Upload 8 files

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:40e486e534b16c41b45d08413ceb5f443c0991730446ff950bd02f1ea93d3a71
+oid sha256:c799f21594ab06d6b91129c311773ce342760d0c02360ffb01ea6208e4faace8
 size 33661637

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd771d4fdcc07b3c7d3128e34406ab567381117cf663ec63f46bc1ee9385a49b
+oid sha256:f329b74b9ae80cb5e6b64ce551d508321a3549bc3d469325a37753fa7dd59f60
 size 14575

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fdf9887d830501ed44fcfc9b1240b3894b0a39f156a92a6534273c610fd7f49b
+oid sha256:57ad5ba60863d9ee4b493aebecbb69b4c26f9bad3b815096052e17626ba0bb35
 size 627

trainer_state.json

@@ -1,1248 +1,696 @@
 {
-  "best_metric": 0.43608614802360535,
-  "best_model_checkpoint": "experiments/checkpoint-1000",
-  "epoch": 16.0,
-  "global_step": 1000,
+  "best_metric": 1.9192386865615845,
+  "best_model_checkpoint": "experiments/checkpoint-500",
+  "epoch": 100.0,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.08,
-      "learning_rate": 2.9999999999999997e-06,
-      "loss": 1.5328,
+      "epoch": 1.0,
+      "learning_rate": 1.8e-06,
+      "loss": 1.7688,
       "step": 5
     },
     {
-      "epoch": 0.16,
-      "learning_rate": 5.999999999999999e-06,
-      "loss": 1.5296,
+      "epoch": 2.0,
+      "learning_rate": 4.8e-06,
+      "loss": 1.7684,
       "step": 10
     },
     {
-      "epoch": 0.24,
-      "learning_rate": 8.999999999999999e-06,
-      "loss": 1.5193,
+      "epoch": 3.0,
+      "learning_rate": 7.799999999999998e-06,
+      "loss": 1.7564,
       "step": 15
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 1.1999999999999999e-05,
-      "loss": 1.5027,
+      "epoch": 4.0,
+      "learning_rate": 1.0799999999999998e-05,
+      "loss": 1.7424,
       "step": 20
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 1.4999999999999999e-05,
-      "loss": 1.5015,
+      "epoch": 5.0,
+      "learning_rate": 1.3799999999999998e-05,
+      "loss": 1.727,
       "step": 25
     },
     {
-      "epoch": 0.48,
-      "learning_rate": 1.7999999999999997e-05,
-      "loss": 1.4799,
+      "epoch": 6.0,
+      "learning_rate": 1.68e-05,
+      "loss": 1.7135,
       "step": 30
     },
     {
-      "epoch": 0.56,
-      "learning_rate": 2.1e-05,
-      "loss": 1.4705,
+      "epoch": 7.0,
+      "learning_rate": 1.98e-05,
+      "loss": 1.701,
       "step": 35
     },
     {
-      "epoch": 0.64,
-      "learning_rate": 2.3999999999999997e-05,
-      "loss": 1.4524,
+      "epoch": 8.0,
+      "learning_rate": 2.28e-05,
+      "loss": 1.6797,
       "step": 40
     },
     {
-      "epoch": 0.72,
-      "learning_rate": 2.6999999999999996e-05,
-      "loss": 1.4231,
+      "epoch": 9.0,
+      "learning_rate": 2.5799999999999997e-05,
+      "loss": 1.6547,
       "step": 45
     },
     {
-      "epoch": 0.8,
-      "learning_rate": 2.9999999999999997e-05,
-      "loss": 1.3934,
+      "epoch": 10.0,
+      "learning_rate": 2.88e-05,
+      "loss": 1.6245,
+      "step": 50
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 1.5684312582015991,
+      "eval_runtime": 1.4166,
+      "eval_samples_per_second": 7.059,
+      "eval_steps_per_second": 1.412,
       "step": 50
     },
     {
-      "epoch": 0.88,
-      "learning_rate": 3.2999999999999996e-05,
-      "loss": 1.3408,
+      "epoch": 11.0,
+      "learning_rate": 3.1799999999999994e-05,
+      "loss": 1.5841,
       "step": 55
     },
     {
-      "epoch": 0.96,
-      "learning_rate": 3.5999999999999994e-05,
-      "loss": 1.2911,
+      "epoch": 12.0,
+      "learning_rate": 3.48e-05,
+      "loss": 1.5316,
       "step": 60
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 3.9e-05,
-      "loss": 1.2097,
+      "epoch": 13.0,
+      "learning_rate": 3.78e-05,
+      "loss": 1.4644,
       "step": 65
     },
     {
-      "epoch": 1.12,
-      "learning_rate": 4.2e-05,
-      "loss": 1.121,
+      "epoch": 14.0,
+      "learning_rate": 4.08e-05,
+      "loss": 1.3728,
       "step": 70
     },
     {
-      "epoch": 1.2,
-      "learning_rate": 4.4999999999999996e-05,
-      "loss": 1.0317,
