Training in progress, epoch 1

all_results.json

@@ -1,18 +1,18 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 80.61162079510703,
-    "eval_average_metrics": 80.61162079510703,
-    "eval_loss": 0.3976394534111023,
-    "eval_runtime": 4.7604,
-    "eval_samples_per_second": 343.462,
-    "test_accuracy": 79.38837920489297,
-    "test_average_metrics": 79.38837920489297,
-    "test_loss": 0.4527607262134552,
-    "test_runtime": 4.773,
-    "test_samples_per_second": 342.554,
-    "train_loss": 0.1054745042930215,
-    "train_runtime": 1312.1786,
+    "eval_accuracy": 72.47706422018348,
+    "eval_average_metrics": 72.47706422018348,
+    "eval_loss": 0.27757981419563293,
+    "eval_runtime": 4.8177,
+    "eval_samples_per_second": 339.375,
+    "test_accuracy": 70.58103975535168,
+    "test_average_metrics": 70.58103975535168,
+    "test_loss": 0.2894817888736725,
+    "test_runtime": 4.7931,
+    "test_samples_per_second": 341.116,
+    "train_loss": 0.2005054286374884,
+    "train_runtime": 1452.5651,
     "train_samples": 9427,
-    "train_samples_per_second": 143.685,
-    "train_steps_per_second": 4.496
+    "train_samples_per_second": 129.798,
+    "train_steps_per_second": 4.062
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 80.61162079510703,
-    "eval_average_metrics": 80.61162079510703,
-    "eval_loss": 0.3976394534111023,
-    "eval_runtime": 4.7604,
-    "eval_samples_per_second": 343.462
+    "eval_accuracy": 72.47706422018348,
+    "eval_average_metrics": 72.47706422018348,
+    "eval_loss": 0.27757981419563293,
+    "eval_runtime": 4.8177,
+    "eval_samples_per_second": 339.375
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70e0afe718876729ceb4662c13216a6f93ad8ab28e4515c3bc06c629b8768ed9
+oid sha256:8b456c85c073efe312e20263b9d0ea6131e6bd25a2ae380d0b5b9a05bb7e4db5
 size 7551621

runs/Jan30_12-18-00_node4/events.out.tfevents.1643516350.node4

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:52a02490ceee213ede6aacdf101a9afb25b9c228e85ccc45b1d795ef38241596
-size 12481
+oid sha256:8965555fd31d5b26e21ef072171bb2ff5a45a773d93493a7fb8265eceea65a1f
+size 12835

runs/Jan30_12-18-00_node4/events.out.tfevents.1643517807.node4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d42972784b2f348a6e72b42c05e6020b438c0d4730b32fcf94321b484386db6e
+size 684

runs/Jan30_13-16-54_node4/1643519896.1204708/events.out.tfevents.1643519896.node4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1df8a71ce578af80859a59a2a941e367844cc6dda48d25bc48803d2581cc216
+size 5020

runs/Jan30_13-16-54_node4/events.out.tfevents.1643519896.node4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6543fc3a7e5836cedb4df61702a6c42055513db13c315b8e422bd1118408297f
+size 4314

