Update.

[picoclvr.git] / main.py

diff --git a/main.py b/main.py
index 58e8046..19f918c 100755 (executable)
--- a/main.py
+++ b/main.py
@@ -14,6 +14,7 @@ import torch, torchvision
 from torch import nn
 from torch.nn import functional as F
 
+import ffutils
 import mygpt, tasks
 
 ######################################################################
@@ -34,8 +35,8 @@ parser = argparse.ArgumentParser(
 parser.add_argument(
     "--task",
     type=str,
-    default="picoclvr",
-    help="picoclvr, mnist, maze, snake, stack, expr, world",
+    default="sandbox",
+    help="sandbox, picoclvr, mnist, maze, snake, stack, expr, world",
 )
 
 parser.add_argument("--log_filename", type=str, default="train.log", help=" ")
@@ -81,6 +82,17 @@ parser.add_argument("--checkpoint_name", type=str, default="checkpoint.pth")
 ##############################
 # picoclvr options
 
+parser.add_argument("--sandbox_level", type=int, default=0)
+
+parser.add_argument("--sandbox_levels_nb_items", type=int, default=25)
+
+parser.add_argument("--sandbox_levels_len_source", type=int, default=6)
+
+parser.add_argument("--sandbox_levels_len_result", type=int, default=8)
+
+##############################
+# picoclvr options
+
 parser.add_argument("--picoclvr_nb_colors", type=int, default=5)
 
 parser.add_argument("--picoclvr_height", type=int, default=12)
@@ -136,7 +148,7 @@ parser.add_argument("--expr_input_file", type=str, default=None)
 ##############################
 # World options
 
-parser.add_argument("--world_vqae_nb_epochs", type=int, default=10)
+parser.add_argument("--world_vqae_nb_epochs", type=int, default=25)
 
 ######################################################################
 
@@ -150,6 +162,12 @@ if args.result_dir is None:
 ######################################################################
 
 default_args = {
+    "sandbox": {
+        "nb_epochs": 50,
+        "batch_size": 25,
+        "nb_train_samples": 100000,
+        "nb_test_samples": 10000,
+    },
     "picoclvr": {
         "nb_epochs": 25,
         "batch_size": 25,
@@ -187,9 +205,9 @@ default_args = {
         "nb_test_samples": 10000,
     },
     "world": {
-        "nb_epochs": 5,
+        "nb_epochs": 10,
         "batch_size": 25,
-        "nb_train_samples": 10000,
+        "nb_train_samples": 25000,
         "nb_test_samples": 1000,
     },
 }
@@ -257,7 +275,38 @@ picoclvr_pruner_eval = (
 
 ######################################################################
 
-if args.task == "picoclvr":
+if args.task == "sandbox":
+    if args.sandbox_level == 0:
+        problem = tasks.ProblemLevel0(
+            nb_sentences=args.sandbox_levels_nb_items,
+            len_prompt=args.sandbox_levels_len_source,
+            len_result=args.sandbox_levels_len_result,
+        )
+    elif args.sandbox_level == 1:
+        problem = tasks.ProblemLevel1(
+            nb_operators=args.sandbox_levels_nb_items,
+            len_source=args.sandbox_levels_len_source,
+            len_result=args.sandbox_levels_len_result,
+        )
+    elif args.sandbox_level == 2:
+        problem = tasks.ProblemLevel2(
+            len_source=args.sandbox_levels_len_source,
+            len_result=args.sandbox_levels_len_result,
+        )
+    else:
+        raise ValueError(f"Unknown sandbox level {args.sandbox_level}")
+
+    task = tasks.SandBox(
+        problem,
+        # tasks.ProblemAddition(zero_padded=False, inverted_result=False),
+        nb_train_samples=args.nb_train_samples,
+        nb_test_samples=args.nb_test_samples,
+        batch_size=args.batch_size,
+        logger=log_string,
+        device=device,
+    )
+
+elif args.task == "picoclvr":
     task = tasks.PicoCLVR(
         nb_train_samples=args.nb_train_samples,
         nb_test_samples=args.nb_test_samples,
@@ -334,6 +383,7 @@ elif args.task == "world":
         nb_test_samples=args.nb_test_samples,
         batch_size=args.batch_size,
         vqae_nb_epochs=args.world_vqae_nb_epochs,
+        logger=log_string,
         device=device,
     )