End of training

all_results.json

@@ -1,11 +1,11 @@
 {
-    "epoch": 20.0,
-    "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
+    "epoch": 50.0,
+    "eval_loss": 6388.02587890625,
+    "eval_runtime": 0.6211,
+    "eval_samples_per_second": 12.88,
+    "eval_steps_per_second": 1.61,
+    "train_loss": 6294.72984375,
+    "train_runtime": 782.6675,
+    "train_samples_per_second": 2.044,
+    "train_steps_per_second": 0.256
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 20.0,
-    "eval_loss": 6395.94580078125,
-    "eval_runtime": 0.6002,
-    "eval_samples_per_second": 13.329,
-    "eval_steps_per_second": 1.666
+    "epoch": 50.0,
+    "eval_loss": 6388.02587890625,
+    "eval_runtime": 0.6211,
+    "eval_samples_per_second": 12.88,
+    "eval_steps_per_second": 1.61
 }

runs/Jul19_14-13-11_amy-2-gpu/events.out.tfevents.1689776807.amy-2-gpu.101345.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a0692b94998ca953e1bcb62053c26f59c67dfa4b008cd0d9b12099ffaa4eb09
+size 359

train_results.json

@@ -1,7 +1,7 @@
 {
-    "epoch": 20.0,
-    "train_loss": 6254.356591796875,
-    "train_runtime": 330.5487,
-    "train_samples_per_second": 1.936,
-    "train_steps_per_second": 0.242
+    "epoch": 50.0,
+    "train_loss": 6294.72984375,
+    "train_runtime": 782.6675,
+    "train_samples_per_second": 2.044,
+    "train_steps_per_second": 0.256
 }

trainer_state.json

@@ -1,233 +1,545 @@
 {
-  "best_metric": 6395.94580078125,
-  "best_model_checkpoint": "./coco_outputs/checkpoint-72",
-  "epoch": 20.0,
-  "global_step": 80,
+  "best_metric": 6388.02587890625,
+  "best_model_checkpoint": "./coco_outputs/checkpoint-148",
+  "epoch": 50.0,
+  "global_step": 200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_loss": 6398.77734375,
-      "eval_runtime": 0.5241,
-      "eval_samples_per_second": 15.264,
-      "eval_steps_per_second": 1.908,
+      "eval_loss": 6397.626953125,
+      "eval_runtime": 0.5268,
+      "eval_samples_per_second": 15.187,
+      "eval_steps_per_second": 1.898,
       "step": 4
     },
     {
       "epoch": 2.0,
-      "eval_loss": 6398.435546875,
-      "eval_runtime": 0.5241,
-      "eval_samples_per_second": 15.265,
-      "eval_steps_per_second": 1.908,
+      "eval_loss": 6396.95263671875,
+      "eval_runtime": 0.5209,
+      "eval_samples_per_second": 15.358,
+      "eval_steps_per_second": 1.92,
       "step": 8
     },
     {
       "epoch": 2.5,
-      "learning_rate": 8.750000000000001e-06,
-      "loss": 6176.0926,
+      "learning_rate": 4.75e-05,
+      "loss": 6175.4402,
       "step": 10
     },
     {
       "epoch": 3.0,
-      "eval_loss": 6398.2265625,
-      "eval_runtime": 0.5251,
-      "eval_samples_per_second": 15.235,
-      "eval_steps_per_second": 1.904,
+      "eval_loss": 6396.91357421875,
+      "eval_runtime": 0.5052,
+      "eval_samples_per_second": 15.836,
+      "eval_steps_per_second": 1.98,
       "step": 12
     },
     {
       "epoch": 4.0,
-      "eval_loss": 6397.8564453125,
-      "eval_runtime": 0.5329,
-      "eval_samples_per_second": 15.013,
-      "eval_steps_per_second": 1.877,
+      "eval_loss": 6396.29931640625,
+      "eval_runtime": 0.527,
+      "eval_samples_per_second": 15.181,
+      "eval_steps_per_second": 1.898,
       "step": 16
     },
     {
       "epoch": 5.0,
-      "learning_rate": 7.500000000000001e-06,
-      "loss": 6292.8656,
+      "learning_rate": 4.5e-05,
+      "loss": 6291.4262,
       "step": 20
     },
     {
       "epoch": 5.0,
-      "eval_loss": 6397.6298828125,
-      "eval_runtime": 0.5291,
-      "eval_samples_per_second": 15.121,
-      "eval_steps_per_second": 1.89,
