Add training data

training_data.json

@@ -1,44 +1,43 @@
 {
   "train_losses": [
-    7049.438330078125,
-    5887.628120117188,
-    5277.642026367187,
-    4594.689287109375,
-    3546.2805078125,
-    1424.719789428711,
-    1127.6089764404296,
-    840.4840567016602,
-    688.1939489746094,
-    602.3454461669921,
-    562.8847897338867,
-    533.8131185913086,
-    522.2901638793945,
-    515.8702670288086,
-    507.2388348388672
+    6835.843090820313,
+    5569.674467773438,
+    4928.053500976563,
+    4278.040825195312,
+    3407.1733154296876,
+    1707.017879638672,
+    1678.5513446044922,
+    1690.340132446289,
+    1636.4783825683594,
+    1561.4304962158203,
+    1430.3591986083984,
+    1286.2933282470703,
+    1134.876439819336,
+    999.4317224121094,
+    912.4670324707031
   ],
   "test_losses": [
-    5279.287158203125,
-    4462.337194824218,
-    4125.337866210937,
-    3586.261376953125,
-    2795.8281372070314,
-    1348.6051147460937,
-    1711.0676147460938,
-    2111.490283203125,
-    2405.6600982666014,
-    2786.551837158203,
-    2962.615338134766,
-    3259.63317565918,
-    3414.8231872558595,
-    3715.608850097656,
-    3996.8146850585936
+    4925.1087646484375,
+    4224.336853027344,
+    3812.184326171875,
+    3274.811474609375,
+    2596.8716430664062,
+    1195.6449188232423,
+    1303.003598022461,
+    1479.3368133544923,
+    1686.8121704101563,
+    1969.0001953125,
+    2295.1075073242187,
+    2735.835479736328,
+    2956.7591918945313,
+    3227.5116333007813,
+    3502.591668701172
   ],
   "config": {
     "epochs": 15,
     "batch_size": 32,
     "learning_rate": 0.0005,
     "sequence_size": 32,
-    "include_inventory": false,
     "adaptive_weighting": {
       "initial_weight_emb": 1.5,
       "final_weight_emb": 0.0,
@@ -47,14 +46,16 @@
       "final_weight_raw": 1.0,
       "weight_raw_shape": "linear",
       "initial_kl_beta": 0.0001,
-      "final_kl_beta": 0.5,
+      "final_kl_beta": 0.4,
       "kl_beta_shape": "cosine",
       "warmup_epoch_ratio": 0.4
-    }
+    },
+    "total_correlation_beta_multiplier": 15.0,
+    "free_bits": 0.15
   },
-  "final_train_loss": 507.2388348388672,
-  "final_test_loss": 3996.8146850585936,
+  "final_train_loss": 912.4670324707031,
+  "final_test_loss": 3502.591668701172,
   "total_epochs": 15,
-  "best_train_loss": 507.2388348388672,
-  "best_test_loss": 1348.6051147460937
+  "best_train_loss": 912.4670324707031,
+  "best_test_loss": 1195.6449188232423
 }