End of training

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 1.0,
-    "eval_loss": 0.6321855187416077,
-    "eval_runtime": 71.348,
-    "eval_samples": 250,
-    "eval_samples_per_second": 3.504,
-    "eval_steps_per_second": 0.224,
-    "eval_wer": 45.68657478305258,
-    "train_loss": 1.2054000387191772,
-    "train_runtime": 15609.5124,
-    "train_samples_per_second": 4.1,
-    "train_steps_per_second": 0.064
+    "eval_loss": 0.6313478946685791,
+    "eval_runtime": 711.2452,
+    "eval_samples": 500,
+    "eval_samples_per_second": 0.703,
+    "eval_steps_per_second": 0.045,
+    "eval_wer": 45.29159303206261,
+    "train_loss": 1.2053918581008911,
+    "train_runtime": 78451.7893,
+    "train_samples_per_second": 0.816,
+    "train_steps_per_second": 0.013
 }

eval_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 1.0,
-    "eval_loss": 0.6321855187416077,
-    "eval_runtime": 71.348,
-    "eval_samples": 250,
-    "eval_samples_per_second": 3.504,
-    "eval_steps_per_second": 0.224,
-    "eval_wer": 45.68657478305258
+    "eval_loss": 0.6313478946685791,
+    "eval_runtime": 711.2452,
+    "eval_samples": 500,
+    "eval_samples_per_second": 0.703,
+    "eval_steps_per_second": 0.045,
+    "eval_wer": 45.29159303206261
 }

runs/Mar29_09-18-59_terra/events.out.tfevents.1680160948.terra.179326.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a0597c554d7d50a6ca7d2c2acbd35225e4bbdb209c21ee5b1f23748eed6c687
+size 406

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 1.0,
-    "train_loss": 1.2054000387191772,
-    "train_runtime": 15609.5124,
-    "train_samples_per_second": 4.1,
-    "train_steps_per_second": 0.064
+    "train_loss": 1.2053918581008911,
+    "train_runtime": 78451.7893,
+    "train_samples_per_second": 0.816,
+    "train_steps_per_second": 0.013
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 45.68657478305258,
+  "best_metric": 45.29159303206261,
   "best_model_checkpoint": "./output/checkpoint-1000",
   "epoch": 1.0,
   "global_step": 1000,
@@ -69,11 +69,11 @@
     },
     {
       "epoch": 0.25,
-      "eval_loss": 1.5133659839630127,
-      "eval_runtime": 49.0888,
-      "eval_samples_per_second": 5.093,
-      "eval_steps_per_second": 0.326,
-      "eval_wer": 79.0199081163859,
+      "eval_loss": 1.4900375604629517,
+      "eval_runtime": 696.9939,
+      "eval_samples_per_second": 0.717,
+      "eval_steps_per_second": 0.046,
+      "eval_wer": 77.9348649330977,
       "step": 250
     },
     {
@@ -85,13 +85,13 @@
     {
       "epoch": 0.3,
       "learning_rate": 5.9600000000000005e-06,
-      "loss": 1.3248,
+      "loss": 1.3247,
       "step": 300
     },
     {
       "epoch": 0.33,
       "learning_rate": 6.460000000000001e-06,
-      "loss": 1.2724,
+      "loss": 1.2725,
       "step": 325
     },
     {
@@ -103,7 +103,7 @@
     {
       "epoch": 0.38,
       "learning_rate": 7.4600000000000006e-06,
-      "loss": 1.0426,
+      "loss": 1.0425,
       "step": 375
     },
     {
@@ -133,16 +133,16 @@
     {
       "epoch": 0.5,
       "learning_rate": 9.960000000000001e-06,
-      "loss": 0.7719,
+      "loss": 0.7722,
       "step": 500
     },
     {
       "epoch": 0.5,
-      "eval_loss": 0.8723946213722229,
-      "eval_runtime": 42.8226,
-      "eval_samples_per_second": 5.838,
-      "eval_steps_per_second": 0.374,
-      "eval_wer": 61.20469627360898,
+      "eval_loss": 0.8731086850166321,
+      "eval_runtime": 702.8815,
+      "eval_samples_per_second": 0.711,
+      "eval_steps_per_second": 0.046,
+      "eval_wer": 60.28780610956829,
       "step": 500
     },
     {
@@ -154,7 +154,7 @@
     {
       "epoch": 0.55,
       "learning_rate": 9.040000000000002e-06,
-      "loss": 0.7805,
+      "loss": 0.7804,
       "step": 550
     },
     {
@@ -166,7 +166,7 @@
     {
       "epoch": 0.6,
       "learning_rate": 8.040000000000001e-06,
-      "loss": 0.8321,
+      "loss": 0.832,
       "step": 600
     },
     {
@@ -196,7 +196,7 @@
     {
       "epoch": 0.72,
       "learning_rate": 5.540000000000001e-06,
-      "loss": 0.7312,
+      "loss": 0.731,
       "step": 725
     },
     {
@@ -207,11 +207,11 @@
     },
     {
       "epoch": 0.75,
-      "eval_loss": 0.6546878814697266,
-      "eval_runtime": 41.6435,
-      "eval_samples_per_second": 6.003,
-      "eval_steps_per_second": 0.384,
-      "eval_wer": 47.32006125574272,
+      "eval_loss": 0.6534215211868286,
+      "eval_runtime": 681.508,
+      "eval_samples_per_second": 0.734,
+      "eval_steps_per_second": 0.047,
+      "eval_wer": 46.50340823024489,
       "step": 750
     },
     {
@@ -223,19 +223,19 @@
     {
       "epoch": 0.8,
       "learning_rate": 4.04e-06,
-      "loss": 0.8042,
+      "loss": 0.8043,
       "step": 800
     },
     {
       "epoch": 0.82,
       "learning_rate": 3.54e-06,
-      "loss": 0.6615,
+      "loss": 0.6616,
       "step": 825
     },
     {
       "epoch": 0.85,
       "learning_rate": 3.04e-06,
-      "loss": 0.663,
+      "loss": 0.6631,
       "step": 850
     },
     {
@@ -265,32 +265,32 @@
     {
       "epoch": 0.97,
       "learning_rate": 5.4e-07,
-      "loss": 0.5841,
+      "loss": 0.5842,
       "step": 975
     },
     {
       "epoch": 1.0,
       "learning_rate": 4e-08,
-      "loss": 0.5734,
+      "loss": 0.5733,
       "step": 1000
     },
     {
       "epoch": 1.0,
-      "eval_loss": 0.6321855187416077,
-      "eval_runtime": 76.4524,
-      "eval_samples_per_second": 3.27,
-      "eval_steps_per_second": 0.209,
-      "eval_wer": 45.68657478305258,
+      "eval_loss": 0.6313478946685791,
+      "eval_runtime": 708.6015,
+      "eval_samples_per_second": 0.706,
+      "eval_steps_per_second": 0.045,
+      "eval_wer": 45.29159303206261,
       "step": 1000
     },
     {
       "epoch": 1.0,
       "step": 1000,
       "total_flos": 1.57560864768e+18,
-      "train_loss": 1.2054000387191772,
-      "train_runtime": 15609.5124,
-      "train_samples_per_second": 4.1,
-      "train_steps_per_second": 0.064
+      "train_loss": 1.2053918581008911,
+      "train_runtime": 78451.7893,
+      "train_samples_per_second": 0.816,
+      "train_steps_per_second": 0.013
     }
   ],
   "max_steps": 1000,