+      "eval_loss": 6395.5556640625,
+      "eval_runtime": 0.5357,
+      "eval_samples_per_second": 14.934,
+      "eval_steps_per_second": 1.867,
       "step": 20
     },
     {
       "epoch": 6.0,
-      "eval_loss": 6397.31787109375,
-      "eval_runtime": 0.5319,
-      "eval_samples_per_second": 15.042,
-      "eval_steps_per_second": 1.88,
+      "eval_loss": 6394.595703125,
+      "eval_runtime": 0.5351,
+      "eval_samples_per_second": 14.951,
+      "eval_steps_per_second": 1.869,
       "step": 24
     },
     {
       "epoch": 7.0,
-      "eval_loss": 6396.95263671875,
-      "eval_runtime": 0.5129,
-      "eval_samples_per_second": 15.596,
-      "eval_steps_per_second": 1.95,
+      "eval_loss": 6393.72900390625,
+      "eval_runtime": 0.5361,
+      "eval_samples_per_second": 14.922,
+      "eval_steps_per_second": 1.865,
       "step": 28
     },
     {
       "epoch": 7.5,
-      "learning_rate": 6.25e-06,
-      "loss": 6231.0688,
+      "learning_rate": 4.25e-05,
+      "loss": 6228.9867,
       "step": 30
     },
     {
       "epoch": 8.0,
-      "eval_loss": 6396.79052734375,
-      "eval_runtime": 0.5163,
-      "eval_samples_per_second": 15.494,
-      "eval_steps_per_second": 1.937,
+      "eval_loss": 6393.19140625,
+      "eval_runtime": 0.5377,
+      "eval_samples_per_second": 14.879,
+      "eval_steps_per_second": 1.86,
       "step": 32
     },
     {
       "epoch": 9.0,
-      "eval_loss": 6396.6748046875,
-      "eval_runtime": 0.5303,
-      "eval_samples_per_second": 15.085,
-      "eval_steps_per_second": 1.886,
+      "eval_loss": 6392.28369140625,
+      "eval_runtime": 0.6156,
+      "eval_samples_per_second": 12.995,
+      "eval_steps_per_second": 1.624,
       "step": 36
     },
     {
       "epoch": 10.0,
-      "learning_rate": 5e-06,
-      "loss": 6090.4715,
+      "learning_rate": 4e-05,
+      "loss": 6087.3184,
       "step": 40
     },
     {
       "epoch": 10.0,
-      "eval_loss": 6396.595703125,
-      "eval_runtime": 0.5278,
-      "eval_samples_per_second": 15.158,
-      "eval_steps_per_second": 1.895,
+      "eval_loss": 6391.6708984375,
+      "eval_runtime": 0.5319,
+      "eval_samples_per_second": 15.04,
+      "eval_steps_per_second": 1.88,
       "step": 40
     },
     {
       "epoch": 11.0,
-      "eval_loss": 6396.53466796875,
-      "eval_runtime": 0.5332,
-      "eval_samples_per_second": 15.004,
-      "eval_steps_per_second": 1.876,
+      "eval_loss": 6390.72705078125,
+      "eval_runtime": 0.5435,
+      "eval_samples_per_second": 14.719,
+      "eval_steps_per_second": 1.84,
       "step": 44
     },
     {
       "epoch": 12.0,
-      "eval_loss": 6396.41064453125,
-      "eval_runtime": 0.5371,
-      "eval_samples_per_second": 14.895,
-      "eval_steps_per_second": 1.862,
+      "eval_loss": 6389.6669921875,
+      "eval_runtime": 0.5411,
+      "eval_samples_per_second": 14.785,
+      "eval_steps_per_second": 1.848,
       "step": 48
     },
     {
       "epoch": 12.5,
-      "learning_rate": 3.7500000000000005e-06,
-      "loss": 6268.359,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 6263.3945,
       "step": 50
     },
     {
       "epoch": 13.0,
-      "eval_loss": 6396.3232421875,
-      "eval_runtime": 0.521,
-      "eval_samples_per_second": 15.354,
-      "eval_steps_per_second": 1.919,
+      "eval_loss": 6389.49072265625,
+      "eval_runtime": 0.5174,
+      "eval_samples_per_second": 15.462,
+      "eval_steps_per_second": 1.933,
       "step": 52
     },
     {
       "epoch": 14.0,
-      "eval_loss": 6396.234375,
-      "eval_runtime": 0.5384,
-      "eval_samples_per_second": 14.858,
-      "eval_steps_per_second": 1.857,
