import Accuracy - - -class LitModel(LightningModule): - def __init__(self) - self.train_acc = Accuracy() - self.val_acc = Accuracy() - - def training_step(self, batch, batch_idx): - ... - acc = self.train_acc(predictions, targets) - self.log("train/acc", acc) - ... - - def validation_step(self, batch, batch_idx): - ... - acc = self.val_acc(predictions, targets) - self.log("val/acc", acc) - ... -``` - -Make sure to use different metric instance for each step to ensure proper value reduction over all GPU processes. - -Torchmetrics provides metrics for most use cases, like F1 score or confusion matrix. Read [documentation](https://torchmetrics.readthedocs.io/en/latest/#more-reading) for more. - -
- -
-Follow PyTorch Lightning style guide - -The style guide is available [here](https://pytorch-lightning.readthedocs.io/en/latest/starter/style_guide.html).
- -1. Be explicit in your init. Try to define all the relevant defaults so that the user doesn’t have to guess. Provide type hints. This way your module is reusable across projects! - - ```python - class LitModel(LightningModule): - def __init__(self, layer_size: int = 256, lr: float = 0.001): - ``` - -2. Preserve the recommended method order. - - ```python - class LitModel(LightningModule): - - def __init__(): - ... - - def forward(): - ... - - def training_step(): - ... - - def training_step_end(): - ... - - def training_epoch_end(): - ... - - def validation_step(): - ... - - def validation_step_end(): - ... - - def validation_epoch_end(): - ... - - def test_step(): - ... - - def test_step_end(): - ... - - def test_epoch_end(): - ... - - def configure_optimizers(): - ... - - def any_extra_hook(): - ... - ``` - -
- -
-Version control your data and models with DVC - -Use [DVC](https://dvc.org) to version control big files, like your data or trained ML models.
-To initialize the dvc repository: - -```bash -dvc init -``` - -To start tracking a file or directory, use `dvc add`: - -```bash -dvc add data/MNIST -``` - -DVC stores information about the added file (or a directory) in a special .dvc file named data/MNIST.dvc, a small text file with a human-readable format. This file can be easily versioned like source code with Git, as a placeholder for the original data: - -```bash -git add data/MNIST.dvc data/.gitignore -git commit -m "Add raw data" -``` - -
- -
-Support installing project as a package - -It allows other people to easily use your modules in their own projects. -Change name of the `src` folder to your project name and add `setup.py` file: - -```python -from setuptools import find_packages, setup - - -setup( - name="src", # change "src" folder name to your project name - version="0.0.0", - description="Describe Your Cool Project", - author="...", - author_email="...", - url="https://github.com/ashleve/lightning-hydra-template", # replace with your own github project link - install_requires=[ - "pytorch>=1.10.0", - "pytorch-lightning>=1.4.0", - "hydra-core>=1.1.0", - ], - packages=find_packages(), -) -``` - -Now your project can be installed from local files: - -```bash -pip install -e . -``` - -Or directly from git repository: - -```bash -pip install git+git://github.com/YourGithubName/your-repo-name.git --upgrade -``` - -So any file can be easily imported into any other file like so: - -```python -from project_name.models.mnist_module import MNISTLitModule -from project_name.datamodules.mnist_datamodule import MNISTDataModule -``` - -
- - - - - -
- -## Other Repositories - -
-Inspirations - -This template was inspired by: -[PyTorchLightning/deep-learninig-project-template](https://github.com/PyTorchLightning/deep-learning-project-template), -[drivendata/cookiecutter-data-science](https://github.com/drivendata/cookiecutter-data-science), -[tchaton/lightning-hydra-seed](https://github.com/tchaton/lightning-hydra-seed), -[Erlemar/pytorch_tempest](https://github.com/Erlemar/pytorch_tempest), -[lucmos/nn-template](https://github.com/lucmos/nn-template). - -
- -
-Useful repositories - -- [pytorch/hydra-torch](https://github.com/pytorch/hydra-torch) - resources for configuring PyTorch classes with Hydra, -- [romesco/hydra-lightning](https://github.com/romesco/hydra-lightning) - resources for configuring PyTorch Lightning classes with Hydra -- [lucmos/nn-template](https://github.com/lucmos/nn-template) - similar template -- [PyTorchLightning/lightning-transformers](https://github.com/PyTorchLightning/lightning-transformers) - official Lightning Transformers repo built with Hydra - -
- - - -
- -## License - -This project is licensed under the MIT License. - -``` -MIT License - -Copyright (c) 2021 ashleve - -Permission is hereby granted, free of charge, to any person obtaining a copy -of this software and associated documentation files (the "Software"), to deal -in the Software without restriction, including without limitation the rights -to use, copy, modify, merge, publish, distribute, sublicense, and/or sell -copies of the Software, and to permit persons to whom the Software is -furnished to do so, subject to the following conditions: - -The above copyright notice and this permission notice shall be included in all -copies or substantial portions of the Software. - -THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR -IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, -FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE -AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER -LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, -OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE -SOFTWARE. -``` - -
-
-
-
- -**DELETE EVERYTHING ABOVE FOR YOUR PROJECT** - ---- - -
- -# Your Project Name - -PyTorch -Lightning -Config: Hydra -Template
-[![Paper](http://img.shields.io/badge/paper-arxiv.1001.2234-B31B1B.svg)](https://www.nature.com/articles/nature14539) -[![Conference](http://img.shields.io/badge/AnyConference-year-4b44ce.svg)](https://papers.nips.cc/paper/2020) - -
- -## Description - -What it does - -## How to run - -Install dependencies - -```bash -# clone project -git clone https://github.com/YourGithubName/your-repo-name -cd your-repo-name - -# [OPTIONAL] create conda environment -conda create -n myenv python=3.8 -conda activate myenv - -# install pytorch according to instructions -# https://pytorch.org/get-started/ - -# install requirements -pip install -r requirements.txt -``` - -Train model with default configuration - -```bash -# train on CPU -python train.py trainer.gpus=0 - -# train on GPU -python train.py trainer.gpus=1 -``` - -Train model with chosen experiment configuration from [configs/experiment/](configs/experiment/) - -```bash -python train.py experiment=experiment_name.yaml -``` - -You can override any parameter from command line like this - -```bash -python train.py trainer.max_epochs=20 datamodule.batch_size=64 -``` diff --git a/models/data/focus/metadata.csv b/models/data/focus/metadata.csv deleted file mode 100644 index 3014ed05ad818414fb7d08f7fc64334013322e88..0000000000000000000000000000000000000000 --- a/models/data/focus/metadata.csv +++ /dev/null @@ -1,571 +0,0 @@ -,image_path,original_filename,study_id,scan_uuid,focus_value,stack_id,obj_name -0,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01631_X013_Y012_Z5107_600_375.jpg,I01631_X013_Y012_Z5107.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.82953,1658220,133 -1,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01632_X013_Y012_Z5175_600_375.jpg,I01632_X013_Y012_Z5175.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.70408,1658220,133 -2,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01633_X013_Y012_Z5722_600_375.jpg,I01633_X013_Y012_Z5722.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.6991799999999997,1658220,133 -3,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01634_X013_Y012_Z5244_600_375.jpg,I01634_X013_Y012_Z5244.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.5026599999999997,1658220,133 -4,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01635_X013_Y012_Z5654_600_375.jpg,I01635_X013_Y012_Z5654.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.3644999999999996,1658220,133 -5,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01636_X013_Y012_Z5312_600_375.jpg,I01636_X013_Y012_Z5312.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.78172,1658220,133 -6,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01637_X013_Y012_Z5585_600_375.jpg,I01637_X013_Y012_Z5585.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.5970499999999999,1658220,133 -7,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01638_X013_Y012_Z5517_600_375.jpg,I01638_X013_Y012_Z5517.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.2322999999999995,1658220,133 -8,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01639_X013_Y012_Z5380_600_375.jpg,I01639_X013_Y012_Z5380.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658220,133 -9,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01640_X013_Y012_Z5449_600_375.jpg,I01640_X013_Y012_Z5449.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.7714100000000004,1658220,133 -10,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01961_X008_Y012_Z4736_600_150.jpg,I01961_X008_Y012_Z4736.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.8167,1647451,13 -11,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01962_X008_Y012_Z4970_600_150.jpg,I01962_X008_Y012_Z4970.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.70088,1647451,13 -12,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01963_X008_Y012_Z5439_600_150.jpg,I01963_X008_Y012_Z5439.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.66089,1647451,13 -13,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01964_X008_Y012_Z5029_600_150.jpg,I01964_X008_Y012_Z5029.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.42352,1647451,13 -14,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01965_X008_Y012_Z5380_600_150.jpg,I01965_X008_Y012_Z5380.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.2935699999999999,1647451,13 -15,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01966_X008_Y012_Z5087_600_150.jpg,I01966_X008_Y012_Z5087.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.9234,1647451,13 -16,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01967_X008_Y012_Z5322_600_150.jpg,I01967_X008_Y012_Z5322.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.8409399999999998,1647451,13 -17,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01968_X008_Y012_Z5146_600_150.jpg,I01968_X008_Y012_Z5146.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.3989099999999999,1647451,13 -18,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01969_X008_Y012_Z5263_600_150.jpg,I01969_X008_Y012_Z5263.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.23914999999999997,1647451,13 -19,31/71a07477-f074-4d0d-96ed-04efe7becf98/I01970_X008_Y012_Z5205_600_150.jpg,I01970_X008_Y012_Z5205.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,0.0,1647451,13 -20,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01981_X014_Y014_Z5058_150_75.jpg,I01981_X014_Y014_Z5058.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.45262,1658248,141 -21,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01982_X014_Y014_Z5673_150_75.jpg,I01982_X014_Y014_Z5673.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.43594,1658248,141 -22,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01983_X014_Y014_Z5126_150_75.jpg,I01983_X014_Y014_Z5126.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.37617,1658248,141 -23,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01984_X014_Y014_Z5605_150_75.jpg,I01984_X014_Y014_Z5605.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.20534,1658248,141 -24,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01985_X014_Y014_Z5195_150_75.jpg,I01985_X014_Y014_Z5195.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.16122,1658248,141 -25,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01986_X014_Y014_Z5537_150_75.jpg,I01986_X014_Y014_Z5537.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.71917,1658248,141 -26,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01987_X014_Y014_Z5263_150_75.jpg,I01987_X014_Y014_Z5263.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.48358,1658248,141 -27,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01988_X014_Y014_Z5468_150_75.jpg,I01988_X014_Y014_Z5468.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.7978399999999999,1658248,141 -28,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01989_X014_Y014_Z5332_150_75.jpg,I01989_X014_Y014_Z5332.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.3852899999999999,1658248,141 -29,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01990_X014_Y014_Z5400_150_75.jpg,I01990_X014_Y014_Z5400.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658248,141 -30,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02631_X016_Y017_Z5195_675_225.jpg,I02631_X016_Y017_Z5195.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.23827,1658297,151 -31,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02632_X016_Y017_Z5810_675_225.jpg,I02632_X016_Y017_Z5810.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.1691000000000003,1658297,151 -32,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02633_X016_Y017_Z5263_675_225.jpg,I02633_X016_Y017_Z5263.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.94231,1658297,151 -33,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02634_X016_Y017_Z5742_675_225.jpg,I02634_X016_Y017_Z5742.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.7410200000000002,1658297,151 -34,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02635_X016_Y017_Z5332_675_225.jpg,I02635_X016_Y017_Z5332.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.3643,1658297,151 -35,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02636_X016_Y017_Z5673_675_225.jpg,I02636_X016_Y017_Z5673.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.09964,1658297,151 -36,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02637_X016_Y017_Z5400_675_225.jpg,I02637_X016_Y017_Z5400.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.5952100000000002,1658297,151 -37,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02638_X016_Y017_Z5605_675_225.jpg,I02638_X016_Y017_Z5605.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.5207000000000002,1658297,151 -38,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02639_X016_Y017_Z5537_675_225.jpg,I02639_X016_Y017_Z5537.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658297,151 -39,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02640_X016_Y017_Z5468_675_225.jpg,I02640_X016_Y017_Z5468.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.08878999999999948,1658297,151 -40,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00091_X017_Y002_Z4453_75_75.jpg,I00091_X017_Y002_Z4453.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-1.7593999999999999,1649053,29 -41,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00092_X017_Y002_Z4511_75_75.jpg,I00092_X017_Y002_Z4511.