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "test_accuracy": 79.38837920489297,
-    "test_average_metrics": 79.38837920489297,
-    "test_loss": 0.4527607262134552,
-    "test_runtime": 4.773,
-    "test_samples_per_second": 342.554
+    "test_accuracy": 70.58103975535168,
+    "test_average_metrics": 70.58103975535168,
+    "test_loss": 0.2894817888736725,
+    "test_runtime": 4.7931,
+    "test_samples_per_second": 341.116
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.1054745042930215,
-    "train_runtime": 1312.1786,
+    "train_loss": 0.2005054286374884,
+    "train_runtime": 1452.5651,
     "train_samples": 9427,
-    "train_samples_per_second": 143.685,
-    "train_steps_per_second": 4.496
+    "train_samples_per_second": 129.798,
+    "train_steps_per_second": 4.062
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 80.61162079510703,
+  "best_metric": 72.47706422018348,
   "best_model_checkpoint": "outputs/adapter/superglue-boolq/checkpoint-3835",
   "epoch": 20.0,
   "global_step": 5900,
@@ -11,261 +11,261 @@
       "epoch": 1.0,
       "eval_accuracy": 62.62996941896024,
       "eval_average_metrics": 62.62996941896024,
-      "eval_loss": 0.2789694368839264,
-      "eval_runtime": 4.4788,
-      "eval_samples_per_second": 365.053,
+      "eval_loss": 0.2848173677921295,
+      "eval_runtime": 4.504,
+      "eval_samples_per_second": 363.014,
       "step": 295
     },
     {
       "epoch": 1.69,
       "learning_rate": 0.00027457627118644066,
-      "loss": 0.3259,
+      "loss": 0.3238,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 68.31804281345566,
-      "eval_average_metrics": 68.31804281345566,
-      "eval_loss": 0.2598051428794861,
-      "eval_runtime": 4.7399,
-      "eval_samples_per_second": 344.945,
+      "eval_accuracy": 62.62996941896024,
+      "eval_average_metrics": 62.62996941896024,
+      "eval_loss": 0.27746617794036865,
+      "eval_runtime": 4.5022,
+      "eval_samples_per_second": 363.153,
       "step": 590
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 74.6788990825688,
-      "eval_average_metrics": 74.6788990825688,
-      "eval_loss": 0.22243507206439972,
-      "eval_runtime": 4.7583,
-      "eval_samples_per_second": 343.612,
+      "eval_accuracy": 62.62996941896024,
+      "eval_average_metrics": 62.62996941896024,
+      "eval_loss": 0.2767568826675415,
+      "eval_runtime": 4.4934,
+      "eval_samples_per_second": 363.871,
       "step": 885
     },
     {
       "epoch": 3.39,
       "learning_rate": 0.00024915254237288135,
-      "loss": 0.2455,
+      "loss": 0.2896,
       "step": 1000
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 74.6788990825688,
-      "eval_average_metrics": 74.6788990825688,
-      "eval_loss": 0.21567188203334808,
-      "eval_runtime": 4.7503,
-      "eval_samples_per_second": 344.185,
+      "eval_accuracy": 62.75229357798165,
+      "eval_average_metrics": 62.75229357798165,
+      "eval_loss": 0.2767893373966217,
+      "eval_runtime": 4.7772,
+      "eval_samples_per_second": 342.253,
       "step": 1180
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 79.32721712538226,
-      "eval_average_metrics": 79.32721712538226,
-      "eval_loss": 0.2073381394147873,
-      "eval_runtime": 4.7337,
-      "eval_samples_per_second": 345.395,
+      "eval_accuracy": 63.36391437308868,
+      "eval_average_metrics": 63.36391437308868,
+      "eval_loss": 0.26716169714927673,
+      "eval_runtime": 4.7616,
+      "eval_samples_per_second": 343.37,
       "step": 1475
     },
     {
       "epoch": 5.08,
       "learning_rate": 0.000223728813559322,
-      "loss": 0.1872,
+      "loss": 0.2823,
       "step": 1500
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 78.71559633027523,
-      "eval_average_metrics": 78.71559633027523,
-      "eval_loss": 0.207603320479393,
-      "eval_runtime": 4.7412,
-      "eval_samples_per_second": 344.851,
+      "eval_accuracy": 66.11620795107034,
+      "eval_average_metrics": 66.11620795107034,
+      "eval_loss": 0.2606015205383301,
+      "eval_runtime": 4.7573,
+      "eval_samples_per_second": 343.683,
       "step": 1770
     },
     {
       "epoch": 6.78,