+      "eval_loss": 6388.85400390625,
+      "eval_runtime": 0.5481,
+      "eval_samples_per_second": 14.597,
+      "eval_steps_per_second": 1.825,
       "step": 56
     },
     {
       "epoch": 15.0,
-      "learning_rate": 2.5e-06,
-      "loss": 6384.7945,
+      "learning_rate": 3.5e-05,
+      "loss": 6378.8203,
       "step": 60
     },
     {
       "epoch": 15.0,
-      "eval_loss": 6396.095703125,
-      "eval_runtime": 0.5186,
-      "eval_samples_per_second": 15.426,
-      "eval_steps_per_second": 1.928,
+      "eval_loss": 6388.69775390625,
+      "eval_runtime": 0.5244,
+      "eval_samples_per_second": 15.255,
+      "eval_steps_per_second": 1.907,
       "step": 60
     },
     {
       "epoch": 16.0,
-      "eval_loss": 6396.09033203125,
-      "eval_runtime": 0.5163,
-      "eval_samples_per_second": 15.496,
-      "eval_steps_per_second": 1.937,
+      "eval_loss": 6388.6611328125,
+      "eval_runtime": 0.5266,
+      "eval_samples_per_second": 15.193,
+      "eval_steps_per_second": 1.899,
       "step": 64
     },
     {
       "epoch": 17.0,
-      "eval_loss": 6396.0029296875,
-      "eval_runtime": 0.5253,
-      "eval_samples_per_second": 15.228,
-      "eval_steps_per_second": 1.904,
+      "eval_loss": 6388.54296875,
+      "eval_runtime": 0.5529,
+      "eval_samples_per_second": 14.47,
+      "eval_steps_per_second": 1.809,
       "step": 68
     },
     {
       "epoch": 17.5,
-      "learning_rate": 1.25e-06,
-      "loss": 6325.7992,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 6319.1008,
       "step": 70
     },
     {
       "epoch": 18.0,
-      "eval_loss": 6395.94580078125,
-      "eval_runtime": 0.5319,
-      "eval_samples_per_second": 15.042,
-      "eval_steps_per_second": 1.88,
+      "eval_loss": 6388.421875,
+      "eval_runtime": 0.5476,
+      "eval_samples_per_second": 14.609,
+      "eval_steps_per_second": 1.826,
       "step": 72
     },
     {
       "epoch": 19.0,
-      "eval_loss": 6395.97802734375,
-      "eval_runtime": 0.5298,
-      "eval_samples_per_second": 15.099,
-      "eval_steps_per_second": 1.887,
+      "eval_loss": 6388.505859375,
+      "eval_runtime": 0.5589,
+      "eval_samples_per_second": 14.314,
+      "eval_steps_per_second": 1.789,
       "step": 76
     },
     {
       "epoch": 20.0,
-      "learning_rate": 0.0,
-      "loss": 6265.4016,
+      "learning_rate": 3e-05,
+      "loss": 6258.6707,
       "step": 80
     },
     {
       "epoch": 20.0,
-      "eval_loss": 6395.9765625,
-      "eval_runtime": 0.5361,
-      "eval_samples_per_second": 14.924,
-      "eval_steps_per_second": 1.865,
+      "eval_loss": 6388.39501953125,
+      "eval_runtime": 0.5558,
+      "eval_samples_per_second": 14.393,
+      "eval_steps_per_second": 1.799,
       "step": 80
     },
     {
-      "epoch": 20.0,
-      "step": 80,
-      "total_flos": 3.060008460288e+17,
-      "train_loss": 6254.356591796875,
-      "train_runtime": 330.5487,
-      "train_samples_per_second": 1.936,
-      "train_steps_per_second": 0.242
+      "epoch": 21.0,
+      "eval_loss": 6388.30419921875,
+      "eval_runtime": 0.6087,
+      "eval_samples_per_second": 13.142,
+      "eval_steps_per_second": 1.643,
+      "step": 84
+    },
+    {
+      "epoch": 22.0,
+      "eval_loss": 6388.27880859375,
+      "eval_runtime": 0.5565,
+      "eval_samples_per_second": 14.376,
+      "eval_steps_per_second": 1.797,
+      "step": 88
+    },
+    {
+      "epoch": 22.5,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 6290.5227,
+      "step": 90
+    },
+    {
+      "epoch": 23.0,
+      "eval_loss": 6388.27001953125,
+      "eval_runtime": 0.5542,
