End of training

all_results.json

@@ -1,14 +1,14 @@
 {
     "epoch": 8.0,
-    "eval_loss": 2.015979766845703,
-    "eval_runtime": 80.6758,
+    "eval_loss": 2.0533692836761475,
+    "eval_runtime": 80.7384,
     "eval_samples": 5193,
-    "eval_samples_per_second": 64.369,
-    "eval_steps_per_second": 8.057,
-    "perplexity": 7.508079946521416,
-    "train_loss": 2.304505779894719,
-    "train_runtime": 46564.1697,
+    "eval_samples_per_second": 64.319,
+    "eval_steps_per_second": 8.051,
+    "perplexity": 7.794117509150382,
+    "train_loss": 2.459805450241544,
+    "train_runtime": 46512.8854,
     "train_samples": 98673,
-    "train_samples_per_second": 16.953,
+    "train_samples_per_second": 16.971,
     "train_steps_per_second": 0.265
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 8.0,
-    "eval_loss": 2.015979766845703,
-    "eval_runtime": 80.6758,
+    "eval_loss": 2.0533692836761475,
+    "eval_runtime": 80.7384,
     "eval_samples": 5193,
-    "eval_samples_per_second": 64.369,
-    "eval_steps_per_second": 8.057,
-    "perplexity": 7.508079946521416
+    "eval_samples_per_second": 64.319,
+    "eval_steps_per_second": 8.051,
+    "perplexity": 7.794117509150382
 }

runs/Dec18_09-11-55_jbuhmann/events.out.tfevents.1639818736.jbuhmann.443.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98a51531722606d55beb72e17dbf4ee9f73b50511acfcd55381bff89cdcbb2ee
-size 9119
+oid sha256:5af5b1309cd2b9f6ac1e6f858c369372d6a5da1864cbdce2bcdfa53c286c951b
+size 9473

runs/Dec18_09-11-55_jbuhmann/events.out.tfevents.1639865330.jbuhmann.443.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11d7cfe4a143d2ad0785d312e9afb4796f2848423a07d7547713df83a780c7d8
+size 311

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 8.0,
-    "train_loss": 2.304505779894719,
-    "train_runtime": 46564.1697,
+    "train_loss": 2.459805450241544,
+    "train_runtime": 46512.8854,
     "train_samples": 98673,
-    "train_samples_per_second": 16.953,
+    "train_samples_per_second": 16.971,
     "train_steps_per_second": 0.265
 }