       "learning_rate": 0.0001983050847457627,
-      "loss": 0.134,
+      "loss": 0.2715,
       "step": 2000
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 80.0,
-      "eval_average_metrics": 80.0,
-      "eval_loss": 0.22447817027568817,
-      "eval_runtime": 4.7578,
-      "eval_samples_per_second": 343.643,
+      "eval_accuracy": 69.0519877675841,
+      "eval_average_metrics": 69.0519877675841,
+      "eval_loss": 0.25209498405456543,
+      "eval_runtime": 4.7545,
+      "eval_samples_per_second": 343.882,
       "step": 2065
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 78.83792048929664,
-      "eval_average_metrics": 78.83792048929664,
-      "eval_loss": 0.2642129361629486,
-      "eval_runtime": 4.7785,
-      "eval_samples_per_second": 342.156,
+      "eval_accuracy": 69.2354740061162,
+      "eval_average_metrics": 69.2354740061162,
+      "eval_loss": 0.25652435421943665,
+      "eval_runtime": 4.7648,
+      "eval_samples_per_second": 343.14,
       "step": 2360
     },
     {
       "epoch": 8.47,
       "learning_rate": 0.0001728813559322034,
-      "loss": 0.0978,
+      "loss": 0.236,
       "step": 2500
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 80.1223241590214,
-      "eval_average_metrics": 80.1223241590214,
-      "eval_loss": 0.26512306928634644,
-      "eval_runtime": 4.7549,
-      "eval_samples_per_second": 343.857,
+      "eval_accuracy": 71.0091743119266,
+      "eval_average_metrics": 71.0091743119266,
+      "eval_loss": 0.24900275468826294,
+      "eval_runtime": 4.7724,
+      "eval_samples_per_second": 342.596,
       "step": 2655
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 79.69418960244649,
-      "eval_average_metrics": 79.69418960244649,
-      "eval_loss": 0.2741082012653351,
-      "eval_runtime": 4.749,
-      "eval_samples_per_second": 344.283,
+      "eval_accuracy": 71.80428134556574,
+      "eval_average_metrics": 71.80428134556574,
+      "eval_loss": 0.2635628283023834,
+      "eval_runtime": 4.7656,
+      "eval_samples_per_second": 343.086,
       "step": 2950
     },
     {
       "epoch": 10.17,
       "learning_rate": 0.00014745762711864405,
-      "loss": 0.0743,
+      "loss": 0.2038,
       "step": 3000
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 80.42813455657493,
-      "eval_average_metrics": 80.42813455657493,
-      "eval_loss": 0.31817376613616943,
-      "eval_runtime": 4.7525,
-      "eval_samples_per_second": 344.028,
+      "eval_accuracy": 70.70336391437309,
+      "eval_average_metrics": 70.70336391437309,
+      "eval_loss": 0.26442670822143555,
+      "eval_runtime": 4.7895,
+      "eval_samples_per_second": 341.37,
       "step": 3245
     },
     {
       "epoch": 11.86,
       "learning_rate": 0.00012203389830508474,
-      "loss": 0.0541,
+      "loss": 0.1743,
       "step": 3500
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 79.81651376146789,
-      "eval_average_metrics": 79.81651376146789,
-      "eval_loss": 0.33367282152175903,
-      "eval_runtime": 4.7584,
-      "eval_samples_per_second": 343.602,
+      "eval_accuracy": 71.49847094801223,
+      "eval_average_metrics": 71.49847094801223,
+      "eval_loss": 0.2672346830368042,
+      "eval_runtime": 4.7696,
+      "eval_samples_per_second": 342.793,
       "step": 3540
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 80.61162079510703,
-      "eval_average_metrics": 80.61162079510703,
-      "eval_loss": 0.3976394534111023,
-      "eval_runtime": 4.7841,
-      "eval_samples_per_second": 341.758,
+      "eval_accuracy": 72.47706422018348,
+      "eval_average_metrics": 72.47706422018348,
+      "eval_loss": 0.27757981419563293,
+      "eval_runtime": 4.7935,
+      "eval_samples_per_second": 341.085,
       "step": 3835
     },
     {
       "epoch": 13.56,
       "learning_rate": 9.661016949152541e-05,
-      "loss": 0.0399,
+      "loss": 0.1493,
       "step": 4000
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 79.8776758409786,
-      "eval_average_metrics": 79.8776758409786,
-      "eval_loss": 0.41730642318725586,
-      "eval_runtime": 4.7467,
-      "eval_samples_per_second": 344.451,
+      "eval_accuracy": 70.8868501529052,