+      "eval_samples_per_second": 14.435,
+      "eval_steps_per_second": 1.804,
+      "step": 92
+    },
+    {
+      "epoch": 24.0,
+      "eval_loss": 6388.3994140625,
+      "eval_runtime": 0.5489,
+      "eval_samples_per_second": 14.573,
+      "eval_steps_per_second": 1.822,
+      "step": 96
+    },
+    {
+      "epoch": 25.0,
+      "learning_rate": 2.5e-05,
+      "loss": 6450.2746,
+      "step": 100
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 6388.41796875,
+      "eval_runtime": 0.5566,
+      "eval_samples_per_second": 14.372,
+      "eval_steps_per_second": 1.797,
+      "step": 100
+    },
+    {
+      "epoch": 26.0,
+      "eval_loss": 6388.23291015625,
+      "eval_runtime": 0.5471,
+      "eval_samples_per_second": 14.622,
+      "eval_steps_per_second": 1.828,
+      "step": 104
+    },
+    {
+      "epoch": 27.0,
+      "eval_loss": 6388.30615234375,
+      "eval_runtime": 0.5404,
+      "eval_samples_per_second": 14.803,
+      "eval_steps_per_second": 1.85,
+      "step": 108
+    },
+    {
+      "epoch": 27.5,
+      "learning_rate": 2.25e-05,
+      "loss": 6321.6988,
+      "step": 110
+    },
+    {
+      "epoch": 28.0,
+      "eval_loss": 6388.2568359375,
+      "eval_runtime": 0.5645,
+      "eval_samples_per_second": 14.172,
+      "eval_steps_per_second": 1.772,
+      "step": 112
+    },
+    {
+      "epoch": 29.0,
+      "eval_loss": 6388.13330078125,
+      "eval_runtime": 0.5582,
+      "eval_samples_per_second": 14.333,
+      "eval_steps_per_second": 1.792,
+      "step": 116
+    },
+    {
+      "epoch": 30.0,
+      "learning_rate": 2e-05,
+      "loss": 6468.3922,
+      "step": 120
+    },
+    {
+      "epoch": 30.0,
+      "eval_loss": 6388.166015625,
+      "eval_runtime": 0.5512,
+      "eval_samples_per_second": 14.515,
+      "eval_steps_per_second": 1.814,
+      "step": 120
+    },
+    {
+      "epoch": 31.0,
+      "eval_loss": 6388.15283203125,
+      "eval_runtime": 0.5777,
+      "eval_samples_per_second": 13.848,
+      "eval_steps_per_second": 1.731,
+      "step": 124
+    },
+    {
+      "epoch": 32.0,
+      "eval_loss": 6388.12109375,
+      "eval_runtime": 0.5552,
+      "eval_samples_per_second": 14.408,
+      "eval_steps_per_second": 1.801,
+      "step": 128
+    },
+    {
+      "epoch": 32.5,
+      "learning_rate": 1.75e-05,
+      "loss": 6305.5602,
+      "step": 130
+    },
+    {
+      "epoch": 33.0,
+      "eval_loss": 6388.09716796875,
+      "eval_runtime": 0.5442,
+      "eval_samples_per_second": 14.7,
+      "eval_steps_per_second": 1.838,
+      "step": 132
+    },
+    {
+      "epoch": 34.0,
+      "eval_loss": 6388.07763671875,
+      "eval_runtime": 0.5513,
+      "eval_samples_per_second": 14.511,
+      "eval_steps_per_second": 1.814,
+      "step": 136
+    },
+    {
+      "epoch": 35.0,
+      "learning_rate": 1.5e-05,
+      "loss": 6349.5672,
+      "step": 140
+    },
+    {
+      "epoch": 35.0,
+      "eval_loss": 6388.04931640625,
+      "eval_runtime": 0.5426,
+      "eval_samples_per_second": 14.744,
+      "eval_steps_per_second": 1.843,
+      "step": 140
+    },
+    {
+      "epoch": 36.0,
+      "eval_loss": 6388.03173828125,
+      "eval_runtime": 0.5452,
+      "eval_samples_per_second": 14.674,
+      "eval_steps_per_second": 1.834,
+      "step": 144
+    },
+    {
+      "epoch": 37.0,
+      "eval_loss": 6388.02587890625,
+      "eval_runtime": 0.5716,
+      "eval_samples_per_second": 13.996,
+      "eval_steps_per_second": 1.75,
+      "step": 148
+    },
+    {
+      "epoch": 37.5,
+      "learning_rate": 1.25e-05,
+      "loss": 6244.9891,
+      "step": 150
+    },
+    {
