End of training

all_results.json

@@ -1,11 +1,11 @@
 {
     "epoch": 20.0,
-    "eval_loss": 6387.95458984375,
-    "eval_runtime": 0.645,
-    "eval_samples_per_second": 12.403,
-    "eval_steps_per_second": 1.55,
-    "train_loss": 6247.897314453125,
-    "train_runtime": 365.3078,
-    "train_samples_per_second": 1.752,
-    "train_steps_per_second": 0.219
+    "eval_loss": 6395.94580078125,
+    "eval_runtime": 0.6002,
+    "eval_samples_per_second": 13.329,
+    "eval_steps_per_second": 1.666,
+    "train_loss": 6254.356591796875,
+    "train_runtime": 330.5487,
+    "train_samples_per_second": 1.936,
+    "train_steps_per_second": 0.242
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 20.0,
-    "eval_loss": 6387.95458984375,
-    "eval_runtime": 0.645,
-    "eval_samples_per_second": 12.403,
-    "eval_steps_per_second": 1.55
+    "eval_loss": 6395.94580078125,
+    "eval_runtime": 0.6002,
+    "eval_samples_per_second": 13.329,
+    "eval_steps_per_second": 1.666
 }

runs/Jul19_11-51-23_amy-2-gpu/events.out.tfevents.1689767842.amy-2-gpu.91834.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b32191095af367bbd256db8a4775adeeef563dfbc2df2da5b575b261eddfa78e
+size 354

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 20.0,
-    "train_loss": 6247.897314453125,
-    "train_runtime": 365.3078,
-    "train_samples_per_second": 1.752,
-    "train_steps_per_second": 0.219
+    "train_loss": 6254.356591796875,
+    "train_runtime": 330.5487,
+    "train_samples_per_second": 1.936,
+    "train_steps_per_second": 0.242
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 6387.95458984375,
-  "best_model_checkpoint": "./coco_outputs/checkpoint-80",
+  "best_metric": 6395.94580078125,
+  "best_model_checkpoint": "./coco_outputs/checkpoint-72",
   "epoch": 20.0,
   "global_step": 80,
   "is_hyper_param_search": false,
@@ -9,220 +9,220 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 6394.3193359375,
-      "eval_runtime": 0.6785,
-      "eval_samples_per_second": 11.791,
-      "eval_steps_per_second": 1.474,
+      "eval_loss": 6398.77734375,
+      "eval_runtime": 0.5241,
+      "eval_samples_per_second": 15.264,
+      "eval_steps_per_second": 1.908,
       "step": 4
     },
     {
       "epoch": 2.0,
-      "eval_loss": 6393.3564453125,
-      "eval_runtime": 0.625,
-      "eval_samples_per_second": 12.799,
-      "eval_steps_per_second": 1.6,
+      "eval_loss": 6398.435546875,
+      "eval_runtime": 0.5241,
+      "eval_samples_per_second": 15.265,
+      "eval_steps_per_second": 1.908,
       "step": 8
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.000175,
-      "loss": 6173.6062,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 6176.0926,
       "step": 10
     },
     {
       "epoch": 3.0,
-      "eval_loss": 6390.9482421875,
-      "eval_runtime": 0.6116,
-      "eval_samples_per_second": 13.08,
-      "eval_steps_per_second": 1.635,
+      "eval_loss": 6398.2265625,
+      "eval_runtime": 0.5251,
+      "eval_samples_per_second": 15.235,
+      "eval_steps_per_second": 1.904,
       "step": 12
     },
     {
       "epoch": 4.0,
-      "eval_loss": 6389.5869140625,
-      "eval_runtime": 0.5503,
-      "eval_samples_per_second": 14.538,
-      "eval_steps_per_second": 1.817,
+      "eval_loss": 6397.8564453125,
+      "eval_runtime": 0.5329,
+      "eval_samples_per_second": 15.013,
+      "eval_steps_per_second": 1.877,
       "step": 16
     },
     {
       "epoch": 5.0,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 6287.2859,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 6292.8656,