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-1.63931,1649053,29 -42,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00093_X017_Y002_Z4980_75_75.jpg,I00093_X017_Y002_Z4980.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-1.6298,1649053,29 -43,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00094_X017_Y002_Z4570_75_75.jpg,I00094_X017_Y002_Z4570.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-1.423,1649053,29 -44,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00095_X017_Y002_Z4921_75_75.jpg,I00095_X017_Y002_Z4921.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-1.26513,1649053,29 -45,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00096_X017_Y002_Z4628_75_75.jpg,I00096_X017_Y002_Z4628.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-0.8576600000000001,1649053,29 -46,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00097_X017_Y002_Z4863_75_75.jpg,I00097_X017_Y002_Z4863.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-0.6610499999999999,1649053,29 -47,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00098_X017_Y002_Z4687_75_75.jpg,I00098_X017_Y002_Z4687.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,-0.14305999999999974,1649053,29 -48,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00099_X017_Y002_Z4804_75_75.jpg,I00099_X017_Y002_Z4804.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,0.0,1649053,29 -49,31/6c58be68-7fd0-4d24-8330-48937f7948d6/I00100_X017_Y002_Z4746_75_75.jpg,I00100_X017_Y002_Z4746.jpg,31,6c58be68-7fd0-4d24-8330-48937f7948d6,0.23066999999999993,1649053,29 -50,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00791_X009_Y007_Z5625_525_0.jpg,I00791_X009_Y007_Z5625.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.1693399999999996,1658104,118 -51,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00792_X009_Y007_Z5009_525_0.jpg,I00792_X009_Y007_Z5009.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.1437099999999996,1658104,118 -52,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00793_X009_Y007_Z5078_525_0.jpg,I00793_X009_Y007_Z5078.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.0274699999999997,1658104,118 -53,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00794_X009_Y007_Z5556_525_0.jpg,I00794_X009_Y007_Z5556.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.82508,1658104,118 -54,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00795_X009_Y007_Z5146_525_0.jpg,I00795_X009_Y007_Z5146.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.6482499999999995,1658104,118 -55,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00796_X009_Y007_Z5488_525_0.jpg,I00796_X009_Y007_Z5488.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.1316299999999995,1658104,118 -56,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00797_X009_Y007_Z5214_525_0.jpg,I00797_X009_Y007_Z5214.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.7807299999999997,1658104,118 -57,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00798_X009_Y007_Z5419_525_0.jpg,I00798_X009_Y007_Z5419.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.7802199999999999,1658104,118 -58,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00799_X009_Y007_Z5283_525_0.jpg,I00799_X009_Y007_Z5283.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658104,118 -59,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I00800_X009_Y007_Z5351_525_0.jpg,I00800_X009_Y007_Z5351.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.3536900000000003,1658104,118 -60,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01551_X003_Y011_Z5166_1200_600.jpg,I01551_X003_Y011_Z5166.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.76354,1666247,189 -61,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01552_X003_Y011_Z5234_1200_600.jpg,I01552_X003_Y011_Z5234.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.56835,1666247,189 -62,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01553_X003_Y011_Z5781_1200_600.jpg,I01553_X003_Y011_Z5781.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.51465,1666247,189 -63,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01554_X003_Y011_Z5302_1200_600.jpg,I01554_X003_Y011_Z5302.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.31699,1666247,189 -64,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01555_X003_Y011_Z5712_1200_600.jpg,I01555_X003_Y011_Z5712.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.70039,1666247,189 -65,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01556_X003_Y011_Z5371_1200_600.jpg,I01556_X003_Y011_Z5371.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.13615,1666247,189 -66,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01557_X003_Y011_Z5644_1200_600.jpg,I01557_X003_Y011_Z5644.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.08569,1666247,189 -67,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01558_X003_Y011_Z5576_1200_600.jpg,I01558_X003_Y011_Z5576.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-1.5852199999999996,1666247,189 -68,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01559_X003_Y011_Z5439_1200_600.jpg,I01559_X003_Y011_Z5439.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-1.5729099999999994,1666247,189 -69,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I01560_X003_Y011_Z5507_1200_600.jpg,I01560_X003_Y011_Z5507.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666247,189 -70,31/19d37361-fee4-4382-b29b-fb657f233636/I02691_X005_Y016_Z4941_975_600.jpg,I02691_X005_Y016_Z4941.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.69095,1649634,46 -71,31/19d37361-fee4-4382-b29b-fb657f233636/I02692_X005_Y016_Z5000_975_600.jpg,I02692_X005_Y016_Z5000.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.5047800000000002,1649634,46 -72,31/19d37361-fee4-4382-b29b-fb657f233636/I02693_X005_Y016_Z5468_975_600.jpg,I02693_X005_Y016_Z5468.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.42314,1649634,46 -73,31/19d37361-fee4-4382-b29b-fb657f233636/I02694_X005_Y016_Z5058_975_600.jpg,I02694_X005_Y016_Z5058.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.13315,1649634,46 -74,31/19d37361-fee4-4382-b29b-fb657f233636/I02695_X005_Y016_Z5117_975_600.jpg,I02695_X005_Y016_Z5117.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.54982,1649634,46 -75,31/19d37361-fee4-4382-b29b-fb657f233636/I02696_X005_Y016_Z5410_975_600.jpg,I02696_X005_Y016_Z5410.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.80402,1649634,46 -76,31/19d37361-fee4-4382-b29b-fb657f233636/I02697_X005_Y016_Z5175_975_600.jpg,I02697_X005_Y016_Z5175.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.7725900000000001,1649634,46 -77,31/19d37361-fee4-4382-b29b-fb657f233636/I02698_X005_Y016_Z5234_975_600.jpg,I02698_X005_Y016_Z5234.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.0,1649634,46 -78,31/19d37361-fee4-4382-b29b-fb657f233636/I02699_X005_Y016_Z5351_975_600.jpg,I02699_X005_Y016_Z5351.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.28047999999999984,1649634,46 -79,31/19d37361-fee4-4382-b29b-fb657f233636/I02700_X005_Y016_Z5292_975_600.jpg,I02700_X005_Y016_Z5292.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.4860800000000003,1649634,46 -80,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01861_X024_Y009_Z4648_1200_375.jpg,I01861_X024_Y009_Z4648.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-4.047470000000001,1689926,251 -81,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01862_X024_Y009_Z4785_1200_375.jpg,I01862_X024_Y009_Z4785.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-4.0098400000000005,1689926,251 -82,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01863_X024_Y009_Z4853_1200_375.jpg,I01863_X024_Y009_Z4853.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-3.8872400000000003,1689926,251 -83,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01864_X024_Y009_Z5332_1200_375.jpg,I01864_X024_Y009_Z5332.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-3.67845,1689926,251 -84,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01865_X024_Y009_Z4921_1200_375.jpg,I01865_X024_Y009_Z4921.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-3.4781400000000002,1689926,251 -85,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01866_X024_Y009_Z5263_1200_375.jpg,I01866_X024_Y009_Z5263.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-2.73144,1689926,251 -86,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01867_X024_Y009_Z4990_1200_375.jpg,I01867_X024_Y009_Z4990.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-2.10975,1689926,251 -87,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01868_X024_Y009_Z5195_1200_375.jpg,I01868_X024_Y009_Z5195.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-0.9890300000000005,1689926,251 -88,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01869_X024_Y009_Z5058_1200_375.jpg,I01869_X024_Y009_Z5058.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,0.0,1689926,251 -89,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01870_X024_Y009_Z5126_1200_375.jpg,I01870_X024_Y009_Z5126.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,0.6395,1689926,251 -90,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02281_X012_Y015_Z5048_225_150.jpg,I02281_X012_Y015_Z5048.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.34154,1677250,238 -91,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02282_X012_Y015_Z5117_225_150.jpg,I02282_X012_Y015_Z5117.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.16596,1677250,238 -92,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02283_X012_Y015_Z5664_225_150.jpg,I02283_X012_Y015_Z5664.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.0846599999999995,1677250,238 -93,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02284_X012_Y015_Z5185_225_150.jpg,I02284_X012_Y015_Z5185.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-3.8749700000000002,1677250,238 -94,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02285_X012_Y015_Z5595_225_150.jpg,I02285_X012_Y015_Z5595.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-3.4891300000000003,1677250,238 -95,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02286_X012_Y015_Z5253_225_150.jpg,I02286_X012_Y015_Z5253.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-2.67,1677250,238 -96,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02287_X012_Y015_Z5527_225_150.jpg,I02287_X012_Y015_Z5527.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-2.32064,1677250,238 -97,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02288_X012_Y015_Z5322_225_150.jpg,I02288_X012_Y015_Z5322.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,0.0,1677250,238 -98,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02289_X012_Y015_Z5458_225_150.jpg,I02289_X012_Y015_Z5458.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,0.17380999999999958,1677250,238 -99,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I02290_X012_Y015_Z5390_225_150.jpg,I02290_X012_Y015_Z5390.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,1.8682899999999991,1677250,238 -100,31/19d37361-fee4-4382-b29b-fb657f233636/I02891_X016_Y017_Z5498_150_675.jpg,I02891_X016_Y017_Z5498.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.88138,1649977,49 -101,31/19d37361-fee4-4382-b29b-fb657f233636/I02892_X016_Y017_Z4970_150_675.jpg,I02892_X016_Y017_Z4970.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.7884700000000002,1649977,49 -102,31/19d37361-fee4-4382-b29b-fb657f233636/I02893_X016_Y017_Z5439_150_675.jpg,I02893_X016_Y017_Z5439.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.5311500000000002,1649977,49 -103,31/19d37361-fee4-4382-b29b-fb657f233636/I02894_X016_Y017_Z5029_150_675.jpg,I02894_X016_Y017_Z5029.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-2.5265000000000004,1649977,49 -104,31/19d37361-fee4-4382-b29b-fb657f233636/I02895_X016_Y017_Z5087_150_675.jpg,I02895_X016_Y017_Z5087.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.8943500000000002,1649977,49 -105,31/19d37361-fee4-4382-b29b-fb657f233636/I02896_X016_Y017_Z5380_150_675.jpg,I02896_X016_Y017_Z5380.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.8757300000000003,1649977,49 -106,31/19d37361-fee4-4382-b29b-fb657f233636/I02897_X016_Y017_Z5322_150_675.jpg,I02897_X016_Y017_Z5322.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.9200600000000003,1649977,49 -107,31/19d37361-fee4-4382-b29b-fb657f233636/I02898_X016_Y017_Z5146_150_675.jpg,I02898_X016_Y017_Z5146.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.8478700000000003,1649977,49 -108,31/19d37361-fee4-4382-b29b-fb657f233636/I02899_X016_Y017_Z5205_150_675.jpg,I02899_X016_Y017_Z5205.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.008770000000000167,1649977,49 -109,31/19d37361-fee4-4382-b29b-fb657f233636/I02900_X016_Y017_Z5263_150_675.jpg,I02900_X016_Y017_Z5263.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.0,1649977,49 -110,31/6e592015-f152-43be-aace-5949fb03fd34/I01251_X011_Y011_Z5664_900_225.jpg,I01251_X011_Y011_Z5664.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.82785,1649497,43 -111,31/6e592015-f152-43be-aace-5949fb03fd34/I01252_X011_Y011_Z5048_900_225.jpg,I01252_X011_Y011_Z5048.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.674810000000001,1649497,43 -112,31/6e592015-f152-43be-aace-5949fb03fd34/I01253_X011_Y011_Z5117_900_225.jpg,I01253_X011_Y011_Z5117.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.2755,1649497,43 -113,31/6e592015-f152-43be-aace-5949fb03fd34/I01254_X011_Y011_Z5595_900_225.jpg,I01254_X011_Y011_Z5595.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-4.29871,1649497,43 -114,31/6e592015-f152-43be-aace-5949fb03fd34/I01255_X011_Y011_Z5185_900_225.jpg,I01255_X011_Y011_Z5185.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-4.19003,1649497,43 -115,31/6e592015-f152-43be-aace-5949fb03fd34/I01256_X011_Y011_Z5527_900_225.jpg,I01256_X011_Y011_Z5527.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-2.9557,1649497,43 -116,31/6e592015-f152-43be-aace-5949fb03fd34/I01257_X011_Y011_Z5458_900_225.jpg,I01257_X011_Y011_Z5458.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.9482600000000003,1649497,43 -117,31/6e592015-f152-43be-aace-5949fb03fd34/I01258_X011_Y011_Z5253_900_225.jpg,I01258_X011_Y011_Z5253.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.9432400000000003,1649497,43 -118,31/6e592015-f152-43be-aace-5949fb03fd34/I01259_X011_Y011_Z5390_900_225.jpg,I01259_X011_Y011_Z5390.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.0,1649497,43 -119,31/6e592015-f152-43be-aace-5949fb03fd34/I01260_X011_Y011_Z5322_900_225.jpg,I01260_X011_Y011_Z5322.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.4332199999999995,1649497,43 -120,31/6e592015-f152-43be-aace-5949fb03fd34/I00441_X011_Y005_Z5585_1200_150.jpg,I00441_X011_Y005_Z5585.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-0.6739199999999999,1649491,42 -121,31/6e592015-f152-43be-aace-5949fb03fd34/I00442_X011_Y005_Z5517_1200_150.jpg,I00442_X011_Y005_Z5517.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.0,1649491,42 -122,31/6e592015-f152-43be-aace-5949fb03fd34/I00443_X011_Y005_Z4970_1200_150.jpg,I00443_X011_Y005_Z4970.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.21967999999999988,1649491,42 -123,31/6e592015-f152-43be-aace-5949fb03fd34/I00444_X011_Y005_Z5449_1200_150.jpg,I00444_X011_Y005_Z5449.