trainer_state.json

@@ -10,218 +10,218 @@
     {
       "epoch": 0.32,
       "learning_rate": 1.9189364461738006e-05,
-      "loss": 2.5752,
+      "loss": 2.8003,
       "step": 500
     },
     {
       "epoch": 0.65,
       "learning_rate": 1.8378728923476004e-05,
-      "loss": 2.4758,
+      "loss": 2.6656,
       "step": 1000
     },
     {
       "epoch": 0.97,
       "learning_rate": 1.756809338521401e-05,
-      "loss": 2.4695,
+      "loss": 2.6501,
       "step": 1500
     },
     {
       "epoch": 1.0,
-      "eval_loss": 2.0689399242401123,
-      "eval_runtime": 81.3366,
-      "eval_samples_per_second": 63.846,
-      "eval_steps_per_second": 7.991,
+      "eval_loss": 2.1010942459106445,
+      "eval_runtime": 81.2019,
+      "eval_samples_per_second": 63.952,
+      "eval_steps_per_second": 8.005,
       "step": 1542
     },
     {
       "epoch": 1.3,
       "learning_rate": 1.675745784695201e-05,
-      "loss": 2.4094,
+      "loss": 2.5775,
       "step": 2000
     },
     {
       "epoch": 1.62,
       "learning_rate": 1.5946822308690015e-05,
-      "loss": 2.3802,
+      "loss": 2.545,
       "step": 2500
     },
     {
       "epoch": 1.95,
       "learning_rate": 1.5136186770428017e-05,
-      "loss": 2.3622,
+      "loss": 2.5183,
       "step": 3000
     },
     {
       "epoch": 2.0,
-      "eval_loss": 2.0509471893310547,
-      "eval_runtime": 81.3625,
-      "eval_samples_per_second": 63.825,
-      "eval_steps_per_second": 7.989,
+      "eval_loss": 2.085500478744507,
+      "eval_runtime": 81.2927,
+      "eval_samples_per_second": 63.88,
+      "eval_steps_per_second": 7.996,
       "step": 3084
     },
     {
       "epoch": 2.27,
       "learning_rate": 1.432555123216602e-05,
-      "loss": 2.3547,
+      "loss": 2.5098,
       "step": 3500
     },
     {
       "epoch": 2.59,
       "learning_rate": 1.3514915693904023e-05,
-      "loss": 2.3319,
+      "loss": 2.4869,
       "step": 4000
     },
     {
       "epoch": 2.92,
       "learning_rate": 1.2704280155642024e-05,
-      "loss": 2.3303,
+      "loss": 2.4835,
       "step": 4500
     },
     {
       "epoch": 3.0,
-      "eval_loss": 2.0795907974243164,
-      "eval_runtime": 81.2667,
-      "eval_samples_per_second": 63.901,
-      "eval_steps_per_second": 7.998,
+      "eval_loss": 2.1157290935516357,
+      "eval_runtime": 81.2608,
+      "eval_samples_per_second": 63.905,
+      "eval_steps_per_second": 7.999,
       "step": 4626
     },
     {
       "epoch": 3.24,
       "learning_rate": 1.1893644617380028e-05,
-      "loss": 2.293,
+      "loss": 2.4413,
       "step": 5000
     },
     {
       "epoch": 3.57,
       "learning_rate": 1.108300907911803e-05,
-      "loss": 2.2819,
+      "loss": 2.4289,
       "step": 5500
     },
     {
       "epoch": 3.89,
       "learning_rate": 1.027237354085603e-05,
-      "loss": 2.2865,
+      "loss": 2.4372,
       "step": 6000
     },
     {
       "epoch": 4.0,
-      "eval_loss": 1.9920992851257324,
-      "eval_runtime": 81.5086,
-      "eval_samples_per_second": 63.711,
-      "eval_steps_per_second": 7.975,
+      "eval_loss": 2.0268728733062744,
+      "eval_runtime": 81.2557,
+      "eval_samples_per_second": 63.909,
+      "eval_steps_per_second": 7.999,
       "step": 6168
     },
     {
       "epoch": 4.22,
       "learning_rate": 9.461738002594033e-06,
-      "loss": 2.2856,
+      "loss": 2.4307,
       "step": 6500
     },
     {
       "epoch": 4.54,
       "learning_rate": 8.651102464332037e-06,
-      "loss": 2.2669,
+      "loss": 2.4153,
       "step": 7000
     },
     {
       "epoch": 4.86,
       "learning_rate": 7.84046692607004e-06,
-      "loss": 2.2598,
+      "loss": 2.4087,
       "step": 7500
     },
     {
       "epoch": 5.0,
-      "eval_loss": 1.9843554496765137,
-      "eval_runtime": 81.4242,
-      "eval_samples_per_second": 63.777,
-      "eval_steps_per_second": 7.983,
+      "eval_loss": 2.0193827152252197,
+      "eval_runtime": 81.2425,
+      "eval_samples_per_second": 63.92,
+      "eval_steps_per_second": 8.001,
       "step": 7710
     },
     {
       "epoch": 5.19,
       "learning_rate": 7.029831387808041e-06,
-      "loss": 2.2482,
+      "loss": 2.393,
       "step": 8000
     },
     {
       "epoch": 5.51,
       "learning_rate": 6.219195849546045e-06,
-      "loss": 2.2285,
+      "loss": 2.3759,
       "step": 8500
     },
     {
       "epoch": 5.84,
       "learning_rate": 5.4085603112840465e-06,
-      "loss": 2.2544,
+      "loss": 2.394,
       "step": 9000
     },
     {
       "epoch": 6.0,
-      "eval_loss": 2.0076751708984375,
-      "eval_runtime": 81.3076,
-      "eval_samples_per_second": 63.869,
-      "eval_steps_per_second": 7.994,
+      "eval_loss": 2.0468995571136475,
+      "eval_runtime": 81.2739,
+      "eval_samples_per_second": 63.895,
+      "eval_steps_per_second": 7.998,
       "step": 9252
     },
     {
       "epoch": 6.16,
       "learning_rate": 4.5979247730220496e-06,
-      "loss": 2.2357,
+      "loss": 2.3816,
       "step": 9500
     },
     {
       "epoch": 6.49,
       "learning_rate": 3.7872892347600522e-06,
-      "loss": 2.218,
+      "loss": 2.3603,
       "step": 10000
     },
     {
       "epoch": 6.81,
       "learning_rate": 2.9766536964980545e-06,
-      "loss": 2.2128,
+      "loss": 2.3576,
       "step": 10500
     },
     {
       "epoch": 7.0,
-      "eval_loss": 1.9913570880889893,
-      "eval_runtime": 81.2325,
-      "eval_samples_per_second": 63.928,
-      "eval_steps_per_second": 8.002,
+      "eval_loss": 2.029569625854492,
+      "eval_runtime": 81.1582,
+      "eval_samples_per_second": 63.986,
+      "eval_steps_per_second": 8.009,
       "step": 10794
     },
     {
       "epoch": 7.13,
       "learning_rate": 2.166018158236057e-06,
-      "loss": 2.2075,
+      "loss": 2.3497,
       "step": 11000
     },
     {
       "epoch": 7.46,
       "learning_rate": 1.3553826199740596e-06,
-      "loss": 2.1936,
+      "loss": 2.3399,
       "step": 11500
     },
     {
       "epoch": 7.78,
       "learning_rate": 5.447470817120623e-07,
-      "loss": 2.1963,
+      "loss": 2.34,
       "step": 12000
     },
     {
       "epoch": 8.0,
-      "eval_loss": 1.9882222414016724,
-      "eval_runtime": 81.2899,
-      "eval_samples_per_second": 63.883,
-      "eval_steps_per_second": 7.996,
+      "eval_loss": 2.0294876098632812,
+      "eval_runtime": 81.1819,
+      "eval_samples_per_second": 63.967,
+      "eval_steps_per_second": 8.007,
       "step": 12336
     },
     {
       "epoch": 8.0,
       "step": 12336,
       "total_flos": 2.0776857912786125e+17,
-      "train_loss": 2.304505779894719,
-      "train_runtime": 46564.1697,
-      "train_samples_per_second": 16.953,
+      "train_loss": 2.459805450241544,
+      "train_runtime": 46512.8854,
+      "train_samples_per_second": 16.971,
       "train_steps_per_second": 0.265
     }
   ],