+      "eval_average_metrics": 70.8868501529052,
+      "eval_loss": 0.29447805881500244,
+      "eval_runtime": 4.7712,
+      "eval_samples_per_second": 342.681,
       "step": 4130
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 80.18348623853211,
-      "eval_average_metrics": 80.18348623853211,
-      "eval_loss": 0.4511031210422516,
-      "eval_runtime": 4.7544,
-      "eval_samples_per_second": 343.89,
+      "eval_accuracy": 70.21406727828746,
+      "eval_average_metrics": 70.21406727828746,
+      "eval_loss": 0.2965507209300995,
+      "eval_runtime": 4.7813,
+      "eval_samples_per_second": 341.96,
       "step": 4425
     },
     {
       "epoch": 15.25,
       "learning_rate": 7.11864406779661e-05,
-      "loss": 0.0303,
+      "loss": 0.1336,
       "step": 4500
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 79.63302752293579,
-      "eval_average_metrics": 79.63302752293579,
-      "eval_loss": 0.4779604375362396,
-      "eval_runtime": 4.7435,
-      "eval_samples_per_second": 344.682,
+      "eval_accuracy": 72.29357798165138,
+      "eval_average_metrics": 72.29357798165138,
+      "eval_loss": 0.3127536177635193,
+      "eval_runtime": 4.7677,
+      "eval_samples_per_second": 342.93,
       "step": 4720
     },
     {
       "epoch": 16.95,
       "learning_rate": 4.576271186440678e-05,
-      "loss": 0.0234,
+      "loss": 0.1166,
       "step": 5000
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 80.18348623853211,
-      "eval_average_metrics": 80.18348623853211,
-      "eval_loss": 0.5195048451423645,
-      "eval_runtime": 4.7472,
-      "eval_samples_per_second": 344.414,
+      "eval_accuracy": 70.9480122324159,
+      "eval_average_metrics": 70.9480122324159,
+      "eval_loss": 0.3478758931159973,
+      "eval_runtime": 4.7648,
+      "eval_samples_per_second": 343.144,
       "step": 5015
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 79.69418960244649,
-      "eval_average_metrics": 79.69418960244649,
-      "eval_loss": 0.5308667421340942,
-      "eval_runtime": 4.7851,
-      "eval_samples_per_second": 341.683,
+      "eval_accuracy": 72.17125382262996,
+      "eval_average_metrics": 72.17125382262996,
+      "eval_loss": 0.33715757727622986,
+      "eval_runtime": 4.7601,
+      "eval_samples_per_second": 343.478,
       "step": 5310
     },
     {
       "epoch": 18.64,
       "learning_rate": 2.0338983050847455e-05,
-      "loss": 0.019,
+      "loss": 0.1059,
       "step": 5500
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 80.06116207951071,
-      "eval_average_metrics": 80.06116207951071,
-      "eval_loss": 0.5453023910522461,
-      "eval_runtime": 4.7413,
-      "eval_samples_per_second": 344.844,
+      "eval_accuracy": 71.92660550458716,
+      "eval_average_metrics": 71.92660550458716,
+      "eval_loss": 0.34181272983551025,
+      "eval_runtime": 4.7537,
+      "eval_samples_per_second": 343.942,
       "step": 5605
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 80.36697247706422,
-      "eval_average_metrics": 80.36697247706422,
-      "eval_loss": 0.5537198185920715,
-      "eval_runtime": 4.7547,
-      "eval_samples_per_second": 343.869,
+      "eval_accuracy": 71.80428134556574,
+      "eval_average_metrics": 71.80428134556574,
+      "eval_loss": 0.34913983941078186,
+      "eval_runtime": 4.7707,
+      "eval_samples_per_second": 342.716,
       "step": 5900
     },
     {
       "epoch": 20.0,
       "step": 5900,
-      "total_flos": 5.80271991662592e+16,
-      "train_loss": 0.1054745042930215,
-      "train_runtime": 1312.1786,
-      "train_samples_per_second": 143.685,
-      "train_steps_per_second": 4.496
+      "total_flos": 5.78564930408256e+16,
+      "train_loss": 0.2005054286374884,
+      "train_runtime": 1452.5651,
+      "train_samples_per_second": 129.798,
+      "train_steps_per_second": 4.062
     }
   ],
   "max_steps": 5900,
   "num_train_epochs": 20,
-  "total_flos": 5.80271991662592e+16,
+  "total_flos": 5.78564930408256e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:035e52789313ccd7d5fc94e73b66a6867aa661d40c1539a49dae23bb7e66974e
+oid sha256:1ce74987ad12799538eb5feab0309115bbfdd2ebfb8a0396d86c0ca1ed29b7b7
 size 3183