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,1.3212899999999999,1649491,42 -124,31/6e592015-f152-43be-aace-5949fb03fd34/I00445_X011_Y005_Z5039_1200_150.jpg,I00445_X011_Y005_Z5039.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,1.9182200000000003,1649491,42 -125,31/6e592015-f152-43be-aace-5949fb03fd34/I00446_X011_Y005_Z5380_1200_150.jpg,I00446_X011_Y005_Z5380.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,2.5491500000000005,1649491,42 -126,31/6e592015-f152-43be-aace-5949fb03fd34/I00447_X011_Y005_Z5107_1200_150.jpg,I00447_X011_Y005_Z5107.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,3.73436,1649491,42 -127,31/6e592015-f152-43be-aace-5949fb03fd34/I00448_X011_Y005_Z5312_1200_150.jpg,I00448_X011_Y005_Z5312.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,3.83736,1649491,42 -128,31/6e592015-f152-43be-aace-5949fb03fd34/I00449_X011_Y005_Z5175_1200_150.jpg,I00449_X011_Y005_Z5175.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,5.14658,1649491,42 -129,31/6e592015-f152-43be-aace-5949fb03fd34/I00450_X011_Y005_Z5244_1200_150.jpg,I00450_X011_Y005_Z5244.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,5.192449999999999,1649491,42 -130,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03771_X005_Y024_Z5029_300_525.jpg,I03771_X005_Y024_Z5029.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.8078100000000004,1658012,105 -131,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03772_X005_Y024_Z5234_300_525.jpg,I03772_X005_Y024_Z5234.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.79233,1658012,105 -132,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03773_X005_Y024_Z5302_300_525.jpg,I03773_X005_Y024_Z5302.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.4966800000000005,1658012,105 -133,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03774_X005_Y024_Z5781_300_525.jpg,I03774_X005_Y024_Z5781.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-3.2115400000000003,1658012,105 -134,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03775_X005_Y024_Z5371_300_525.jpg,I03775_X005_Y024_Z5371.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.99358,1658012,105 -135,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03776_X005_Y024_Z5712_300_525.jpg,I03776_X005_Y024_Z5712.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.2270600000000003,1658012,105 -136,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03777_X005_Y024_Z5439_300_525.jpg,I03777_X005_Y024_Z5439.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.5542800000000003,1658012,105 -137,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03778_X005_Y024_Z5644_300_525.jpg,I03778_X005_Y024_Z5644.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.8657500000000002,1658012,105 -138,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03779_X005_Y024_Z5507_300_525.jpg,I03779_X005_Y024_Z5507.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.2804500000000001,1658012,105 -139,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03780_X005_Y024_Z5576_300_525.jpg,I03780_X005_Y024_Z5576.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658012,105 -140,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04611_X018_Y026_Z5888_1050_150.jpg,I04611_X018_Y026_Z5888.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-6.04539,1663070,181 -141,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04612_X018_Y026_Z5273_1050_150.jpg,I04612_X018_Y026_Z5273.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-5.90599,1663070,181 -142,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04613_X018_Y026_Z5820_1050_150.jpg,I04613_X018_Y026_Z5820.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-5.7744,1663070,181 -143,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04614_X018_Y026_Z5341_1050_150.jpg,I04614_X018_Y026_Z5341.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-5.6486,1663070,181 -144,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04615_X018_Y026_Z5751_1050_150.jpg,I04615_X018_Y026_Z5751.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-5.07846,1663070,181 -145,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04616_X018_Y026_Z5410_1050_150.jpg,I04616_X018_Y026_Z5410.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-4.761290000000001,1663070,181 -146,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04617_X018_Y026_Z5683_1050_150.jpg,I04617_X018_Y026_Z5683.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-3.7954700000000003,1663070,181 -147,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04618_X018_Y026_Z5478_1050_150.jpg,I04618_X018_Y026_Z5478.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-1.84544,1663070,181 -148,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04619_X018_Y026_Z5615_1050_150.jpg,I04619_X018_Y026_Z5615.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-1.5001100000000003,1663070,181 -149,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I04620_X018_Y026_Z5546_1050_150.jpg,I04620_X018_Y026_Z5546.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,0.0,1663070,181 -150,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02221_X006_Y015_Z4970_750_600.jpg,I02221_X006_Y015_Z4970.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.5894400000000002,1661046,154 -151,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02222_X006_Y015_Z5029_750_600.jpg,I02222_X006_Y015_Z5029.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.5204900000000001,1661046,154 -152,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02223_X006_Y015_Z5498_750_600.jpg,I02223_X006_Y015_Z5498.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.4110500000000001,1661046,154 -153,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02224_X006_Y015_Z5087_750_600.jpg,I02224_X006_Y015_Z5087.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.37813,1661046,154 -154,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02225_X006_Y015_Z5439_750_600.jpg,I02225_X006_Y015_Z5439.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.0470300000000003,1661046,154 -155,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02226_X006_Y015_Z5146_750_600.jpg,I02226_X006_Y015_Z5146.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.9025600000000003,1661046,154 -156,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02227_X006_Y015_Z5380_750_600.jpg,I02227_X006_Y015_Z5380.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.5114800000000002,1661046,154 -157,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02228_X006_Y015_Z5205_750_600.jpg,I02228_X006_Y015_Z5205.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.2644200000000003,1661046,154 -158,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02229_X006_Y015_Z5322_750_600.jpg,I02229_X006_Y015_Z5322.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,0.0,1661046,154 -159,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02230_X006_Y015_Z5263_750_600.jpg,I02230_X006_Y015_Z5263.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,0.11827999999999994,1661046,154 -160,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01191_X007_Y007_Z4755_525_75.jpg,I01191_X007_Y007_Z4755.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-5.82597,1689205,243 -161,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01192_X007_Y007_Z4892_525_75.jpg,I01192_X007_Y007_Z4892.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-5.80757,1689205,243 -162,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01193_X007_Y007_Z4960_525_75.jpg,I01193_X007_Y007_Z4960.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-5.5717799999999995,1689205,243 -163,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01194_X007_Y007_Z5439_525_75.jpg,I01194_X007_Y007_Z5439.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-5.500929999999999,1689205,243 -164,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01195_X007_Y007_Z5029_525_75.jpg,I01195_X007_Y007_Z5029.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-4.94757,1689205,243 -165,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01196_X007_Y007_Z5371_525_75.jpg,I01196_X007_Y007_Z5371.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-4.39333,1689205,243 -166,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01197_X007_Y007_Z5097_525_75.jpg,I01197_X007_Y007_Z5097.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-2.84908,1689205,243 -167,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01198_X007_Y007_Z5302_525_75.jpg,I01198_X007_Y007_Z5302.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-2.0695999999999994,1689205,243 -168,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01199_X007_Y007_Z5166_525_75.jpg,I01199_X007_Y007_Z5166.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,-0.3270199999999992,1689205,243 -169,31/9faa9262-1268-4de1-afb0-d38f9eb3b692/I01200_X007_Y007_Z5234_525_75.jpg,I01200_X007_Y007_Z5234.jpg,31,9faa9262-1268-4de1-afb0-d38f9eb3b692,0.0,1689205,243 -170,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05051_X014_Y026_Z5673_525_0.jpg,I05051_X014_Y026_Z5673.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.53444,1647674,20 -171,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05052_X014_Y026_Z5146_525_0.jpg,I05052_X014_Y026_Z5146.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.45405,1647674,20 -172,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05053_X014_Y026_Z5615_525_0.jpg,I05053_X014_Y026_Z5615.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.3418,1647674,20 -173,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05054_X014_Y026_Z5205_525_0.jpg,I05054_X014_Y026_Z5205.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.25067,1647674,20 -174,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05055_X014_Y026_Z5556_525_0.jpg,I05055_X014_Y026_Z5556.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.9882900000000001,1647674,20 -175,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05056_X014_Y026_Z5263_525_0.jpg,I05056_X014_Y026_Z5263.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.95303,1647674,20 -176,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05057_X014_Y026_Z5322_525_0.jpg,I05057_X014_Y026_Z5322.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.48587999999999987,1647674,20 -177,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05058_X014_Y026_Z5498_525_0.jpg,I05058_X014_Y026_Z5498.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.44835,1647674,20 -178,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05059_X014_Y026_Z5380_525_0.jpg,I05059_X014_Y026_Z5380.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.021560000000000024,1647674,20 -179,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05060_X014_Y026_Z5439_525_0.jpg,I05060_X014_Y026_Z5439.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,0.0,1647674,20 -180,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05051_X006_Y028_Z5273_300_450.jpg,I05051_X006_Y028_Z5273.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-3.85633,1662708,172 -181,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05052_X006_Y028_Z5888_300_450.jpg,I05052_X006_Y028_Z5888.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-3.7541199999999995,1662708,172 -182,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05053_X006_Y028_Z5341_300_450.jpg,I05053_X006_Y028_Z5341.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-3.67863,1662708,172 -183,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05054_X006_Y028_Z5410_300_450.jpg,I05054_X006_Y028_Z5410.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-3.3142199999999997,1662708,172 -184,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05055_X006_Y028_Z5820_300_450.jpg,I05055_X006_Y028_Z5820.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-3.2557099999999997,1662708,172 -185,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05056_X006_Y028_Z5478_300_450.jpg,I05056_X006_Y028_Z5478.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-2.29783,1662708,172 -186,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05057_X006_Y028_Z5751_300_450.jpg,I05057_X006_Y028_Z5751.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-2.2071899999999998,1662708,172 -187,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05058_X006_Y028_Z5683_300_450.jpg,I05058_X006_Y028_Z5683.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,-0.6894399999999994,1662708,172 -188,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05059_X006_Y028_Z5546_300_450.jpg,I05059_X006_Y028_Z5546.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,0.0,1662708,172 -189,31/d340cee7-10b0-4a4a-bb66-6724528574a0/I05060_X006_Y028_Z5615_300_450.jpg,I05060_X006_Y028_Z5615.jpg,31,d340cee7-10b0-4a4a-bb66-6724528574a0,1.1080800000000002,1662708,172 -190,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05281_X012_Y027_Z5605_750_450.jpg,I05281_X012_Y027_Z5605.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-2.24207,1647605,17 -191,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05282_X012_Y027_Z5078_750_450.jpg,I05282_X012_Y027_Z5078.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-2.2308,1647605,17 -192,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05283_X012_Y027_Z5546_750_450.jpg,I05283_X012_Y027_Z5546.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.8912,1647605,17 -193,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05284_X012_Y027_Z5136_750_450.jpg,I05284_X012_Y027_Z5136.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.5498399999999997,1647605,17 -194,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05285_X012_Y027_Z5488_750_450.jpg,I05285_X012_Y027_Z5488.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-1.28504,1647605,17 -195,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05286_X012_Y027_Z5195_750_450.jpg,I05286_X012_Y027_Z5195.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.7591799999999997,1647605,17 -196,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05287_X012_Y027_Z5429_750_450.jpg,I05287_X012_Y027_Z5429.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,-0.35535000000000005,1647605,17 -197,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05288_X012_Y027_Z5253_750_450.jpg,I05288_X012_Y027_Z5253.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,0.0,1647605,17 -198,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05289_X012_Y027_Z5371_750_450.jpg,I05289_X012_Y027_Z5371.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,0.25980000000000025,1647605,17 -199,31/71a07477-f074-4d0d-96ed-04efe7becf98/I05290_X012_Y027_Z5312_750_450.jpg,I05290_X012_Y027_Z5312.jpg,31,71a07477-f074-4d0d-96ed-04efe7becf98,0.33237000000000005,1647605,17 -200,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03221_X005_Y021_Z5839_900_375.jpg,I03221_X005_Y021_Z5839.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.76335,1658009,102 -201,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03222_X005_Y021_Z5224_900_375.jpg,I03222_X005_Y021_Z5224.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.7464600000000001,1658009,102 -202,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03223_X005_Y021_Z5292_900_375.jpg,I03223_X005_Y021_Z5292.