Update.

[mygptrnn.git] / main.py

diff --git a/main.py b/main.py
index cae20f8..c51035c 100755 (executable)
--- a/main.py
+++ b/main.py
@@ -107,8 +107,6 @@ parser.add_argument("--caterpillar_height", type=int, default=None)
 
 parser.add_argument("--rho", type=float, default=0.0)
 
-parser.add_argument("--dim_rec_v", type=int, default=None)
-
 parser.add_argument("--nb_blocks", type=int, default=None)
 
 parser.add_argument("--dropout", type=float, default=0.1)
@@ -119,7 +117,7 @@ parser.add_argument("--deterministic_synthesis", action="store_true", default=Fa
 
 parser.add_argument("--no_checkpoint", action="store_true", default=False)
 
-parser.add_argument("--overwrite_results", action="store_true", default=False)
+parser.add_argument("--continue_training", action="store_true", default=False)
 
 parser.add_argument("--checkpoint_name", type=str, default="checkpoint.pth")
 
@@ -332,7 +330,6 @@ default_model_args = {
         "dim_keys": 32,
         "dim_hidden": 32,
         "nb_heads": 2,
-        "dim_rec_v": 16,
         "nb_blocks": 2,
     },
     "17K-C": {
@@ -343,7 +340,6 @@ default_model_args = {
         "nb_heads": 2,
         "nb_lines": 16,
         "caterpillar_height": 4,
-        "dim_rec_v": 16,
         "nb_blocks": 2,
     },
     "4M": {
@@ -352,7 +348,6 @@ default_model_args = {
         "dim_keys": 32,
         "dim_hidden": 1024,
         "nb_heads": 4,
-        "dim_rec_v": 64,
         "nb_blocks": 6,
     },
     "4M-C": {
@@ -363,7 +358,6 @@ default_model_args = {
         "nb_heads": 4,
         "nb_lines": 32,
         "caterpillar_height": 4,
-        "dim_rec_v": 64,  # dim_model / nb_heads
         "nb_blocks": 6,
     },
     "37M": {
@@ -372,7 +366,6 @@ default_model_args = {
         "dim_keys": 64,
         "dim_hidden": 2048,
         "nb_heads": 8,
-        "dim_rec_v": 64,
         "nb_blocks": 12,
     },
     "37M-C": {
@@ -383,7 +376,6 @@ default_model_args = {
         "nb_heads": 8,
         "nb_lines": 256,
         "caterpillar_height": 32,
-        "dim_rec_v": 64,
         "nb_blocks": 12,
     },
     "122M": {
@@ -392,7 +384,6 @@ default_model_args = {
         "dim_keys": 64,
         "dim_hidden": 2048,
         "nb_heads": 8,
-        "dim_rec_v": 96,
         "nb_blocks": 24,
     },
     "122M-C": {
@@ -402,7 +393,6 @@ default_model_args = {
         "dim_hidden": 2048,
         "nb_heads": 8,
         "nb_lines": 128,
-        "dim_rec_v": 96,
         "nb_blocks": 24,
     },
     "352M": {
@@ -411,7 +401,6 @@ default_model_args = {
         "dim_keys": 64,
         "dim_hidden": 2048,
         "nb_heads": 8,
-        "dim_rec_v": 128,
         "nb_blocks": 48,
     },
     "352M-C": {
@@ -421,7 +410,6 @@ default_model_args = {
         "dim_hidden": 2048,
         "nb_heads": 8,
         "nb_lines": 128,
-        "dim_rec_v": 128,
         "nb_blocks": 48,
     },
 }
@@ -438,7 +426,7 @@ else:
 try:
     os.mkdir(args.result_dir)
 except FileExistsError:
-    if not args.overwrite_results:
+    if not args.continue_training:
         print(f"result directory {args.result_dir} already exists")
         exit(1)
 
@@ -736,7 +724,6 @@ model = mygpt.MyGPT(
     nb_heads=args.nb_heads,
     nb_lines=args.nb_lines,
     caterpillar_height=args.caterpillar_height,
-    dim_rec_v=args.dim_rec_v,
     nb_blocks=args.nb_blocks,
     causal=True,
     dropout=args.dropout,