End of training

all_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 8.0,
-    "total_flos": 3.7940453059015475e+17,
-    "train_loss": 0.10690519699813626,
-    "train_runtime": 1502.7914,
+    "epoch": 7.0,
+    "total_flos": 3.9803472298849075e+17,
+    "train_loss": 0.11263628595385025,
+    "train_runtime": 1312.5646,
     "train_samples": 9000,
-    "train_samples_per_second": 119.777,
-    "train_steps_per_second": 3.753
+    "train_samples_per_second": 137.136,
+    "train_steps_per_second": 4.297
 }

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 8.0,
-    "total_flos": 3.7940453059015475e+17,
-    "train_loss": 0.10690519699813626,
-    "train_runtime": 1502.7914,
+    "epoch": 7.0,
+    "total_flos": 3.9803472298849075e+17,
+    "train_loss": 0.11263628595385025,
+    "train_runtime": 1312.5646,
     "train_samples": 9000,
-    "train_samples_per_second": 119.777,
-    "train_steps_per_second": 3.753
+    "train_samples_per_second": 137.136,
+    "train_steps_per_second": 4.297
 }

trainer_state.json

@@ -1,129 +1,112 @@
 {
-  "best_metric": 0.7964841108857336,
-  "best_model_checkpoint": "logs/ecthr_b/MHGanainy/xmod-shared-roberta-base-legal-multi/seed_1/checkpoint-1410",
-  "epoch": 8.0,
+  "best_metric": 0.7924403183023873,
+  "best_model_checkpoint": "logs/ecthr_b/MHGanainy/xmod-roberta-base-legal-multi/seed_1/checkpoint-1128",
+  "epoch": 7.0,
   "eval_steps": 500,
-  "global_step": 2256,
+  "global_step": 1974,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 0.1879543662071228,
-      "eval_macro-f1": 0.637208986436089,
-      "eval_micro-f1": 0.7414155571128241,
-      "eval_runtime": 9.9431,
-      "eval_samples_per_second": 100.573,
-      "eval_steps_per_second": 3.218,
+      "eval_loss": 0.18594583868980408,
+      "eval_macro-f1": 0.6610565197291588,
+      "eval_micro-f1": 0.7678571428571429,
+      "eval_runtime": 9.434,
+      "eval_samples_per_second": 105.999,
+      "eval_steps_per_second": 3.392,
       "step": 282
     },
     {
       "epoch": 1.773049645390071,
-      "grad_norm": 1.1033704280853271,
+      "grad_norm": 1.6924870014190674,
       "learning_rate": 2.7345744680851065e-05,
-      "loss": 0.1761,
+      "loss": 0.1734,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_loss": 0.19721177220344543,
-      "eval_macro-f1": 0.6534255535190332,
-      "eval_micro-f1": 0.76043673731535,
-      "eval_runtime": 13.6625,
-      "eval_samples_per_second": 73.193,
-      "eval_steps_per_second": 2.342,
+      "eval_loss": 0.204640731215477,
+      "eval_macro-f1": 0.6510354952925879,
+      "eval_micro-f1": 0.7552314521242867,
+      "eval_runtime": 10.5166,
+      "eval_samples_per_second": 95.088,
+      "eval_steps_per_second": 3.043,
       "step": 564
     },
     {
       "epoch": 3.0,
-      "eval_loss": 0.1727759689092636,
-      "eval_macro-f1": 0.7189646935925815,
-      "eval_micro-f1": 0.7865168539325843,
-      "eval_runtime": 9.5271,
-      "eval_samples_per_second": 104.964,
-      "eval_steps_per_second": 3.359,
+      "eval_loss": 0.17442859709262848,
+      "eval_macro-f1": 0.7103393399151534,
+      "eval_micro-f1": 0.7856437273625289,
+      "eval_runtime": 9.5348,
+      "eval_samples_per_second": 104.879,
+      "eval_steps_per_second": 3.356,
       "step": 846
     },
     {
       "epoch": 3.546099290780142,
-      "grad_norm": 1.1534513235092163,
+      "grad_norm": 0.9688987135887146,
       "learning_rate": 2.4686170212765958e-05,
-      "loss": 0.1155,
+      "loss": 0.1145,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_loss": 0.1755395233631134,
