Update.

[culture.git] / main.py

diff --git a/main.py b/main.py
index 918f75d..43241dd 100755 (executable)
--- a/main.py
+++ b/main.py
@@ -13,7 +13,7 @@ from torch.nn import functional as F
 
 import ffutils
 import mygpt
-import sky, wireworld, quizz_machine
+import sky, reasoning, quizz_machine
 
 # world quizzes vs. culture quizzes
 
@@ -79,23 +79,23 @@ parser.add_argument("--dropout", type=float, default=0.1)
 
 parser.add_argument("--deterministic_synthesis", action="store_true", default=False)
 
-parser.add_argument("--both_directions", action="store_true", default=False)
-
 parser.add_argument("--problem", type=str, default="sky")
 
 parser.add_argument("--nb_gpts", type=int, default=5)
 
-parser.add_argument("--min_to_validate", type=int, default=4)
+parser.add_argument("--min_to_validate", type=int, default=None)
 
-parser.add_argument("--max_to_validate", type=int, default=4)
+parser.add_argument("--max_to_validate", type=int, default=None)
 
 parser.add_argument("--accuracy_to_make_c_quizzes", type=float, default=0.975)
 
-parser.add_argument("--dirty_debug", action="store_true", default=False)
+parser.add_argument("--generation_temperature", type=float, default=2.0)
+
+parser.add_argument("--deterministic_validation", action="store_true", default=False)
 
-parser.add_argument("--generation_temperature", type=float, default=1.0)
+parser.add_argument("--bidirectional_validation", action="store_true", default=False)
 
-parser.add_argument("--stochastic_validation", action="store_true", default=False)
+parser.add_argument("--dirty_debug", action="store_true", default=False)
 
 ######################################################################
 
@@ -113,6 +113,12 @@ parser.add_argument("--sky_speed", type=int, default=3)
 
 args = parser.parse_args()
 
+if args.min_to_validate is None:
+    args.min_to_validate = args.nb_gpts - 1
+
+if args.max_to_validate is None:
+    args.max_to_validate = args.nb_gpts - 1
+
 if args.result_dir is None:
     args.result_dir = f"results_culture"
 
@@ -243,8 +249,10 @@ if args.problem == "sky":
         nb_iterations=args.sky_nb_iterations,
         speed=args.sky_speed,
     )
-elif args.problem == "wireworld":
-    problem = wireworld.Wireworld(height=8, width=10, nb_iterations=2, speed=5)
+    back_accuracy = False
+elif args.problem == "reasoning":
+    problem = reasoning.Reasoning(device=device)
+    back_accuracy = True
 else:
     raise ValueError
 
@@ -252,6 +260,7 @@ quizz_machine = quizz_machine.QuizzMachine(
     problem=problem,
     nb_train_samples=args.nb_train_samples,
     nb_test_samples=args.nb_test_samples,
+    back_accuracy=back_accuracy,
     batch_size=args.physical_batch_size,
     result_dir=args.result_dir,
     logger=log_string,
@@ -423,8 +432,8 @@ def create_c_quizzes(
             nb_correct, seq_logproba = quizz_machine.compute_correctness(
                 c_quizzes,
                 models,
-                both_directions=args.both_directions,
-                deterministic_validation=not args.stochastic_validation,
+                bidirectional_validation=args.bidirectional_validation,
+                deterministic_validation=args.deterministic_validation,
             )
 
             for n, l in zip(nb_correct, seq_logproba):
@@ -502,6 +511,9 @@ log_string(f"nb_parameters {nb_parameters} ({int(nb_parameters/1e6)}M)")
 for n_epoch in range(args.nb_epochs):
     log_string(f"--- epoch {n_epoch} ----------------------------------------")
 
+    cta = " ".join([f"{float(m.main_test_accuracy):.04f}" for m in models])
+    log_string(f"current_test_accuracies {cta}")
+
     # Select, improve, and eval the worst model
 
     weakest_model = min(models, key=lambda m: float(m.main_test_accuracy))
@@ -522,9 +534,6 @@ for n_epoch in range(args.nb_epochs):
         f"test_set_composition w_quizzes {quizz_machine.nb_batch_w_quizzes} c_quizzes {quizz_machine.nb_batch_c_quizzes}"
     )
 
-    cta = " ".join([f"{float(m.main_test_accuracy):.04f}" for m in models])
-    log_string(f"current_test_accuracies {cta}")
-
     # Replace a fraction of the w_quizzes with fresh ones
 
     quizz_machine.renew_w_quizzes(args.nb_train_samples // args.nb_gpts)