       "step": 20
     },
     {
       "epoch": 5.0,
-      "eval_loss": 6388.72802734375,
-      "eval_runtime": 0.5904,
-      "eval_samples_per_second": 13.55,
-      "eval_steps_per_second": 1.694,
+      "eval_loss": 6397.6298828125,
+      "eval_runtime": 0.5291,
+      "eval_samples_per_second": 15.121,
+      "eval_steps_per_second": 1.89,
       "step": 20
     },
     {
       "epoch": 6.0,
-      "eval_loss": 6388.6220703125,
-      "eval_runtime": 0.5392,
-      "eval_samples_per_second": 14.837,
-      "eval_steps_per_second": 1.855,
+      "eval_loss": 6397.31787109375,
+      "eval_runtime": 0.5319,
+      "eval_samples_per_second": 15.042,
+      "eval_steps_per_second": 1.88,
       "step": 24
     },
     {
       "epoch": 7.0,
-      "eval_loss": 6388.43994140625,
-      "eval_runtime": 0.5365,
-      "eval_samples_per_second": 14.91,
-      "eval_steps_per_second": 1.864,
+      "eval_loss": 6396.95263671875,
+      "eval_runtime": 0.5129,
+      "eval_samples_per_second": 15.596,
+      "eval_steps_per_second": 1.95,
       "step": 28
     },
     {
       "epoch": 7.5,
-      "learning_rate": 0.000125,
-      "loss": 6224.3344,
+      "learning_rate": 6.25e-06,
+      "loss": 6231.0688,
       "step": 30
     },
     {
       "epoch": 8.0,
-      "eval_loss": 6388.2119140625,
-      "eval_runtime": 0.5405,
-      "eval_samples_per_second": 14.802,
-      "eval_steps_per_second": 1.85,
+      "eval_loss": 6396.79052734375,
+      "eval_runtime": 0.5163,
+      "eval_samples_per_second": 15.494,
+      "eval_steps_per_second": 1.937,
       "step": 32
     },
     {
       "epoch": 9.0,
-      "eval_loss": 6388.544921875,
-      "eval_runtime": 0.5344,
-      "eval_samples_per_second": 14.971,
-      "eval_steps_per_second": 1.871,
+      "eval_loss": 6396.6748046875,
+      "eval_runtime": 0.5303,
+      "eval_samples_per_second": 15.085,
+      "eval_steps_per_second": 1.886,
       "step": 36
     },
     {
       "epoch": 10.0,
-      "learning_rate": 0.0001,
-      "loss": 6083.1363,
+      "learning_rate": 5e-06,
+      "loss": 6090.4715,
       "step": 40
     },
     {
       "epoch": 10.0,
-      "eval_loss": 6388.326171875,
-      "eval_runtime": 0.5236,
-      "eval_samples_per_second": 15.28,
-      "eval_steps_per_second": 1.91,
+      "eval_loss": 6396.595703125,
+      "eval_runtime": 0.5278,
+      "eval_samples_per_second": 15.158,
+      "eval_steps_per_second": 1.895,
       "step": 40
     },
     {
       "epoch": 11.0,
-      "eval_loss": 6390.20556640625,
-      "eval_runtime": 0.5417,
-      "eval_samples_per_second": 14.767,
-      "eval_steps_per_second": 1.846,
+      "eval_loss": 6396.53466796875,
+      "eval_runtime": 0.5332,
+      "eval_samples_per_second": 15.004,
+      "eval_steps_per_second": 1.876,
       "step": 44
     },
     {
       "epoch": 12.0,
-      "eval_loss": 6388.66064453125,
-      "eval_runtime": 0.5437,
-      "eval_samples_per_second": 14.713,
-      "eval_steps_per_second": 1.839,
+      "eval_loss": 6396.41064453125,
+      "eval_runtime": 0.5371,
+      "eval_samples_per_second": 14.895,
+      "eval_steps_per_second": 1.862,
       "step": 48
     },
     {
       "epoch": 12.5,
-      "learning_rate": 7.500000000000001e-05,
-      "loss": 6260.7426,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 6268.359,
       "step": 50
     },
     {
       "epoch": 13.0,
-      "eval_loss": 6388.1484375,
-      "eval_runtime": 0.5303,
-      "eval_samples_per_second": 15.086,