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.51606,1658009,102 -203,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03224_X005_Y021_Z5771_900_375.jpg,I03224_X005_Y021_Z5771.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.3718599999999999,1658009,102 -204,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03225_X005_Y021_Z5361_900_375.jpg,I03225_X005_Y021_Z5361.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.9329300000000003,1658009,102 -205,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03226_X005_Y021_Z5703_900_375.jpg,I03226_X005_Y021_Z5703.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.6175899999999999,1658009,102 -206,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03227_X005_Y021_Z5429_900_375.jpg,I03227_X005_Y021_Z5429.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658009,102 -207,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03228_X005_Y021_Z5634_900_375.jpg,I03228_X005_Y021_Z5634.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.17383999999999977,1658009,102 -208,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03229_X005_Y021_Z5566_900_375.jpg,I03229_X005_Y021_Z5566.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.9256199999999999,1658009,102 -209,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03230_X005_Y021_Z5498_900_375.jpg,I03230_X005_Y021_Z5498.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.9784299999999999,1658009,102 -210,31/6e592015-f152-43be-aace-5949fb03fd34/I00381_X005_Y005_Z4980_825_600.jpg,I00381_X005_Y005_Z4980.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.568530000000001,1649367,34 -211,31/6e592015-f152-43be-aace-5949fb03fd34/I00382_X005_Y005_Z5595_825_600.jpg,I00382_X005_Y005_Z5595.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.45105,1649367,34 -212,31/6e592015-f152-43be-aace-5949fb03fd34/I00383_X005_Y005_Z5048_825_600.jpg,I00383_X005_Y005_Z5048.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-4.93506,1649367,34 -213,31/6e592015-f152-43be-aace-5949fb03fd34/I00384_X005_Y005_Z5527_825_600.jpg,I00384_X005_Y005_Z5527.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-4.412840000000001,1649367,34 -214,31/6e592015-f152-43be-aace-5949fb03fd34/I00385_X005_Y005_Z5117_825_600.jpg,I00385_X005_Y005_Z5117.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-3.7957,1649367,34 -215,31/6e592015-f152-43be-aace-5949fb03fd34/I00386_X005_Y005_Z5458_825_600.jpg,I00386_X005_Y005_Z5458.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-2.9582500000000005,1649367,34 -216,31/6e592015-f152-43be-aace-5949fb03fd34/I00387_X005_Y005_Z5390_825_600.jpg,I00387_X005_Y005_Z5390.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.9122500000000002,1649367,34 -217,31/6e592015-f152-43be-aace-5949fb03fd34/I00388_X005_Y005_Z5185_825_600.jpg,I00388_X005_Y005_Z5185.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.6961000000000004,1649367,34 -218,31/6e592015-f152-43be-aace-5949fb03fd34/I00389_X005_Y005_Z5322_825_600.jpg,I00389_X005_Y005_Z5322.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.0,1649367,34 -219,31/6e592015-f152-43be-aace-5949fb03fd34/I00390_X005_Y005_Z5253_825_600.jpg,I00390_X005_Y005_Z5253.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.3178200000000002,1649367,34 -220,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03881_X009_Y024_Z5244_150_0.jpg,I03881_X009_Y024_Z5244.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8805400000000001,1661140,157 -221,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03882_X009_Y024_Z5771_150_0.jpg,I03882_X009_Y024_Z5771.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8165000000000002,1661140,157 -222,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03883_X009_Y024_Z5302_150_0.jpg,I03883_X009_Y024_Z5302.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8111000000000002,1661140,157 -223,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03884_X009_Y024_Z5361_150_0.jpg,I03884_X009_Y024_Z5361.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.6547000000000003,1661140,157 -224,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03885_X009_Y024_Z5712_150_0.jpg,I03885_X009_Y024_Z5712.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.5956100000000002,1661140,157 -225,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03886_X009_Y024_Z5654_150_0.jpg,I03886_X009_Y024_Z5654.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.3885700000000001,1661140,157 -226,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03887_X009_Y024_Z5419_150_0.jpg,I03887_X009_Y024_Z5419.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.3884300000000003,1661140,157 -227,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03888_X009_Y024_Z5595_150_0.jpg,I03888_X009_Y024_Z5595.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.13427000000000033,1661140,157 -228,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03889_X009_Y024_Z5478_150_0.jpg,I03889_X009_Y024_Z5478.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.13366999999999996,1661140,157 -229,31/274d8969-aa7c-4ac0-be60-e753579393ad/I03890_X009_Y024_Z5537_150_0.jpg,I03890_X009_Y024_Z5537.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,0.0,1661140,157 -230,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03801_X003_Y021_Z5166_600_525.jpg,I03801_X003_Y021_Z5166.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-6.318910000000001,1653738,67 -231,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03802_X003_Y021_Z5234_600_525.jpg,I03802_X003_Y021_Z5234.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-6.15174,1653738,67 -232,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03803_X003_Y021_Z5781_600_525.jpg,I03803_X003_Y021_Z5781.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-6.142900000000001,1653738,67 -233,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03804_X003_Y021_Z5302_600_525.jpg,I03804_X003_Y021_Z5302.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.752790000000001,1653738,67 -234,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03805_X003_Y021_Z5712_600_525.jpg,I03805_X003_Y021_Z5712.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.548690000000001,1653738,67 -235,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03806_X003_Y021_Z5644_600_525.jpg,I03806_X003_Y021_Z5644.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.38612,1653738,67 -236,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03807_X003_Y021_Z5371_600_525.jpg,I03807_X003_Y021_Z5371.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.277840000000001,1653738,67 -237,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03808_X003_Y021_Z5576_600_525.jpg,I03808_X003_Y021_Z5576.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-2.1588700000000003,1653738,67 -238,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03809_X003_Y021_Z5439_600_525.jpg,I03809_X003_Y021_Z5439.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-1.1545700000000005,1653738,67 -239,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03810_X003_Y021_Z5507_600_525.jpg,I03810_X003_Y021_Z5507.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,0.0,1653738,67 -240,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06191_X020_Y030_Z5478_450_150.jpg,I06191_X020_Y030_Z5478.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-1.6979400000000002,1657822,89 -241,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06192_X020_Y030_Z5419_450_150.jpg,I06192_X020_Y030_Z5419.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-1.44022,1657822,89 -242,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06193_X020_Y030_Z5361_450_150.jpg,I06193_X020_Y030_Z5361.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.9538600000000002,1657822,89 -243,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06194_X020_Y030_Z5302_450_150.jpg,I06194_X020_Y030_Z5302.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.49914000000000014,1657822,89 -244,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06195_X020_Y030_Z4951_450_150.jpg,I06195_X020_Y030_Z4951.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.17367,1657822,89 -245,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06196_X020_Y030_Z5009_450_150.jpg,I06196_X020_Y030_Z5009.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.16887000000000008,1657822,89 -246,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06197_X020_Y030_Z5068_450_150.jpg,I06197_X020_Y030_Z5068.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.16425,1657822,89 -247,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06198_X020_Y030_Z5244_450_150.jpg,I06198_X020_Y030_Z5244.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.1229300000000002,1657822,89 -248,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06199_X020_Y030_Z5126_450_150.jpg,I06199_X020_Y030_Z5126.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,-0.09434000000000031,1657822,89 -249,31/02061a59-9897-436e-a9ec-8a8937ab2e15/I06200_X020_Y030_Z5185_450_150.jpg,I06200_X020_Y030_Z5185.jpg,31,02061a59-9897-436e-a9ec-8a8937ab2e15,0.0,1657822,89 -250,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03971_X020_Y021_Z5078_975_450.jpg,I03971_X020_Y021_Z5078.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.06038,1654261,82 -251,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03972_X020_Y021_Z5693_975_450.jpg,I03972_X020_Y021_Z5693.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.93449,1654261,82 -252,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03973_X020_Y021_Z5146_975_450.jpg,I03973_X020_Y021_Z5146.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.805540000000001,1654261,82 -253,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03974_X020_Y021_Z5625_975_450.jpg,I03974_X020_Y021_Z5625.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.244400000000001,1654261,82 -254,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03975_X020_Y021_Z5214_975_450.jpg,I03975_X020_Y021_Z5214.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.2369,1654261,82 -255,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03976_X020_Y021_Z5556_975_450.jpg,I03976_X020_Y021_Z5556.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-2.89733,1654261,82 -256,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03977_X020_Y021_Z5283_975_450.jpg,I03977_X020_Y021_Z5283.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-2.49329,1654261,82 -257,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03978_X020_Y021_Z5488_975_450.jpg,I03978_X020_Y021_Z5488.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-0.7850400000000004,1654261,82 -258,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03979_X020_Y021_Z5351_975_450.jpg,I03979_X020_Y021_Z5351.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,0.0,1654261,82 -259,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03980_X020_Y021_Z5419_975_450.jpg,I03980_X020_Y021_Z5419.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,0.6877999999999993,1654261,82 -260,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03971_X020_Y021_Z5078_975_600.jpg,I03971_X020_Y021_Z5078.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.74818,1654261,82 -261,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03972_X020_Y021_Z5693_975_600.jpg,I03972_X020_Y021_Z5693.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.62229,1654261,82 -262,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03973_X020_Y021_Z5146_975_600.jpg,I03973_X020_Y021_Z5146.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.49334,1654261,82 -263,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03974_X020_Y021_Z5625_975_600.jpg,I03974_X020_Y021_Z5625.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.9322,1654261,82 -264,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03975_X020_Y021_Z5214_975_600.jpg,I03975_X020_Y021_Z5214.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.9247,1654261,82 -265,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03976_X020_Y021_Z5556_975_600.jpg,I03976_X020_Y021_Z5556.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-3.5851299999999995,1654261,82 -266,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03977_X020_Y021_Z5283_975_600.jpg,I03977_X020_Y021_Z5283.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-3.1810899999999993,1654261,82 -267,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03978_X020_Y021_Z5488_975_600.jpg,I03978_X020_Y021_Z5488.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-1.4728399999999997,1654261,82 -268,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03979_X020_Y021_Z5351_975_600.jpg,I03979_X020_Y021_Z5351.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-0.6877999999999993,1654261,82 -269,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I03980_X020_Y021_Z5419_975_600.jpg,I03980_X020_Y021_Z5419.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,0.0,1654261,82 -270,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03571_X008_Y023_Z5351_750_450.jpg,I03571_X008_Y023_Z5351.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.5553299999999999,1658092,115 -271,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03572_X008_Y023_Z5966_750_450.jpg,I03572_X008_Y023_Z5966.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.4767599999999996,1658092,115 -272,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03573_X008_Y023_Z5419_750_450.jpg,I03573_X008_Y023_Z5419.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.9444899999999996,1658092,115 -273,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03574_X008_Y023_Z5898_750_450.jpg,I03574_X008_Y023_Z5898.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.79358,1658092,115 -274,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03575_X008_Y023_Z5488_750_450.jpg,I03575_X008_Y023_Z5488.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658092,115 -275,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03576_X008_Y023_Z5830_750_450.jpg,I03576_X008_Y023_Z5830.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.11581000000000019,1658092,115 -276,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03577_X008_Y023_Z5556_750_450.jpg,I03577_X008_Y023_Z5556.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.7383899999999999,1658092,115 -277,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03578_X008_Y023_Z5625_750_450.jpg,I03578_X008_Y023_Z5625.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.8902600000000005,1658092,115 -278,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03579_X008_Y023_Z5761_750_450.jpg,I03579_X008_Y023_Z5761.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,1.0994700000000002,1658092,115 -279,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I03580_X008_Y023_Z5693_750_450.jpg,I03580_X008_Y023_Z5693.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,1.1608100000000006,1658092,115 -280,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03161_X005_Y018_Z5341_0_675.jpg,I03161_X005_Y018_Z5341.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.3557,1666305,199 -281,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03162_X005_Y018_Z5957_0_675.jpg,I03162_X005_Y018_Z5957.