From d97bf031901c2576fe47aad6d37bdf2bae93dd48 Mon Sep 17 00:00:00 2001
From: Yongbin Sun <yb_sun@mit.edu>
Date: Sun, 4 Feb 2018 20:38:46 -0500
Subject: [PATCH] Update and rename Readme.md to README.md

---
 part_seg/{Readme.md => README.md} | 18 ++++++------------
 1 file changed, 6 insertions(+), 12 deletions(-)
 rename part_seg/{Readme.md => README.md} (60%)

diff --git a/part_seg/Readme.md b/part_seg/README.md
similarity index 60%
rename from part_seg/Readme.md
rename to part_seg/README.md
index 0761044..6b1c6db 100644
--- a/part_seg/Readme.md
+++ b/part_seg/README.md
@@ -1,10 +1,6 @@
-## Run part segmentation experiments
+## Part segmentation
 
-### Requirements
-
-2 GPUs (each with 12 GB memory) for distributed training.
-
-### Data preparation 
+### Dataset 
 
 Load the data for part segmentation.
 
@@ -12,9 +8,9 @@ Load the data for part segmentation.
 sh +x download_data.sh
 ```
 
-### Training 
+### Train
 
-Train the model using 2 GPUs. 
+Train the model on 2 GPUs, each with 12 GB memeory. 
 
 ```
 python train_multi_gpu.py
@@ -22,16 +18,14 @@ python train_multi_gpu.py
 
 Model parameters are saved every 10 epochs in "train_results/trained_models/".
 
-### Testing
+### Evaluation
 
-To test the model parameter set saved after epoch n, 
+To evaluate the model saved after epoch n, 
 
 ```
 python test.py --model_path train_results/trained_models/epoch_n.ckpt
 ```
 
-n = 0, 10, 20, ..., 190, 200.
-
 For example, if we want to test the model saved after 160 epochs, 
 
 ```