-      "eval_macro-f1": 0.7373513816426139,
-      "eval_micro-f1": 0.7873910127431254,
-      "eval_runtime": 10.0776,
-      "eval_samples_per_second": 99.23,
-      "eval_steps_per_second": 3.175,
+      "eval_loss": 0.18016603589057922,
+      "eval_macro-f1": 0.739214459676758,
+      "eval_micro-f1": 0.7924403183023873,
+      "eval_runtime": 10.4478,
+      "eval_samples_per_second": 95.714,
+      "eval_steps_per_second": 3.063,
       "step": 1128
     },
     {
       "epoch": 5.0,
-      "eval_loss": 0.17423860728740692,
-      "eval_macro-f1": 0.7353822537422677,
-      "eval_micro-f1": 0.7964841108857336,
-      "eval_runtime": 10.1727,
-      "eval_samples_per_second": 98.302,
-      "eval_steps_per_second": 3.146,
+      "eval_loss": 0.19197727739810944,
+      "eval_macro-f1": 0.721481034377041,
+      "eval_micro-f1": 0.7874427730542839,
+      "eval_runtime": 10.2803,
+      "eval_samples_per_second": 97.273,
+      "eval_steps_per_second": 3.113,
       "step": 1410
     },
     {
       "epoch": 5.319148936170213,
-      "grad_norm": 1.2388067245483398,
-      "learning_rate": 2.2026595744680854e-05,
-      "loss": 0.091,
+      "grad_norm": 0.9491854310035706,
+      "learning_rate": 2.203191489361702e-05,
+      "loss": 0.0906,
       "step": 1500
     },
     {
       "epoch": 6.0,
-      "eval_loss": 0.19554495811462402,
-      "eval_macro-f1": 0.7451892220071915,
-      "eval_micro-f1": 0.7875816993464052,
-      "eval_runtime": 13.6476,
-      "eval_samples_per_second": 73.273,
-      "eval_steps_per_second": 2.345,
+      "eval_loss": 0.19124138355255127,
+      "eval_macro-f1": 0.7336639771212009,
+      "eval_micro-f1": 0.7867017774851877,
+      "eval_runtime": 13.0899,
+      "eval_samples_per_second": 76.395,
+      "eval_steps_per_second": 2.445,
       "step": 1692
     },
     {
       "epoch": 7.0,
-      "eval_loss": 0.1909777820110321,
-      "eval_macro-f1": 0.7475109109665823,
-      "eval_micro-f1": 0.7894201424211598,
-      "eval_runtime": 9.5034,
-      "eval_samples_per_second": 105.225,
-      "eval_steps_per_second": 3.367,
+      "eval_loss": 0.1959591656923294,
+      "eval_macro-f1": 0.7411436475174427,
+      "eval_micro-f1": 0.7906208718626155,
+      "eval_runtime": 9.3742,
+      "eval_samples_per_second": 106.676,
+      "eval_steps_per_second": 3.414,
       "step": 1974
     },
     {
-      "epoch": 7.092198581560283,
-      "grad_norm": 1.6909549236297607,
-      "learning_rate": 1.9367021276595743e-05,
-      "loss": 0.071,
-      "step": 2000
-    },
-    {
-      "epoch": 8.0,
-      "eval_loss": 0.2065315544605255,
-      "eval_macro-f1": 0.7429317897586639,
-      "eval_micro-f1": 0.785572468563865,
-      "eval_runtime": 9.4369,
-      "eval_samples_per_second": 105.967,
-      "eval_steps_per_second": 3.391,
-      "step": 2256
-    },
-    {
-      "epoch": 8.0,
-      "step": 2256,
-      "total_flos": 3.7940453059015475e+17,
-      "train_loss": 0.10690519699813626,
-      "train_runtime": 1502.7914,
-      "train_samples_per_second": 119.777,
-      "train_steps_per_second": 3.753
+      "epoch": 7.0,
+      "step": 1974,
+      "total_flos": 3.9803472298849075e+17,
+      "train_loss": 0.11263628595385025,
+      "train_runtime": 1312.5646,
+      "train_samples_per_second": 137.136,
+      "train_steps_per_second": 4.297
     }
   ],
   "logging_steps": 500,
@@ -152,7 +135,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3.7940453059015475e+17,
+  "total_flos": 3.9803472298849075e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null