End of training

all_results.json

@@ -1,11 +1,11 @@
 {
     "epoch": 5.0,
-    "eval_loss": 5474.208984375,
-    "eval_runtime": 2.3559,
-    "eval_samples_per_second": 13.583,
-    "eval_steps_per_second": 1.698,
-    "train_loss": 4986.834765625,
-    "train_runtime": 89.7426,
-    "train_samples_per_second": 1.783,
-    "train_steps_per_second": 0.223
+    "eval_loss": 6787.48388671875,
+    "eval_runtime": 2.1947,
+    "eval_samples_per_second": 14.58,
+    "eval_steps_per_second": 1.823,
+    "train_loss": 6234.263671875,
+    "train_runtime": 98.3256,
+    "train_samples_per_second": 1.627,
+    "train_steps_per_second": 0.203
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 5.0,
-    "eval_loss": 5474.208984375,
-    "eval_runtime": 2.3559,
-    "eval_samples_per_second": 13.583,
-    "eval_steps_per_second": 1.698
+    "eval_loss": 6787.48388671875,
+    "eval_runtime": 2.1947,
+    "eval_samples_per_second": 14.58,
+    "eval_steps_per_second": 1.823
 }

runs/Jul19_09-26-30_amy-2-gpu/events.out.tfevents.1689758914.amy-2-gpu.65271.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e68a853e459590d055a3178d151a8352e199e4bedeb72327642d9dc66ec1591e
+size 354

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 5.0,
-    "train_loss": 4986.834765625,
-    "train_runtime": 89.7426,
-    "train_samples_per_second": 1.783,
-    "train_steps_per_second": 0.223
+    "train_loss": 6234.263671875,
+    "train_runtime": 98.3256,
+    "train_samples_per_second": 1.627,
+    "train_steps_per_second": 0.203
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 5474.208984375,
+  "best_metric": 6787.48388671875,
   "best_model_checkpoint": "./coco_outputs/checkpoint-20",
   "epoch": 5.0,
   "global_step": 20,
@@ -9,64 +9,64 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 5475.6171875,
-      "eval_runtime": 2.202,
-      "eval_samples_per_second": 14.532,
-      "eval_steps_per_second": 1.816,
+      "eval_loss": 6788.2998046875,
+      "eval_runtime": 2.0726,
+      "eval_samples_per_second": 15.44,
+      "eval_steps_per_second": 1.93,
       "step": 4
     },
     {
       "epoch": 2.0,
-      "eval_loss": 5475.1103515625,
-      "eval_runtime": 2.2765,
-      "eval_samples_per_second": 14.056,
-      "eval_steps_per_second": 1.757,
+      "eval_loss": 6787.97509765625,
+      "eval_runtime": 2.0833,
+      "eval_samples_per_second": 15.361,
+      "eval_steps_per_second": 1.92,
       "step": 8
     },
     {
       "epoch": 2.5,
       "learning_rate": 1e-05,
-      "loss": 4957.6352,
+      "loss": 6175.932,
       "step": 10
     },
     {
       "epoch": 3.0,
-      "eval_loss": 5474.685546875,
-      "eval_runtime": 2.2499,
-      "eval_samples_per_second": 14.223,
-      "eval_steps_per_second": 1.778,
+      "eval_loss": 6787.7900390625,
+      "eval_runtime": 2.1929,
+      "eval_samples_per_second": 14.592,
+      "eval_steps_per_second": 1.824,
       "step": 12
     },
     {
       "epoch": 4.0,
-      "eval_loss": 5474.34375,
-      "eval_runtime": 2.2284,
-      "eval_samples_per_second": 14.36,
-      "eval_steps_per_second": 1.795,
+      "eval_loss": 6787.5068359375,
+      "eval_runtime": 2.0906,
+      "eval_samples_per_second": 15.307,
+      "eval_steps_per_second": 1.913,
       "step": 16
     },
     {
       "epoch": 5.0,
       "learning_rate": 0.0,
-      "loss": 5016.0344,
+      "loss": 6292.5953,
       "step": 20
     },
     {
       "epoch": 5.0,
-      "eval_loss": 5474.208984375,
-      "eval_runtime": 2.3127,
-      "eval_samples_per_second": 13.837,
-      "eval_steps_per_second": 1.73,
+      "eval_loss": 6787.48388671875,
+      "eval_runtime": 2.2266,
+      "eval_samples_per_second": 14.372,
+      "eval_steps_per_second": 1.796,
       "step": 20
     },
     {
       "epoch": 5.0,
       "step": 20,
       "total_flos": 7.65002115072e+16,
-      "train_loss": 4986.834765625,
-      "train_runtime": 89.7426,
-      "train_samples_per_second": 1.783,
-      "train_steps_per_second": 0.223
+      "train_loss": 6234.263671875,
+      "train_runtime": 98.3256,
+      "train_samples_per_second": 1.627,
+      "train_steps_per_second": 0.203
     }
   ],
   "max_steps": 20,