+      "epoch": 38.0,
+      "eval_loss": 6388.037109375,
+      "eval_runtime": 0.5397,
+      "eval_samples_per_second": 14.822,
+      "eval_steps_per_second": 1.853,
+      "step": 152
+    },
+    {
+      "epoch": 39.0,
+      "eval_loss": 6388.02734375,
+      "eval_runtime": 0.5505,
+      "eval_samples_per_second": 14.532,
+      "eval_steps_per_second": 1.817,
+      "step": 156
+    },
+    {
+      "epoch": 40.0,
+      "learning_rate": 1e-05,
+      "loss": 6383.7191,
+      "step": 160
+    },
+    {
+      "epoch": 40.0,
+      "eval_loss": 6388.146484375,
+      "eval_runtime": 0.537,
+      "eval_samples_per_second": 14.898,
+      "eval_steps_per_second": 1.862,
+      "step": 160
+    },
+    {
+      "epoch": 41.0,
+      "eval_loss": 6388.25048828125,
+      "eval_runtime": 0.5406,
+      "eval_samples_per_second": 14.797,
+      "eval_steps_per_second": 1.85,
+      "step": 164
+    },
+    {
+      "epoch": 42.0,
+      "eval_loss": 6388.1259765625,
+      "eval_runtime": 0.5956,
+      "eval_samples_per_second": 13.433,
+      "eval_steps_per_second": 1.679,
+      "step": 168
+    },
+    {
+      "epoch": 42.5,
+      "learning_rate": 7.5e-06,
+      "loss": 6260.3203,
+      "step": 170
+    },
+    {
+      "epoch": 43.0,
+      "eval_loss": 6388.0517578125,
+      "eval_runtime": 0.557,
+      "eval_samples_per_second": 14.363,
+      "eval_steps_per_second": 1.795,
+      "step": 172
+    },
+    {
+      "epoch": 44.0,
+      "eval_loss": 6388.0322265625,
+      "eval_runtime": 0.5537,
+      "eval_samples_per_second": 14.447,
+      "eval_steps_per_second": 1.806,
+      "step": 176
+    },
+    {
+      "epoch": 45.0,
+      "learning_rate": 5e-06,
+      "loss": 6254.6055,
+      "step": 180
+    },
+    {
+      "epoch": 45.0,
+      "eval_loss": 6388.0625,
+      "eval_runtime": 0.5885,
+      "eval_samples_per_second": 13.595,
+      "eval_steps_per_second": 1.699,
+      "step": 180
+    },
+    {
+      "epoch": 46.0,
+      "eval_loss": 6388.06884765625,
+      "eval_runtime": 0.5716,
+      "eval_samples_per_second": 13.995,
+      "eval_steps_per_second": 1.749,
+      "step": 184
+    },
+    {
+      "epoch": 47.0,
+      "eval_loss": 6388.06005859375,
+      "eval_runtime": 0.5704,
+      "eval_samples_per_second": 14.026,
+      "eval_steps_per_second": 1.753,
+      "step": 188
+    },
+    {
+      "epoch": 47.5,
+      "learning_rate": 2.5e-06,
+      "loss": 6351.991,
+      "step": 190
+    },
+    {
+      "epoch": 48.0,
+      "eval_loss": 6388.08251953125,
+      "eval_runtime": 0.5514,
+      "eval_samples_per_second": 14.507,
+      "eval_steps_per_second": 1.813,
+      "step": 192
+    },
+    {
+      "epoch": 49.0,
+      "eval_loss": 6388.09375,
+      "eval_runtime": 0.5515,
+      "eval_samples_per_second": 14.506,
+      "eval_steps_per_second": 1.813,
+      "step": 196
+    },
+    {
+      "epoch": 50.0,
+      "learning_rate": 0.0,
+      "loss": 6209.7984,
+      "step": 200
+    },
+    {
+      "epoch": 50.0,
+      "eval_loss": 6388.09228515625,
+      "eval_runtime": 0.566,
+      "eval_samples_per_second": 14.135,
+      "eval_steps_per_second": 1.767,
+      "step": 200
+    },
+    {
+      "epoch": 50.0,
+      "step": 200,
+      "total_flos": 7.65002115072e+17,
+      "train_loss": 6294.72984375,
+      "train_runtime": 782.6675,
+      "train_samples_per_second": 2.044,
+      "train_steps_per_second": 0.256
     }
   ],
-  "max_steps": 80,
-  "num_train_epochs": 20,
-  "total_flos": 3.060008460288e+17,
+  "max_steps": 200,
+  "num_train_epochs": 50,
+  "total_flos": 7.65002115072e+17,
   "trial_name": null,
   "trial_params": null
 }