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.3423,1666305,199 -282,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03163_X005_Y018_Z5410_0_675.jpg,I03163_X005_Y018_Z5410.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.0997900000000005,1666305,199 -283,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03164_X005_Y018_Z5888_0_675.jpg,I03164_X005_Y018_Z5888.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.94116,1666305,199 -284,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03165_X005_Y018_Z5820_0_675.jpg,I03165_X005_Y018_Z5820.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.19355,1666305,199 -285,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03166_X005_Y018_Z5478_0_675.jpg,I03166_X005_Y018_Z5478.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.17094,1666305,199 -286,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03167_X005_Y018_Z5751_0_675.jpg,I03167_X005_Y018_Z5751.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-2.3911999999999995,1666305,199 -287,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03168_X005_Y018_Z5546_0_675.jpg,I03168_X005_Y018_Z5546.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-1.8792600000000004,1666305,199 -288,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03169_X005_Y018_Z5683_0_675.jpg,I03169_X005_Y018_Z5683.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-0.6559600000000003,1666305,199 -289,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03170_X005_Y018_Z5615_0_675.jpg,I03170_X005_Y018_Z5615.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666305,199 -290,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02091_X003_Y014_Z5126_450_0.jpg,I02091_X003_Y014_Z5126.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.7188700000000003,1657955,93 -291,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02092_X003_Y014_Z5195_450_0.jpg,I02092_X003_Y014_Z5195.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.5748400000000005,1657955,93 -292,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02093_X003_Y014_Z5742_450_0.jpg,I02093_X003_Y014_Z5742.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.4312300000000002,1657955,93 -293,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02094_X003_Y014_Z5263_450_0.jpg,I02094_X003_Y014_Z5263.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.2130400000000003,1657955,93 -294,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02095_X003_Y014_Z5673_450_0.jpg,I02095_X003_Y014_Z5673.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.7181200000000003,1657955,93 -295,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02096_X003_Y014_Z5332_450_0.jpg,I02096_X003_Y014_Z5332.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.4144700000000006,1657955,93 -296,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02097_X003_Y014_Z5605_450_0.jpg,I02097_X003_Y014_Z5605.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.6838000000000002,1657955,93 -297,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02098_X003_Y014_Z5537_450_0.jpg,I02098_X003_Y014_Z5537.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.016170000000000684,1657955,93 -298,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02099_X003_Y014_Z5400_450_0.jpg,I02099_X003_Y014_Z5400.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1657955,93 -299,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02100_X003_Y014_Z5468_450_0.jpg,I02100_X003_Y014_Z5468.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.4722899999999992,1657955,93 -300,31/6e592015-f152-43be-aace-5949fb03fd34/I00251_X011_Y004_Z4882_375_150.jpg,I00251_X011_Y004_Z4882.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-2.0642199999999997,1649490,41 -301,31/6e592015-f152-43be-aace-5949fb03fd34/I00252_X011_Y004_Z4951_375_150.jpg,I00252_X011_Y004_Z4951.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.8673600000000001,1649490,41 -302,31/6e592015-f152-43be-aace-5949fb03fd34/I00253_X011_Y004_Z5498_375_150.jpg,I00253_X011_Y004_Z5498.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.6457799999999998,1649490,41 -303,31/6e592015-f152-43be-aace-5949fb03fd34/I00254_X011_Y004_Z5019_375_150.jpg,I00254_X011_Y004_Z5019.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.3883999999999999,1649490,41 -304,31/6e592015-f152-43be-aace-5949fb03fd34/I00255_X011_Y004_Z5429_375_150.jpg,I00255_X011_Y004_Z5429.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-0.8519399999999999,1649490,41 -305,31/6e592015-f152-43be-aace-5949fb03fd34/I00256_X011_Y004_Z5087_375_150.jpg,I00256_X011_Y004_Z5087.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-0.13412999999999986,1649490,41 -306,31/6e592015-f152-43be-aace-5949fb03fd34/I00257_X011_Y004_Z5361_375_150.jpg,I00257_X011_Y004_Z5361.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.0,1649490,41 -307,31/6e592015-f152-43be-aace-5949fb03fd34/I00258_X011_Y004_Z5292_375_150.jpg,I00258_X011_Y004_Z5292.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,1.9309399999999997,1649490,41 -308,31/6e592015-f152-43be-aace-5949fb03fd34/I00259_X011_Y004_Z5156_375_150.jpg,I00259_X011_Y004_Z5156.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,1.97227,1649490,41 -309,31/6e592015-f152-43be-aace-5949fb03fd34/I00260_X011_Y004_Z5224_375_150.jpg,I00260_X011_Y004_Z5224.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,3.09396,1649490,41 -310,31/08549899-f59e-4139-93be-71e91ba83c24/I06311_X013_Y033_Z5156_150_300.jpg,I06311_X013_Y033_Z5156.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-3.0658399999999997,1652774,61 -311,31/08549899-f59e-4139-93be-71e91ba83c24/I06312_X013_Y033_Z5771_150_300.jpg,I06312_X013_Y033_Z5771.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-3.0549099999999996,1652774,61 -312,31/08549899-f59e-4139-93be-71e91ba83c24/I06313_X013_Y033_Z5224_150_300.jpg,I06313_X013_Y033_Z5224.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-2.97239,1652774,61 -313,31/08549899-f59e-4139-93be-71e91ba83c24/I06314_X013_Y033_Z5292_150_300.jpg,I06314_X013_Y033_Z5292.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-2.83464,1652774,61 -314,31/08549899-f59e-4139-93be-71e91ba83c24/I06315_X013_Y033_Z5361_150_300.jpg,I06315_X013_Y033_Z5361.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-2.44279,1652774,61 -315,31/08549899-f59e-4139-93be-71e91ba83c24/I06316_X013_Y033_Z5703_150_300.jpg,I06316_X013_Y033_Z5703.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-2.3220699999999996,1652774,61 -316,31/08549899-f59e-4139-93be-71e91ba83c24/I06317_X013_Y033_Z5634_150_300.jpg,I06317_X013_Y033_Z5634.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-1.2824999999999998,1652774,61 -317,31/08549899-f59e-4139-93be-71e91ba83c24/I06318_X013_Y033_Z5429_150_300.jpg,I06318_X013_Y033_Z5429.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,-1.1824499999999998,1652774,61 -318,31/08549899-f59e-4139-93be-71e91ba83c24/I06319_X013_Y033_Z5566_150_300.jpg,I06319_X013_Y033_Z5566.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,0.0,1652774,61 -319,31/08549899-f59e-4139-93be-71e91ba83c24/I06320_X013_Y033_Z5498_150_300.jpg,I06320_X013_Y033_Z5498.jpg,31,08549899-f59e-4139-93be-71e91ba83c24,0.27627999999999986,1652774,61 -320,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01631_X013_Y012_Z5107_750_300.jpg,I01631_X013_Y012_Z5107.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.82953,1658220,132 -321,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01632_X013_Y012_Z5175_750_300.jpg,I01632_X013_Y012_Z5175.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.70408,1658220,132 -322,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01633_X013_Y012_Z5722_750_300.jpg,I01633_X013_Y012_Z5722.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.6991799999999997,1658220,132 -323,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01634_X013_Y012_Z5244_750_300.jpg,I01634_X013_Y012_Z5244.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.5026599999999997,1658220,132 -324,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01635_X013_Y012_Z5654_750_300.jpg,I01635_X013_Y012_Z5654.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.3644999999999996,1658220,132 -325,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01636_X013_Y012_Z5312_750_300.jpg,I01636_X013_Y012_Z5312.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.78172,1658220,132 -326,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01637_X013_Y012_Z5585_750_300.jpg,I01637_X013_Y012_Z5585.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.5970499999999999,1658220,132 -327,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01638_X013_Y012_Z5517_750_300.jpg,I01638_X013_Y012_Z5517.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.2322999999999995,1658220,132 -328,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01639_X013_Y012_Z5380_750_300.jpg,I01639_X013_Y012_Z5380.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658220,132 -329,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01640_X013_Y012_Z5449_750_300.jpg,I01640_X013_Y012_Z5449.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.7714100000000004,1658220,132 -330,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00791_X004_Y007_Z5058_300_0.jpg,I00791_X004_Y007_Z5058.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.117609999999999,1666267,194 -331,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00792_X004_Y007_Z5673_300_0.jpg,I00792_X004_Y007_Z5673.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.9395999999999995,1666267,194 -332,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00793_X004_Y007_Z5126_300_0.jpg,I00793_X004_Y007_Z5126.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.9111199999999995,1666267,194 -333,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00794_X004_Y007_Z5195_300_0.jpg,I00794_X004_Y007_Z5195.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.3923,1666267,194 -334,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00795_X004_Y007_Z5605_300_0.jpg,I00795_X004_Y007_Z5605.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.22054,1666267,194 -335,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00796_X004_Y007_Z5263_300_0.jpg,I00796_X004_Y007_Z5263.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.9565599999999996,1666267,194 -336,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00797_X004_Y007_Z5537_300_0.jpg,I00797_X004_Y007_Z5537.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.8842799999999995,1666267,194 -337,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00798_X004_Y007_Z5468_300_0.jpg,I00798_X004_Y007_Z5468.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-2.0984999999999996,1666267,194 -338,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00799_X004_Y007_Z5332_300_0.jpg,I00799_X004_Y007_Z5332.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-0.7643899999999997,1666267,194 -339,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I00800_X004_Y007_Z5400_300_0.jpg,I00800_X004_Y007_Z5400.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666267,194 -340,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02371_X018_Y016_Z4882_525_525.jpg,I02371_X018_Y016_Z4882.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.0811899999999999,1661361,163 -341,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02372_X018_Y016_Z5000_525_525.jpg,I02372_X018_Y016_Z5000.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.0489099999999998,1661361,163 -342,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02373_X018_Y016_Z5058_525_525.jpg,I02373_X018_Y016_Z5058.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.9565499999999998,1661361,163 -343,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02374_X018_Y016_Z5468_525_525.jpg,I02374_X018_Y016_Z5468.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8777499999999998,1661361,163 -344,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02375_X018_Y016_Z5117_525_525.jpg,I02375_X018_Y016_Z5117.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.7010199999999998,1661361,163 -345,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02376_X018_Y016_Z5410_525_525.jpg,I02376_X018_Y016_Z5410.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.5387799999999996,1661361,163 -346,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02377_X018_Y016_Z5175_525_525.jpg,I02377_X018_Y016_Z5175.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.34043999999999963,1661361,163 -347,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02378_X018_Y016_Z5351_525_525.jpg,I02378_X018_Y016_Z5351.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.2080599999999997,1661361,163 -348,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02379_X018_Y016_Z5234_525_525.jpg,I02379_X018_Y016_Z5234.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.06192999999999982,1661361,163 -349,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02380_X018_Y016_Z5292_525_525.jpg,I02380_X018_Y016_Z5292.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,0.0,1661361,163 -350,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04211_X018_Y023_Z4794_975_450.jpg,I04211_X018_Y023_Z4794.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-5.15921,1652336,53 -351,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04212_X018_Y023_Z4863_975_450.jpg,I04212_X018_Y023_Z4863.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-5.15663,1652336,53 -352,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04213_X018_Y023_Z4931_975_450.jpg,I04213_X018_Y023_Z4931.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-5.14436,1652336,53 -353,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04214_X018_Y023_Z5410_975_450.jpg,I04214_X018_Y023_Z5410.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-5.026389999999999,1652336,53 -354,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04215_X018_Y023_Z5000_975_450.jpg,I04215_X018_Y023_Z5000.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-4.97594,1652336,53 -355,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04216_X018_Y023_Z5068_975_450.jpg,I04216_X018_Y023_Z5068.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-4.02257,1652336,53 -356,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04217_X018_Y023_Z5341_975_450.jpg,I04217_X018_Y023_Z5341.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-3.16357,1652336,53 -357,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04218_X018_Y023_Z5136_975_450.jpg,I04218_X018_Y023_Z5136.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-2.0135899999999998,1652336,53 -358,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04219_X018_Y023_Z5273_975_450.jpg,I04219_X018_Y023_Z5273.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,-0.3254199999999994,1652336,53 -359,31/7197b078-de83-46e5-bfe5-0a3fee4b5454/I04220_X018_Y023_Z5205_975_450.jpg,I04220_X018_Y023_Z5205.jpg,31,7197b078-de83-46e5-bfe5-0a3fee4b5454,0.0,1652336,53 -360,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03291_X009_Y019_Z5205_750_450.jpg,I03291_X009_Y019_Z5205.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.14318,1666427,203 -361,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03292_X009_Y019_Z5273_750_450.jpg,I03292_X009_Y019_Z5273.