-      "eval_steps_per_second": 1.886,
+      "eval_loss": 6396.3232421875,
+      "eval_runtime": 0.521,
+      "eval_samples_per_second": 15.354,
+      "eval_steps_per_second": 1.919,
       "step": 52
     },
     {
       "epoch": 14.0,
-      "eval_loss": 6388.0478515625,
-      "eval_runtime": 0.5365,
-      "eval_samples_per_second": 14.913,
-      "eval_steps_per_second": 1.864,
+      "eval_loss": 6396.234375,
+      "eval_runtime": 0.5384,
+      "eval_samples_per_second": 14.858,
+      "eval_steps_per_second": 1.857,
       "step": 56
     },
     {
       "epoch": 15.0,
-      "learning_rate": 5e-05,
-      "loss": 6377.543,
+      "learning_rate": 2.5e-06,
+      "loss": 6384.7945,
       "step": 60
     },
     {
       "epoch": 15.0,
-      "eval_loss": 6388.0302734375,
-      "eval_runtime": 0.5355,
-      "eval_samples_per_second": 14.94,
-      "eval_steps_per_second": 1.867,
+      "eval_loss": 6396.095703125,
+      "eval_runtime": 0.5186,
+      "eval_samples_per_second": 15.426,
+      "eval_steps_per_second": 1.928,
       "step": 60
     },
     {
       "epoch": 16.0,
-      "eval_loss": 6388.1337890625,
-      "eval_runtime": 0.5423,
-      "eval_samples_per_second": 14.751,
-      "eval_steps_per_second": 1.844,
+      "eval_loss": 6396.09033203125,
+      "eval_runtime": 0.5163,
+      "eval_samples_per_second": 15.496,
+      "eval_steps_per_second": 1.937,
       "step": 64
     },
     {
       "epoch": 17.0,
-      "eval_loss": 6388.0595703125,
-      "eval_runtime": 0.5448,
-      "eval_samples_per_second": 14.684,
-      "eval_steps_per_second": 1.835,
+      "eval_loss": 6396.0029296875,
+      "eval_runtime": 0.5253,
+      "eval_samples_per_second": 15.228,
+      "eval_steps_per_second": 1.904,
       "step": 68
     },
     {
       "epoch": 17.5,
-      "learning_rate": 2.5e-05,
-      "loss": 6318.3168,
+      "learning_rate": 1.25e-06,
+      "loss": 6325.7992,
       "step": 70
     },
     {
       "epoch": 18.0,
-      "eval_loss": 6388.01953125,
-      "eval_runtime": 0.5448,
-      "eval_samples_per_second": 14.685,
-      "eval_steps_per_second": 1.836,
+      "eval_loss": 6395.94580078125,
+      "eval_runtime": 0.5319,
+      "eval_samples_per_second": 15.042,
+      "eval_steps_per_second": 1.88,
       "step": 72
     },
     {
       "epoch": 19.0,
-      "eval_loss": 6387.9638671875,
-      "eval_runtime": 0.5362,
-      "eval_samples_per_second": 14.918,
-      "eval_steps_per_second": 1.865,
+      "eval_loss": 6395.97802734375,
+      "eval_runtime": 0.5298,
+      "eval_samples_per_second": 15.099,
+      "eval_steps_per_second": 1.887,
       "step": 76
     },
     {
       "epoch": 20.0,
       "learning_rate": 0.0,
-      "loss": 6258.2133,
+      "loss": 6265.4016,
       "step": 80
     },
     {
       "epoch": 20.0,
-      "eval_loss": 6387.95458984375,
-      "eval_runtime": 0.544,
-      "eval_samples_per_second": 14.705,
-      "eval_steps_per_second": 1.838,
+      "eval_loss": 6395.9765625,
+      "eval_runtime": 0.5361,
+      "eval_samples_per_second": 14.924,
+      "eval_steps_per_second": 1.865,
       "step": 80
     },
     {
       "epoch": 20.0,
       "step": 80,
       "total_flos": 3.060008460288e+17,
-      "train_loss": 6247.897314453125,
-      "train_runtime": 365.3078,
-      "train_samples_per_second": 1.752,
-      "train_steps_per_second": 0.219
+      "train_loss": 6254.356591796875,
+      "train_runtime": 330.5487,
+      "train_samples_per_second": 1.936,
+      "train_steps_per_second": 0.242
     }
   ],
   "max_steps": 80,