From 929f2dd097a44e12016b11729c11a022ddfd1618 Mon Sep 17 00:00:00 2001
From: Sun Jin Kim <sk2521@ic.ac.uk>
Date: Fri, 8 Apr 2022 22:33:24 +0900
Subject: [PATCH] mnist 0.01 randomsearch vs gru graph data
---
gru_learner.pkl => 0_01pkls/gru_learner.pkl | Bin
gru_logs.pkl => 0_01pkls/gru_logs.pkl | Bin
.../randomsearch_logs.pkl | Bin
MetaAugment/autoaugment_learners/aa_learner.py | 4 ++--
plot_pickles.py | 9 +++++++++
5 files changed, 11 insertions(+), 2 deletions(-)
rename gru_learner.pkl => 0_01pkls/gru_learner.pkl (100%)
rename gru_logs.pkl => 0_01pkls/gru_logs.pkl (100%)
rename randomsearch_logs.pkl => 0_01pkls/randomsearch_logs.pkl (100%)
diff --git a/gru_learner.pkl b/0_01pkls/gru_learner.pkl
similarity index 100%
rename from gru_learner.pkl
rename to 0_01pkls/gru_learner.pkl
diff --git a/gru_logs.pkl b/0_01pkls/gru_logs.pkl
similarity index 100%
rename from gru_logs.pkl
rename to 0_01pkls/gru_logs.pkl
diff --git a/randomsearch_logs.pkl b/0_01pkls/randomsearch_logs.pkl
similarity index 100%
rename from randomsearch_logs.pkl
rename to 0_01pkls/randomsearch_logs.pkl
diff --git a/MetaAugment/autoaugment_learners/aa_learner.py b/MetaAugment/autoaugment_learners/aa_learner.py
index 6e7874e9..909b05b9 100644
--- a/MetaAugment/autoaugment_learners/aa_learner.py
+++ b/MetaAugment/autoaugment_learners/aa_learner.py
@@ -233,7 +233,7 @@ class aa_learner:
train_loader, test_loader = create_toy(train_dataset,
test_dataset,
batch_size=64,
- n_samples=0.01,
+ n_samples=0.04,
seed=100)
# train the child network with the dataloaders equipped with our specific policy
@@ -244,7 +244,7 @@ class aa_learner:
# sgd = optim.Adadelta(child_network.parameters(), lr=1e-2),
cost = nn.CrossEntropyLoss(),
max_epochs = 3000000,
- early_stop_num = 120,
+ early_stop_num = 60,
logging = logging)
# if logging is true, 'accuracy' is actually a tuple: (accuracy, accuracy_log)
diff --git a/plot_pickles.py b/plot_pickles.py
index 7f0f2d7c..90462cfa 100644
--- a/plot_pickles.py
+++ b/plot_pickles.py
@@ -1,6 +1,7 @@
import pickle
from pprint import pprint
import matplotlib.pyplot as plt
+from torch import gru
def get_maxacc(log):
output = []
@@ -25,6 +26,14 @@ plt.xlabel('number of policies tested')
plt.legend()
plt.show()
+plt.plot([acc for pol,acc in rs_list], label='randomsearcher')
+plt.plot([acc for pol,acc in gru_list], label='gru learner')
+plt.title('Comparing two agents')
+plt.ylabel('best accuracy to date')
+plt.xlabel('number of policies tested')
+plt.legend()
+plt.show()
+
def get_best5(log):
l = sorted(log, reverse=True, key=lambda x:x[1])
--
GitLab