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.00772,1666427,203 -362,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03293_X009_Y019_Z5820_750_450.jpg,I03293_X009_Y019_Z5820.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.9653400000000003,1666427,203 -363,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03294_X009_Y019_Z5341_750_450.jpg,I03294_X009_Y019_Z5341.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.5847800000000003,1666427,203 -364,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03295_X009_Y019_Z5751_750_450.jpg,I03295_X009_Y019_Z5751.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.50815,1666427,203 -365,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03296_X009_Y019_Z5683_750_450.jpg,I03296_X009_Y019_Z5683.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-2.62929,1666427,203 -366,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03297_X009_Y019_Z5410_750_450.jpg,I03297_X009_Y019_Z5410.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-2.30819,1666427,203 -367,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03298_X009_Y019_Z5615_750_450.jpg,I03298_X009_Y019_Z5615.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-0.8292700000000002,1666427,203 -368,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03299_X009_Y019_Z5478_750_450.jpg,I03299_X009_Y019_Z5478.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666427,203 -369,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03300_X009_Y019_Z5546_750_450.jpg,I03300_X009_Y019_Z5546.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.8541600000000003,1666427,203 -370,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01981_X014_Y014_Z5058_600_600.jpg,I01981_X014_Y014_Z5058.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.06733,1658248,142 -371,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01982_X014_Y014_Z5673_600_600.jpg,I01982_X014_Y014_Z5673.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.05065,1658248,142 -372,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01983_X014_Y014_Z5126_600_600.jpg,I01983_X014_Y014_Z5126.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.9908800000000002,1658248,142 -373,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01984_X014_Y014_Z5605_600_600.jpg,I01984_X014_Y014_Z5605.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.8200500000000002,1658248,142 -374,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01985_X014_Y014_Z5195_600_600.jpg,I01985_X014_Y014_Z5195.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.7759300000000002,1658248,142 -375,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01986_X014_Y014_Z5537_600_600.jpg,I01986_X014_Y014_Z5537.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.3338800000000002,1658248,142 -376,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01987_X014_Y014_Z5263_600_600.jpg,I01987_X014_Y014_Z5263.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.09829,1658248,142 -377,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01988_X014_Y014_Z5468_600_600.jpg,I01988_X014_Y014_Z5468.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.41255,1658248,142 -378,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01989_X014_Y014_Z5332_600_600.jpg,I01989_X014_Y014_Z5332.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658248,142 -379,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I01990_X014_Y014_Z5400_600_600.jpg,I01990_X014_Y014_Z5400.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.3852899999999999,1658248,142 -380,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02461_X017_Y015_Z4941_825_525.jpg,I02461_X017_Y015_Z4941.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-6.10772,1646182,9 -381,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02462_X017_Y015_Z5556_825_525.jpg,I02462_X017_Y015_Z5556.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-6.0020299999999995,1646182,9 -382,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02463_X017_Y015_Z5009_825_525.jpg,I02463_X017_Y015_Z5009.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-5.99302,1646182,9 -383,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02464_X017_Y015_Z5078_825_525.jpg,I02464_X017_Y015_Z5078.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-5.763489999999999,1646182,9 -384,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02465_X017_Y015_Z5146_825_525.jpg,I02465_X017_Y015_Z5146.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-5.272639999999999,1646182,9 -385,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02466_X017_Y015_Z5488_825_525.jpg,I02466_X017_Y015_Z5488.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-4.3412999999999995,1646182,9 -386,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02467_X017_Y015_Z5214_825_525.jpg,I02467_X017_Y015_Z5214.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-3.3186199999999992,1646182,9 -387,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02468_X017_Y015_Z5419_825_525.jpg,I02468_X017_Y015_Z5419.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-2.2662499999999994,1646182,9 -388,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02469_X017_Y015_Z5283_825_525.jpg,I02469_X017_Y015_Z5283.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,-0.14960999999999913,1646182,9 -389,31/10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02/I02470_X017_Y015_Z5351_825_525.jpg,I02470_X017_Y015_Z5351.jpg,31,10460be0-f5f3-4ac2-9ddf-b4ea9fa1ea02,0.0,1646182,9 -390,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02251_X007_Y013_Z5673_825_375.jpg,I02251_X007_Y013_Z5673.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-1.4255600000000002,1674584,212 -391,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02252_X007_Y013_Z5058_825_375.jpg,I02252_X007_Y013_Z5058.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-1.32227,1674584,212 -392,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02253_X007_Y013_Z5126_825_375.jpg,I02253_X007_Y013_Z5126.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-1.1189900000000002,1674584,212 -393,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02254_X007_Y013_Z5605_825_375.jpg,I02254_X007_Y013_Z5605.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-0.9820899999999999,1674584,212 -394,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02255_X007_Y013_Z5195_825_375.jpg,I02255_X007_Y013_Z5195.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-0.6792800000000003,1674584,212 -395,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02256_X007_Y013_Z5537_825_375.jpg,I02256_X007_Y013_Z5537.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,0.0,1674584,212 -396,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02257_X007_Y013_Z5263_825_375.jpg,I02257_X007_Y013_Z5263.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,1.1599999999999997,1674584,212 -397,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02258_X007_Y013_Z5468_825_375.jpg,I02258_X007_Y013_Z5468.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,1.6472000000000002,1674584,212 -398,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02259_X007_Y013_Z5332_825_375.jpg,I02259_X007_Y013_Z5332.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,3.8092099999999998,1674584,212 -399,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I02260_X007_Y013_Z5400_825_375.jpg,I02260_X007_Y013_Z5400.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,3.8625000000000003,1674584,212 -400,31/19d37361-fee4-4382-b29b-fb657f233636/I00241_X012_Y003_Z4501_675_375.jpg,I00241_X012_Y003_Z4501.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.83518,1649839,47 -401,31/19d37361-fee4-4382-b29b-fb657f233636/I00242_X012_Y003_Z5029_675_375.jpg,I00242_X012_Y003_Z5029.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.75365,1649839,47 -402,31/19d37361-fee4-4382-b29b-fb657f233636/I00243_X012_Y003_Z4560_675_375.jpg,I00243_X012_Y003_Z4560.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.68273,1649839,47 -403,31/19d37361-fee4-4382-b29b-fb657f233636/I00244_X012_Y003_Z4619_675_375.jpg,I00244_X012_Y003_Z4619.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.37309,1649839,47 -404,31/19d37361-fee4-4382-b29b-fb657f233636/I00245_X012_Y003_Z4970_675_375.jpg,I00245_X012_Y003_Z4970.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.2805300000000002,1649839,47 -405,31/19d37361-fee4-4382-b29b-fb657f233636/I00246_X012_Y003_Z4677_675_375.jpg,I00246_X012_Y003_Z4677.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.72376,1649839,47 -406,31/19d37361-fee4-4382-b29b-fb657f233636/I00247_X012_Y003_Z4912_675_375.jpg,I00247_X012_Y003_Z4912.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.5521599999999998,1649839,47 -407,31/19d37361-fee4-4382-b29b-fb657f233636/I00248_X012_Y003_Z4736_675_375.jpg,I00248_X012_Y003_Z4736.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.0,1649839,47 -408,31/19d37361-fee4-4382-b29b-fb657f233636/I00249_X012_Y003_Z4853_675_375.jpg,I00249_X012_Y003_Z4853.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.19214999999999982,1649839,47 -409,31/19d37361-fee4-4382-b29b-fb657f233636/I00250_X012_Y003_Z4794_675_375.jpg,I00250_X012_Y003_Z4794.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.64744,1649839,47 -410,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03111_X010_Y018_Z5458_0_75.jpg,I03111_X010_Y018_Z5458.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.46579,1666458,205 -411,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03112_X010_Y018_Z6074_0_75.jpg,I03112_X010_Y018_Z6074.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.26736,1666458,205 -412,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03113_X010_Y018_Z5527_0_75.jpg,I03113_X010_Y018_Z5527.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-6.17171,1666458,205 -413,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03114_X010_Y018_Z6005_0_75.jpg,I03114_X010_Y018_Z6005.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.94633,1666458,205 -414,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03115_X010_Y018_Z5595_0_75.jpg,I03115_X010_Y018_Z5595.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.66737,1666458,205 -415,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03116_X010_Y018_Z5937_0_75.jpg,I03116_X010_Y018_Z5937.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.94393,1666458,205 -416,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03117_X010_Y018_Z5664_0_75.jpg,I03117_X010_Y018_Z5664.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.82679,1666458,205 -417,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03118_X010_Y018_Z5869_0_75.jpg,I03118_X010_Y018_Z5869.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-2.6256500000000003,1666458,205 -418,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03119_X010_Y018_Z5732_0_75.jpg,I03119_X010_Y018_Z5732.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-0.8657300000000001,1666458,205 -419,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03120_X010_Y018_Z5800_0_75.jpg,I03120_X010_Y018_Z5800.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666458,205 -420,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02661_X007_Y015_Z5068_375_450.jpg,I02661_X007_Y015_Z5068.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-6.711459999999999,1697200,262 -421,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02662_X007_Y015_Z5136_375_450.jpg,I02662_X007_Y015_Z5136.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-6.469709999999999,1697200,262 -422,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02663_X007_Y015_Z5683_375_450.jpg,I02663_X007_Y015_Z5683.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-6.364189999999999,1697200,262 -423,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02664_X007_Y015_Z5615_375_450.jpg,I02664_X007_Y015_Z5615.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-5.39711,1697200,262 -424,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02665_X007_Y015_Z5205_375_450.jpg,I02665_X007_Y015_Z5205.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-4.6773299999999995,1697200,262 -425,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02666_X007_Y015_Z5546_375_450.jpg,I02666_X007_Y015_Z5546.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-3.927999999999999,1697200,262 -426,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02667_X007_Y015_Z5478_375_450.jpg,I02667_X007_Y015_Z5478.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-2.5435799999999995,1697200,262 -427,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02668_X007_Y015_Z5273_375_450.jpg,I02668_X007_Y015_Z5273.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-1.678259999999999,1697200,262 -428,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02669_X007_Y015_Z5410_375_450.jpg,I02669_X007_Y015_Z5410.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-0.7523699999999991,1697200,262 -429,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02670_X007_Y015_Z5341_375_450.jpg,I02670_X007_Y015_Z5341.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,0.0,1697200,262 -430,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02661_X007_Y015_Z5068_525_450.jpg,I02661_X007_Y015_Z5068.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-6.711459999999999,1697200,262 -431,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02662_X007_Y015_Z5136_525_450.jpg,I02662_X007_Y015_Z5136.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-6.469709999999999,1697200,262 -432,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02663_X007_Y015_Z5683_525_450.jpg,I02663_X007_Y015_Z5683.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-6.364189999999999,1697200,262 -433,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02664_X007_Y015_Z5615_525_450.jpg,I02664_X007_Y015_Z5615.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-5.39711,1697200,262 -434,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02665_X007_Y015_Z5205_525_450.jpg,I02665_X007_Y015_Z5205.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-4.6773299999999995,1697200,262 -435,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02666_X007_Y015_Z5546_525_450.jpg,I02666_X007_Y015_Z5546.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-3.927999999999999,1697200,262 -436,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02667_X007_Y015_Z5478_525_450.jpg,I02667_X007_Y015_Z5478.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-2.5435799999999995,1697200,262 -437,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02668_X007_Y015_Z5273_525_450.jpg,I02668_X007_Y015_Z5273.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-1.678259999999999,1697200,262 -438,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02669_X007_Y015_Z5410_525_450.jpg,I02669_X007_Y015_Z5410.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,-0.7523699999999991,1697200,262 -439,31/e62edecf-597c-4f50-93b7-816659e9a88d/I02670_X007_Y015_Z5341_525_450.jpg,I02670_X007_Y015_Z5341.jpg,31,e62edecf-597c-4f50-93b7-816659e9a88d,0.0,1697200,262 -440,31/19d37361-fee4-4382-b29b-fb657f233636/I01941_X005_Y013_Z4824_1050_450.jpg,I01941_X005_Y013_Z4824.