config files

README.md

@@ -5,12 +5,12 @@ This repository is based on the official implementation of [Discrete Point Flow]
 
 # Environment
 
-This repository requires:
+We provide all necessary requirements in form of a `environment.yml`.
 
-- pytorch
-- ...
-
-We further provide all necessary requirements in for of a `requirements.txt`.
+For our evaluation we rely on the efficient implementation of the EMD metric provided by [PointFlow](https://github.com/stevenygd/PointFlow).
+To this end, we refer to the installation instructions provided there. 
+Alternatively, the precompiled code can be downloaded [here](https://drive.google.com/drive/folders/1jFo6gSuQNjVq-8oB0iZ2YkMRFVP7t8GC?usp=sharing),
+which needs to be unzipped and placed in `lib/metrics/` and is expected to work with the provided `environment.yml`. 
 
 # Datasets
 
@@ -38,14 +38,12 @@ Since the preprocessing takes up to a week, we provide the preprocessed datasets
 # Pretrained models
 
 All pretrained models including the corresponding config files can be downloaded [here](https://drive.google.com/drive/folders/1fkVBVqxy2_zTevwd3WdnROPreYke-zuU?usp=sharing).
-To use the models, you need to download the models and put the files in the root directory `./`.
-Then, specify the `path2data` storing preprocessed data and path2save directory storing all saved
-checkpoints.
-
+To use the models during evaluation, specify your path to the preprocessed data `path2data` in the configs of the pretrained models.
+
 # Training
 
 All training configurations can be found in `configs/`. Prior to training/evaluation remember to set
-`path2data` in the resp. config file accordingly.
+`path2data` in the resp. config file accordingly. Note, `path2save` specifies the logging directory and defaults to `./results`.
 
 ## Generative modeling
 

configs/config_generative_modeling_airplane.yaml

@@ -41,11 +41,11 @@ num_workers: 8
 p_decoder_base_type: free
 p_decoder_base_var: -3.9551
 p_decoder_n_features: 64
-p_decoder_n_flows: 1 #21
+p_decoder_n_flows: 21
 p_latent_space_size: 3
 p_prior_n_layers: 1
 params_reduce_mode: depth_and_feature
-path2data: /usr/local/google/home/postels/research/data/ShapeNet # your/path/to/data
+path2data: your/path/to/data
 path2save: ./results
 pc_enc_init_n_channels: 3
 pc_enc_init_n_features: 64

scripts/run_evaluate_ae.sh

@@ -1 +1 @@
-python evaluate_ae.py ./configs/config_autoencoding.yaml path_to_trained_model test 2048 2048 autoencoding --weights_type learned_weights --reps 1 --f1_threshold_lst 0.0001 --cd --f1 --emd
+python evaluate_ae.py path_to_experiment_root name_of_trained_model test 2048 2048 autoencoding --weights_type learned_weights --reps 1 --f1_threshold_lst 0.0001 --cd --f1 --emd

scripts/run_evaluate_gen.sh

@@ -1 +1 @@
-python evaluate_ae.py ./configs/config_generative_modeling_airplane.yaml path_to_trained_model test 2048 2048 generating --weights_type learned_weights --reps 10 --f1_threshold_lst 0.0001 --cd --emd
+python evaluate_ae.py path_to_experiment_root name_of_trained_model test 2048 2048 generating --weights_type learned_weights --reps 10 --f1_threshold_lst 0.0001 --cd --emd

scripts/run_evaluate_svr.sh

@@ -1 +1 @@
-python evaluate_ae.py ./configs/config_SVR.yaml path_to_trained_model test 2500 2500 reconstruction --weights_type learned_weights --reps 1 --f1_threshold_lst 0.001 --cd --f1 --emd --unit_scale_evaluation
+python evaluate_ae.py path_to_experiment_root name_of_trained_model test 2500 2500 reconstruction --weights_type learned_weights --reps 1 --f1_threshold_lst 0.001 --cd --f1 --emd --unit_scale_evaluation