Training in progress, epoch 3

checkpoint-652/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05b941a9d3d5f0904b9be07c1e13a18bae3fac4b04b65ace1eb83cbbb5b629ee
+oid sha256:52b188021dbfa241f0a678e35b7af59e11aff7057c3843277ef9933346f641a0
 size 2681485310

checkpoint-652/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:743c3cf66ebcaa120ff0414f34b1f2897e119bb20e1c6c2c08e5861da71cdcb0
+oid sha256:d8ec99e6ce268c28a5ab518a36acce6ef6019d14ec92ba4b6dfe673399644daf
 size 1340743917

checkpoint-652/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a157602c28dafa1a0271775fa975aab22e5621dd748103545f800233df8f5450
+oid sha256:37f37fdf2b898420edca00f388fedb6a726205e8670b9bf7173ceeb8aa3d80ff
 size 623

checkpoint-652/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.3068828284740448,
+  "best_metric": 0.30892524123191833,
   "best_model_checkpoint": "outputs/electra-nli-efl-tweeteval/checkpoint-652",
   "epoch": 3.9973231357552583,
   "global_step": 652,
@@ -9,71 +9,71 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "learning_rate": 9.499999999999999e-07,
-      "loss": 0.4384,
+      "learning_rate": 9e-07,
+      "loss": 0.439,
       "step": 163
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7444061962134251,
-      "eval_f1": 0.7308261375858633,
-      "eval_loss": 0.39615127444267273,
-      "eval_runtime": 9.0599,
-      "eval_samples_per_second": 256.514,
-      "eval_steps_per_second": 32.119,
+      "eval_accuracy": 0.745697074010327,
+      "eval_f1": 0.732240056847258,
+      "eval_loss": 0.3982622027397156,
+      "eval_runtime": 8.9237,
+      "eval_samples_per_second": 260.431,
+      "eval_steps_per_second": 32.61,
       "step": 163
     },
     {
       "epoch": 2.0,
-      "learning_rate": 9e-07,
-      "loss": 0.3447,
+      "learning_rate": 8e-07,
+      "loss": 0.3465,
       "step": 326
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.76592082616179,
-      "eval_f1": 0.7552159046464709,
-      "eval_loss": 0.3409559428691864,
-      "eval_runtime": 9.4378,
-      "eval_samples_per_second": 246.244,
-      "eval_steps_per_second": 30.833,
+      "eval_accuracy": 0.7620481927710844,
+      "eval_f1": 0.750740067157349,
+      "eval_loss": 0.3448249399662018,
+      "eval_runtime": 9.0059,
+      "eval_samples_per_second": 258.053,
+      "eval_steps_per_second": 32.312,
       "step": 326
     },
     {
       "epoch": 3.0,
-      "learning_rate": 8.499999999999999e-07,
-      "loss": 0.3057,
+      "learning_rate": 7e-07,
+      "loss": 0.3089,
       "step": 489
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.7749569707401033,
-      "eval_f1": 0.768808341108185,
-      "eval_loss": 0.32338443398475647,
-      "eval_runtime": 9.0418,
-      "eval_samples_per_second": 257.028,
-      "eval_steps_per_second": 32.184,
+      "eval_accuracy": 0.7693631669535284,
+      "eval_f1": 0.7633651185887134,
+      "eval_loss": 0.3303545117378235,
+      "eval_runtime": 9.3737,
+      "eval_samples_per_second": 247.927,
+      "eval_steps_per_second": 31.044,
       "step": 489
     },
     {
       "epoch": 4.0,
-      "learning_rate": 8e-07,
-      "loss": 0.287,
+      "learning_rate": 6e-07,
+      "loss": 0.2916,
       "step": 652
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.7857142857142857,
-      "eval_f1": 0.7778970154753132,
-      "eval_loss": 0.3068828284740448,
-      "eval_runtime": 9.3662,
-      "eval_samples_per_second": 248.128,
-      "eval_steps_per_second": 31.069,
+      "eval_accuracy": 0.7839931153184165,
+      "eval_f1": 0.7736843738845695,
+      "eval_loss": 0.30892524123191833,
+      "eval_runtime": 9.0737,
+      "eval_samples_per_second": 256.125,
+      "eval_steps_per_second": 32.071,
       "step": 652
     }
   ],
-  "max_steps": 3260,
-  "num_train_epochs": 20,
+  "max_steps": 1630,
+  "num_train_epochs": 10,
   "total_flos": 1.2184053565654464e+16,
   "trial_name": null,
   "trial_params": null

checkpoint-652/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82126e010e751131ea0bffbc21275298a6bdbc244157fe27eed5dd4b7b0d6e0f
+oid sha256:cd48e93c542d5f8f840918341d20dc98e6dbd60ec7052cf9f5610075d1655eaf
 size 3119

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c0974d1d7e3c00ca3ef5b4f10e74487753227c1cf0a5211b14949a292a65de3
+oid sha256:d8ec99e6ce268c28a5ab518a36acce6ef6019d14ec92ba4b6dfe673399644daf
 size 1340743917