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.1490200000000002,1649631,45 -441,31/19d37361-fee4-4382-b29b-fb657f233636/I01942_X005_Y013_Z4882_1050_450.jpg,I01942_X005_Y013_Z4882.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-1.09607,1649631,45 -442,31/19d37361-fee4-4382-b29b-fb657f233636/I01943_X005_Y013_Z4941_1050_450.jpg,I01943_X005_Y013_Z4941.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.7287400000000002,1649631,45 -443,31/19d37361-fee4-4382-b29b-fb657f233636/I01944_X005_Y013_Z5351_1050_450.jpg,I01944_X005_Y013_Z5351.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,-0.5274700000000001,1649631,45 -444,31/19d37361-fee4-4382-b29b-fb657f233636/I01945_X005_Y013_Z5000_1050_450.jpg,I01945_X005_Y013_Z5000.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.0,1649631,45 -445,31/19d37361-fee4-4382-b29b-fb657f233636/I01946_X005_Y013_Z5292_1050_450.jpg,I01946_X005_Y013_Z5292.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.5652399999999997,1649631,45 -446,31/19d37361-fee4-4382-b29b-fb657f233636/I01947_X005_Y013_Z5058_1050_450.jpg,I01947_X005_Y013_Z5058.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,0.8891099999999996,1649631,45 -447,31/19d37361-fee4-4382-b29b-fb657f233636/I01948_X005_Y013_Z5234_1050_450.jpg,I01948_X005_Y013_Z5234.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,1.6017200000000003,1649631,45 -448,31/19d37361-fee4-4382-b29b-fb657f233636/I01949_X005_Y013_Z5117_1050_450.jpg,I01949_X005_Y013_Z5117.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,1.6078599999999996,1649631,45 -449,31/19d37361-fee4-4382-b29b-fb657f233636/I01950_X005_Y013_Z5175_1050_450.jpg,I01950_X005_Y013_Z5175.jpg,31,19d37361-fee4-4382-b29b-fb657f233636,1.99911,1649631,45 -450,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02251_X014_Y015_Z5058_1200_375.jpg,I02251_X014_Y015_Z5058.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.7335700000000003,1658249,143 -451,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02252_X014_Y015_Z5673_1200_375.jpg,I02252_X014_Y015_Z5673.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.6783200000000003,1658249,143 -452,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02253_X014_Y015_Z5126_1200_375.jpg,I02253_X014_Y015_Z5126.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.6175800000000002,1658249,143 -453,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02254_X014_Y015_Z5605_1200_375.jpg,I02254_X014_Y015_Z5605.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.36231,1658249,143 -454,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02255_X014_Y015_Z5195_1200_375.jpg,I02255_X014_Y015_Z5195.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-2.19135,1658249,143 -455,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02256_X014_Y015_Z5537_1200_375.jpg,I02256_X014_Y015_Z5537.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.77366,1658249,143 -456,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02257_X014_Y015_Z5263_1200_375.jpg,I02257_X014_Y015_Z5263.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-1.10119,1658249,143 -457,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02258_X014_Y015_Z5468_1200_375.jpg,I02258_X014_Y015_Z5468.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.9268100000000001,1658249,143 -458,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02259_X014_Y015_Z5400_1200_375.jpg,I02259_X014_Y015_Z5400.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,-0.038269999999999804,1658249,143 -459,31/fba56d57-656e-4b6f-ba63-e4ba3ad083f5/I02260_X014_Y015_Z5332_1200_375.jpg,I02260_X014_Y015_Z5332.jpg,31,fba56d57-656e-4b6f-ba63-e4ba3ad083f5,0.0,1658249,143 -460,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00431_X007_Y004_Z5048_900_225.jpg,I00431_X007_Y004_Z5048.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.68069,1653833,77 -461,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00432_X007_Y004_Z5664_900_225.jpg,I00432_X007_Y004_Z5664.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.5910400000000005,1653833,77 -462,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00433_X007_Y004_Z5117_900_225.jpg,I00433_X007_Y004_Z5117.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.556570000000001,1653833,77 -463,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00434_X007_Y004_Z5185_900_225.jpg,I00434_X007_Y004_Z5185.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.27214,1653833,77 -464,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00435_X007_Y004_Z5595_900_225.jpg,I00435_X007_Y004_Z5595.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-5.176080000000001,1653833,77 -465,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00436_X007_Y004_Z5527_900_225.jpg,I00436_X007_Y004_Z5527.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.33522,1653833,77 -466,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00437_X007_Y004_Z5253_900_225.jpg,I00437_X007_Y004_Z5253.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-4.30171,1653833,77 -467,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00438_X007_Y004_Z5458_900_225.jpg,I00438_X007_Y004_Z5458.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-2.14862,1653833,77 -468,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00439_X007_Y004_Z5322_900_225.jpg,I00439_X007_Y004_Z5322.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,-1.3281900000000002,1653833,77 -469,31/679d4bad-e8d7-4094-8c70-486e4c091e21/I00440_X007_Y004_Z5390_900_225.jpg,I00440_X007_Y004_Z5390.jpg,31,679d4bad-e8d7-4094-8c70-486e4c091e21,0.0,1653833,77 -470,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01791_X001_Y013_Z5078_1125_75.jpg,I01791_X001_Y013_Z5078.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.87233,1676971,220 -471,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01792_X001_Y013_Z5146_1125_75.jpg,I01792_X001_Y013_Z5146.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.78289,1676971,220 -472,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01793_X001_Y013_Z5693_1125_75.jpg,I01793_X001_Y013_Z5693.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.77005,1676971,220 -473,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01794_X001_Y013_Z5214_1125_75.jpg,I01794_X001_Y013_Z5214.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.44409,1676971,220 -474,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01795_X001_Y013_Z5625_1125_75.jpg,I01795_X001_Y013_Z5625.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-4.20589,1676971,220 -475,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01796_X001_Y013_Z5283_1125_75.jpg,I01796_X001_Y013_Z5283.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-3.4436600000000004,1676971,220 -476,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01797_X001_Y013_Z5556_1125_75.jpg,I01797_X001_Y013_Z5556.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-3.2127800000000004,1676971,220 -477,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01798_X001_Y013_Z5488_1125_75.jpg,I01798_X001_Y013_Z5488.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-1.37507,1676971,220 -478,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01799_X001_Y013_Z5351_1125_75.jpg,I01799_X001_Y013_Z5351.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,-0.6813900000000004,1676971,220 -479,31/db6784e7-fa75-49aa-804f-464fefe1dfc2/I01800_X001_Y013_Z5419_1125_75.jpg,I01800_X001_Y013_Z5419.jpg,31,db6784e7-fa75-49aa-804f-464fefe1dfc2,0.0,1676971,220 -480,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02221_X006_Y015_Z4970_825_600.jpg,I02221_X006_Y015_Z4970.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.7077200000000001,1661046,154 -481,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02222_X006_Y015_Z5029_825_600.jpg,I02222_X006_Y015_Z5029.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.63877,1661046,154 -482,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02223_X006_Y015_Z5498_825_600.jpg,I02223_X006_Y015_Z5498.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.52933,1661046,154 -483,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02224_X006_Y015_Z5087_825_600.jpg,I02224_X006_Y015_Z5087.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.49641,1661046,154 -484,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02225_X006_Y015_Z5439_825_600.jpg,I02225_X006_Y015_Z5439.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.1653100000000003,1661046,154 -485,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02226_X006_Y015_Z5146_825_600.jpg,I02226_X006_Y015_Z5146.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-1.0208400000000002,1661046,154 -486,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02227_X006_Y015_Z5380_825_600.jpg,I02227_X006_Y015_Z5380.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.6297600000000001,1661046,154 -487,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02228_X006_Y015_Z5205_825_600.jpg,I02228_X006_Y015_Z5205.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.38270000000000026,1661046,154 -488,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02229_X006_Y015_Z5322_825_600.jpg,I02229_X006_Y015_Z5322.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.11827999999999994,1661046,154 -489,31/274d8969-aa7c-4ac0-be60-e753579393ad/I02230_X006_Y015_Z5263_825_600.jpg,I02230_X006_Y015_Z5263.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,0.0,1661046,154 -490,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03231_X003_Y019_Z5498_825_525.jpg,I03231_X003_Y019_Z5498.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.540649999999999,1666255,192 -491,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03232_X003_Y019_Z6113_825_525.jpg,I03232_X003_Y019_Z6113.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.22211,1666255,192 -492,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03233_X003_Y019_Z5566_825_525.jpg,I03233_X003_Y019_Z5566.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.20641,1666255,192 -493,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03234_X003_Y019_Z6044_825_525.jpg,I03234_X003_Y019_Z6044.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.42676,1666255,192 -494,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03235_X003_Y019_Z5634_825_525.jpg,I03235_X003_Y019_Z5634.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.2971,1666255,192 -495,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03236_X003_Y019_Z5703_825_525.jpg,I03236_X003_Y019_Z5703.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.00256,1666255,192 -496,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03237_X003_Y019_Z5976_825_525.jpg,I03237_X003_Y019_Z5976.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-2.8840899999999996,1666255,192 -497,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03238_X003_Y019_Z5908_825_525.jpg,I03238_X003_Y019_Z5908.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-1.5780599999999998,1666255,192 -498,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03239_X003_Y019_Z5771_825_525.jpg,I03239_X003_Y019_Z5771.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-0.21882000000000001,1666255,192 -499,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03240_X003_Y019_Z5839_825_525.jpg,I03240_X003_Y019_Z5839.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666255,192 -500,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05521_X007_Y028_Z5097_225_375.jpg,I05521_X007_Y028_Z5097.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-2.01538,1654469,83 -501,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05522_X007_Y028_Z5166_225_375.jpg,I05522_X007_Y028_Z5166.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-1.9193399999999998,1654469,83 -502,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05523_X007_Y028_Z5712_225_375.jpg,I05523_X007_Y028_Z5712.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-1.8066,1654469,83 -503,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05524_X007_Y028_Z5234_225_375.jpg,I05524_X007_Y028_Z5234.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-1.7116699999999998,1654469,83 -504,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05525_X007_Y028_Z5644_225_375.jpg,I05525_X007_Y028_Z5644.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-1.3220299999999998,1654469,83 -505,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05526_X007_Y028_Z5302_225_375.jpg,I05526_X007_Y028_Z5302.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-1.0526,1654469,83 -506,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05527_X007_Y028_Z5576_225_375.jpg,I05527_X007_Y028_Z5576.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,-0.5564299999999998,1654469,83 -507,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05528_X007_Y028_Z5371_225_375.jpg,I05528_X007_Y028_Z5371.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,0.0,1654469,83 -508,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05529_X007_Y028_Z5507_225_375.jpg,I05529_X007_Y028_Z5507.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,0.4536000000000002,1654469,83 -509,31/0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd/I05530_X007_Y028_Z5439_225_375.jpg,I05530_X007_Y028_Z5439.jpg,31,0ce6bee3-089e-48a8-ac55-9d45f4d9d9fd,0.7401600000000004,1654469,83 -510,31/6e592015-f152-43be-aace-5949fb03fd34/I00561_X004_Y006_Z4980_525_600.jpg,I00561_X004_Y006_Z4980.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.750489999999999,1649348,32 -511,31/6e592015-f152-43be-aace-5949fb03fd34/I00562_X004_Y006_Z5048_525_600.jpg,I00562_X004_Y006_Z5048.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.4710399999999995,1649348,32 -512,31/6e592015-f152-43be-aace-5949fb03fd34/I00563_X004_Y006_Z5595_525_600.jpg,I00563_X004_Y006_Z5595.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.32499,1649348,32 -513,31/6e592015-f152-43be-aace-5949fb03fd34/I00564_X004_Y006_Z5117_525_600.jpg,I00564_X004_Y006_Z5117.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-4.887359999999999,1649348,32 -514,31/6e592015-f152-43be-aace-5949fb03fd34/I00565_X004_Y006_Z5527_525_600.jpg,I00565_X004_Y006_Z5527.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-4.09256,1649348,32 -515,31/6e592015-f152-43be-aace-5949fb03fd34/I00566_X004_Y006_Z5185_525_600.jpg,I00566_X004_Y006_Z5185.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-3.1511299999999993,1649348,32 -516,31/6e592015-f152-43be-aace-5949fb03fd34/I00567_X004_Y006_Z5458_525_600.jpg,I00567_X004_Y006_Z5458.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-2.2828999999999997,1649348,32 -517,31/6e592015-f152-43be-aace-5949fb03fd34/I00568_X004_Y006_Z5390_525_600.jpg,I00568_X004_Y006_Z5390.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.0582399999999996,1649348,32 -518,31/6e592015-f152-43be-aace-5949fb03fd34/I00569_X004_Y006_Z5253_525_600.jpg,I00569_X004_Y006_Z5253.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.0,1649348,32 -519,31/6e592015-f152-43be-aace-5949fb03fd34/I00570_X004_Y006_Z5322_525_600.jpg,I00570_X004_Y006_Z5322.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.42213999999999974,1649348,32 -520,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04351_X010_Y028_Z5361_975_375.jpg,I04351_X010_Y028_Z5361.