From: François Fleuret <francois@fleuret.org>
Date: Sun, 23 Jun 2024 07:06:29 +0000 (+0200)
Subject: Update.
X-Git-Url: https://fleuret.org/cgi-bin/gitweb/gitweb.cgi?a=commitdiff_plain;h=fdce96c3960f5f544bd68c0f18cc5cd096ecbfb3;p=culture.git

Update.
---

diff --git a/README.txt b/README.txt
index d5cf6d8..4a5ca30 100644
--- a/README.txt
+++ b/README.txt
@@ -38,12 +38,12 @@ My home-baked GPT-37M trained with 250k solves this with ~99% success
 At every iteration, we select the GPT with the lowest test accuracy,
 and run one epoch.
 
+* Creating new quizzes
+
 If its test accuracy got higher than 97.5%, it will create new
 quizzes. To do so, it generates a large number of pairs of frames, and
 checks which ones of these quizzes are hard but not too hard, which
-means
-
-[THIS IS THE IMPORTANT BIT]:
+means [THIS IS THE IMPORTANT BIT]:
 
 it can be solved, in both time directions, by all the other GPTs **but
 one**
diff --git a/tasks.py b/tasks.py
index 1254323..ecf0a65 100755
--- a/tasks.py
+++ b/tasks.py
@@ -101,8 +101,8 @@ class World(Task):
 
         self.batch_size = batch_size
         self.device = device
-        self.height = 7
-        self.width = 9
+        self.height = 6
+        self.width = 8
 
         self.train_input = world.generate_seq(
             nb_train_samples, height=self.height, width=self.width