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8991200000000001,1661173,160 -521,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04352_X010_Y028_Z5888_975_375.jpg,I04352_X010_Y028_Z5888.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8380900000000002,1661173,160 -522,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04353_X010_Y028_Z5419_975_375.jpg,I04353_X010_Y028_Z5419.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.8046700000000002,1661173,160 -523,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04354_X010_Y028_Z5830_975_375.jpg,I04354_X010_Y028_Z5830.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.6336100000000002,1661173,160 -524,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04355_X010_Y028_Z5478_975_375.jpg,I04355_X010_Y028_Z5478.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.6198100000000002,1661173,160 -525,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04356_X010_Y028_Z5771_975_375.jpg,I04356_X010_Y028_Z5771.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.41593000000000013,1661173,160 -526,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04357_X010_Y028_Z5537_975_375.jpg,I04357_X010_Y028_Z5537.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.3497500000000002,1661173,160 -527,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04358_X010_Y028_Z5712_975_375.jpg,I04358_X010_Y028_Z5712.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.1528400000000003,1661173,160 -528,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04359_X010_Y028_Z5595_975_375.jpg,I04359_X010_Y028_Z5595.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,-0.09433000000000025,1661173,160 -529,31/274d8969-aa7c-4ac0-be60-e753579393ad/I04360_X010_Y028_Z5654_975_375.jpg,I04360_X010_Y028_Z5654.jpg,31,274d8969-aa7c-4ac0-be60-e753579393ad,0.0,1661173,160 -530,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03201_X001_Y018_Z5292_525_75.jpg,I03201_X001_Y018_Z5292.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.4066,1666217,185 -531,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03202_X001_Y018_Z5361_525_75.jpg,I03202_X001_Y018_Z5361.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.237299999999999,1666217,185 -532,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03203_X001_Y018_Z5908_525_75.jpg,I03203_X001_Y018_Z5908.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-5.07073,1666217,185 -533,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03204_X001_Y018_Z5429_525_75.jpg,I03204_X001_Y018_Z5429.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.836919999999999,1666217,185 -534,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03205_X001_Y018_Z5839_525_75.jpg,I03205_X001_Y018_Z5839.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-4.34463,1666217,185 -535,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03206_X001_Y018_Z5498_525_75.jpg,I03206_X001_Y018_Z5498.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.8039199999999997,1666217,185 -536,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03207_X001_Y018_Z5771_525_75.jpg,I03207_X001_Y018_Z5771.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-3.06764,1666217,185 -537,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03208_X001_Y018_Z5703_525_75.jpg,I03208_X001_Y018_Z5703.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-1.8345899999999995,1666217,185 -538,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03209_X001_Y018_Z5566_525_75.jpg,I03209_X001_Y018_Z5566.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,-1.2095500000000001,1666217,185 -539,31/2e66350d-9b1d-4f78-b66f-14f0d6eb23fc/I03210_X001_Y018_Z5634_525_75.jpg,I03210_X001_Y018_Z5634.jpg,31,2e66350d-9b1d-4f78-b66f-14f0d6eb23fc,0.0,1666217,185 -540,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03141_X019_Y017_Z5009_150_300.jpg,I03141_X019_Y017_Z5009.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-4.672890000000001,1694486,254 -541,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03142_X019_Y017_Z5078_150_300.jpg,I03142_X019_Y017_Z5078.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-4.53617,1694486,254 -542,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03143_X019_Y017_Z5625_150_300.jpg,I03143_X019_Y017_Z5625.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-4.43132,1694486,254 -543,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03144_X019_Y017_Z5146_150_300.jpg,I03144_X019_Y017_Z5146.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-4.30693,1694486,254 -544,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03145_X019_Y017_Z5556_150_300.jpg,I03145_X019_Y017_Z5556.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-3.7882100000000003,1694486,254 -545,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03146_X019_Y017_Z5214_150_300.jpg,I03146_X019_Y017_Z5214.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-3.65682,1694486,254 -546,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03147_X019_Y017_Z5488_150_300.jpg,I03147_X019_Y017_Z5488.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-2.55783,1694486,254 -547,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03148_X019_Y017_Z5283_150_300.jpg,I03148_X019_Y017_Z5283.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-2.0602,1694486,254 -548,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03149_X019_Y017_Z5419_150_300.jpg,I03149_X019_Y017_Z5419.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,-0.46848000000000045,1694486,254 -549,31/8e572b7a-08fa-4c17-a2e3-7bf606c52a30/I03150_X019_Y017_Z5351_150_300.jpg,I03150_X019_Y017_Z5351.jpg,31,8e572b7a-08fa-4c17-a2e3-7bf606c52a30,0.0,1694486,254 -550,31/6e592015-f152-43be-aace-5949fb03fd34/I00271_X009_Y004_Z5703_750_150.jpg,I00271_X009_Y004_Z5703.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.912370000000001,1649450,38 -551,31/6e592015-f152-43be-aace-5949fb03fd34/I00272_X009_Y004_Z5087_750_150.jpg,I00272_X009_Y004_Z5087.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.790280000000001,1649450,38 -552,31/6e592015-f152-43be-aace-5949fb03fd34/I00273_X009_Y004_Z5634_750_150.jpg,I00273_X009_Y004_Z5634.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.529480000000001,1649450,38 -553,31/6e592015-f152-43be-aace-5949fb03fd34/I00274_X009_Y004_Z5156_750_150.jpg,I00274_X009_Y004_Z5156.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.2340800000000005,1649450,38 -554,31/6e592015-f152-43be-aace-5949fb03fd34/I00275_X009_Y004_Z5566_750_150.jpg,I00275_X009_Y004_Z5566.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-5.140070000000001,1649450,38 -555,31/6e592015-f152-43be-aace-5949fb03fd34/I00276_X009_Y004_Z5498_750_150.jpg,I00276_X009_Y004_Z5498.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-3.3135400000000006,1649450,38 -556,31/6e592015-f152-43be-aace-5949fb03fd34/I00277_X009_Y004_Z5224_750_150.jpg,I00277_X009_Y004_Z5224.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-3.2179200000000003,1649450,38 -557,31/6e592015-f152-43be-aace-5949fb03fd34/I00278_X009_Y004_Z5429_750_150.jpg,I00278_X009_Y004_Z5429.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-1.0945500000000008,1649450,38 -558,31/6e592015-f152-43be-aace-5949fb03fd34/I00279_X009_Y004_Z5292_750_150.jpg,I00279_X009_Y004_Z5292.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,-0.5576700000000008,1649450,38 -559,31/6e592015-f152-43be-aace-5949fb03fd34/I00280_X009_Y004_Z5361_750_150.jpg,I00280_X009_Y004_Z5361.jpg,31,6e592015-f152-43be-aace-5949fb03fd34,0.0,1649450,38 -560,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01401_X016_Y009_Z5498_300_450.jpg,I01401_X016_Y009_Z5498.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-4.82709,1674918,217 -561,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01402_X016_Y009_Z4882_300_450.jpg,I01402_X016_Y009_Z4882.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-4.79788,1674918,217 -562,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01403_X016_Y009_Z4951_300_450.jpg,I01403_X016_Y009_Z4951.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-4.6450700000000005,1674918,217 -563,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01404_X016_Y009_Z5429_300_450.jpg,I01404_X016_Y009_Z5429.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-4.4177599999999995,1674918,217 -564,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01405_X016_Y009_Z5019_300_450.jpg,I01405_X016_Y009_Z5019.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-4.358879999999999,1674918,217 -565,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01406_X016_Y009_Z5361_300_450.jpg,I01406_X016_Y009_Z5361.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-3.41147,1674918,217 -566,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01407_X016_Y009_Z5087_300_450.jpg,I01407_X016_Y009_Z5087.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-3.0542399999999996,1674918,217 -567,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01408_X016_Y009_Z5292_300_450.jpg,I01408_X016_Y009_Z5292.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-1.4860799999999994,1674918,217 -568,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01409_X016_Y009_Z5156_300_450.jpg,I01409_X016_Y009_Z5156.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,-0.5280399999999998,1674918,217 -569,31/4c7e9e66-61a1-47ca-aa4e-340b0eef8db1/I01410_X016_Y009_Z5224_300_450.jpg,I01410_X016_Y009_Z5224.jpg,31,4c7e9e66-61a1-47ca-aa4e-340b0eef8db1,0.0,1674918,217 diff --git a/models/notebooks/.gitkeep b/notebooks/.gitkeep similarity index 100% rename from models/notebooks/.gitkeep rename to notebooks/.gitkeep diff --git a/models/notebooks/1.0-hfk-datamodules-exploration.ipynb b/notebooks/1.0-hfk-datamodules-exploration.ipynb similarity index 100% rename from models/notebooks/1.0-hfk-datamodules-exploration.ipynb rename to notebooks/1.0-hfk-datamodules-exploration.ipynb diff --git a/models/requirements.txt b/requirements.txt similarity index 100% rename from models/requirements.txt rename to requirements.txt diff --git a/models/scripts/schedule.sh b/scripts/schedule.sh similarity index 100% rename from models/scripts/schedule.sh rename to scripts/schedule.sh diff --git a/models/setup.cfg b/setup.cfg similarity index 100% rename from models/setup.cfg rename to setup.cfg diff --git a/models/src/__init__.py b/src/__init__.py similarity index 100% rename from models/src/__init__.py rename to src/__init__.py diff --git a/models/src/datamodules/__init__.py b/src/datamodules/__init__.py similarity index 100% rename from models/src/datamodules/__init__.py rename to src/datamodules/__init__.py diff --git a/models/src/datamodules/components/__init__.py b/src/datamodules/components/__init__.py similarity index 100% rename from models/src/datamodules/components/__init__.py rename to src/datamodules/components/__init__.py diff --git a/models/src/datamodules/focus_datamodule.py b/src/datamodules/focus_datamodule.py similarity index 100% rename from models/src/datamodules/focus_datamodule.py rename to src/datamodules/focus_datamodule.py diff --git a/models/src/datamodules/mnist_datamodule.py b/src/datamodules/mnist_datamodule.py similarity index 100% rename from models/src/datamodules/mnist_datamodule.py rename to src/datamodules/mnist_datamodule.py diff --git a/models/src/models/__init__.py b/src/models/__init__.py similarity index 100% rename from models/src/models/__init__.py rename to src/models/__init__.py diff --git a/models/src/models/components/__init__.py b/src/models/components/__init__.py similarity index 100% rename from models/src/models/components/__init__.py rename to src/models/components/__init__.py diff --git a/models/src/models/components/simple_dense_net.py b/src/models/components/simple_dense_net.py similarity index 100% rename from models/src/models/components/simple_dense_net.py rename to src/models/components/simple_dense_net.py diff --git a/models/src/models/focus_module.py b/src/models/focus_module.py similarity index 100% rename from models/src/models/focus_module.py rename to src/models/focus_module.py diff --git a/models/src/models/mnist_module.py b/src/models/mnist_module.py similarity index 100% rename from models/src/models/mnist_module.py rename to src/models/mnist_module.py diff --git a/models/src/testing_pipeline.py b/src/testing_pipeline.py similarity index 100% rename from models/src/testing_pipeline.py rename to src/testing_pipeline.py diff --git a/models/src/training_pipeline.py b/src/training_pipeline.py similarity index 100% rename from models/src/training_pipeline.py rename to src/training_pipeline.py diff --git a/models/src/utils/__init__.py b/src/utils/__init__.py similarity index 100% rename from models/src/utils/__init__.py rename to src/utils/__init__.py diff --git a/models/src/vendor/__init__.py b/src/vendor/__init__.py similarity index 100% rename from models/src/vendor/__init__.py rename to src/vendor/__init__.py diff --git a/models/test.py b/test.py similarity index 100% rename from models/test.py rename to test.py diff --git a/models/tests/__init__.py b/tests/__init__.py similarity index 100% rename from models/tests/__init__.py rename to tests/__init__.py diff --git a/models/tests/helpers/__init__.py b/tests/helpers/__init__.py similarity index 100% rename from models/tests/helpers/__init__.py rename to tests/helpers/__init__.py diff --git a/models/tests/helpers/module_available.py b/tests/helpers/module_available.py similarity index 100% rename from models/tests/helpers/module_available.py rename to tests/helpers/module_available.py diff --git a/models/tests/helpers/run_command.py b/tests/helpers/run_command.py similarity index 100% rename from models/tests/helpers/run_command.py rename to tests/helpers/run_command.py diff --git a/models/tests/helpers/runif.py b/tests/helpers/runif.py similarity index 100% rename from models/tests/helpers/runif.py rename to tests/helpers/runif.py diff --git a/models/tests/shell/__init__.py b/tests/shell/__init__.py similarity index 100% rename from models/tests/shell/__init__.py rename to tests/shell/__init__.py diff --git a/models/tests/shell/test_basic_commands.py b/tests/shell/test_basic_commands.py similarity index 100% rename from models/tests/shell/test_basic_commands.py rename to tests/shell/test_basic_commands.py diff --git a/models/tests/shell/test_debug_configs.py b/tests/shell/test_debug_configs.py similarity index 100% rename from models/tests/shell/test_debug_configs.py rename to tests/shell/test_debug_configs.py diff --git a/models/tests/shell/test_sweeps.py b/tests/shell/test_sweeps.py similarity index 100% rename from models/tests/shell/test_sweeps.py rename to tests/shell/test_sweeps.py diff --git a/models/tests/unit/__init__.py b/tests/unit/__init__.py similarity index 100% rename from models/tests/unit/__init__.py rename to tests/unit/__init__.py diff --git a/models/tests/unit/test_mnist_datamodule.py b/tests/unit/test_mnist_datamodule.py similarity index 100% rename from models/tests/unit/test_mnist_datamodule.py rename to tests/unit/test_mnist_datamodule.py diff --git a/models/train.py b/train.py similarity index 100% rename from models/train.py rename to train.py