+      "epoch": 15.0,
+      "learning_rate": 4.3799999999999994e-05,
+      "loss": 1.2692,
       "step": 75
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.7999999999999994e-05,
-      "loss": 0.9421,
+      "epoch": 16.0,
+      "learning_rate": 4.56e-05,
+      "loss": 1.1998,
       "step": 80
     },
     {
-      "epoch": 1.36,
-      "learning_rate": 5.1e-05,
-      "loss": 0.883,
+      "epoch": 17.0,
+      "learning_rate": 4.8599999999999995e-05,
+      "loss": 1.1159,
       "step": 85
     },
     {
-      "epoch": 1.44,
-      "learning_rate": 5.399999999999999e-05,
-      "loss": 0.8388,
+      "epoch": 18.0,
+      "learning_rate": 5.1599999999999994e-05,
+      "loss": 1.0442,
       "step": 90
     },
     {
-      "epoch": 1.52,
-      "learning_rate": 5.6999999999999996e-05,
-      "loss": 0.8171,
+      "epoch": 19.0,
+      "learning_rate": 5.459999999999999e-05,
+      "loss": 0.9944,
       "step": 95
     },
     {
-      "epoch": 1.6,
-      "learning_rate": 5.9999999999999995e-05,
-      "loss": 0.7882,
+      "epoch": 20.0,
+      "learning_rate": 5.76e-05,
+      "loss": 0.9518,
       "step": 100
     },
     {
-      "epoch": 1.68,
-      "learning_rate": 6.299999999999999e-05,
-      "loss": 0.7571,
+      "epoch": 20.0,
+      "eval_loss": 0.9818406105041504,
+      "eval_runtime": 1.3835,
+      "eval_samples_per_second": 7.228,
+      "eval_steps_per_second": 1.446,
+      "step": 100
+    },
+    {
+      "epoch": 21.0,
+      "learning_rate": 6.0599999999999996e-05,
+      "loss": 0.908,
       "step": 105
     },
     {
-      "epoch": 1.76,
-      "learning_rate": 6.599999999999999e-05,
-      "loss": 0.7267,
+      "epoch": 22.0,
+      "learning_rate": 6.359999999999999e-05,
+      "loss": 0.8678,
       "step": 110
     },
     {
-      "epoch": 1.84,
-      "learning_rate": 6.9e-05,
-      "loss": 0.7099,
+      "epoch": 23.0,
+      "learning_rate": 6.659999999999999e-05,
+      "loss": 0.8303,
       "step": 115
     },
     {
-      "epoch": 1.92,
-      "learning_rate": 7.199999999999999e-05,
-      "loss": 0.69,
+      "epoch": 24.0,
+      "learning_rate": 6.96e-05,
+      "loss": 0.7928,
       "step": 120
     },
     {
-      "epoch": 2.0,
-      "learning_rate": 7.5e-05,
-      "loss": 0.6809,
+      "epoch": 25.0,
+      "learning_rate": 7.259999999999999e-05,
+      "loss": 0.7594,
       "step": 125
     },
     {
-      "epoch": 2.08,
-      "learning_rate": 7.8e-05,
-      "loss": 0.6636,
+      "epoch": 26.0,
+      "learning_rate": 7.56e-05,
+      "loss": 0.73,
       "step": 130
     },
     {
-      "epoch": 2.16,
-      "learning_rate": 8.1e-05,
-      "loss": 0.6522,
+      "epoch": 27.0,
+      "learning_rate": 7.86e-05,
+      "loss": 0.7034,
       "step": 135
     },
     {
-      "epoch": 2.24,
-      "learning_rate": 8.4e-05,
-      "loss": 0.6398,
+      "epoch": 28.0,
+      "learning_rate": 8.16e-05,
+      "loss": 0.6777,
       "step": 140
     },
     {
-      "epoch": 2.32,
-      "learning_rate": 8.699999999999999e-05,
-      "loss": 0.6322,
+      "epoch": 29.0,
+      "learning_rate": 8.459999999999998e-05,
+      "loss": 0.6493,
       "step": 145
     },
     {
-      "epoch": 2.4,
-      "learning_rate": 8.999999999999999e-05,
-      "loss": 0.6268,
+      "epoch": 30.0,
+      "learning_rate": 8.759999999999999e-05,
+      "loss": 0.6249,
       "step": 150
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 9.3e-05,
-      "loss": 0.6159,
+      "epoch": 30.0,
+      "eval_loss": 0.8599645495414734,
+      "eval_runtime": 1.4099,
+      "eval_samples_per_second": 7.093,
+      "eval_steps_per_second": 1.419,
+      "step": 150
+    },
+    {
+      "epoch": 31.0,
+      "learning_rate": 9.059999999999999e-05,
+      "loss": 0.6007,
       "step": 155
     },
     {
-      "epoch": 2.56,
-      "learning_rate": 9.599999999999999e-05,
-      "loss": 0.6111,
+      "epoch": 32.0,
+      "learning_rate": 9.36e-05,
+      "loss": 0.5716,
       "step": 160
     },
     {
-      "epoch": 2.64,
-      "learning_rate": 9.9e-05,
-      "loss": 0.5995,
+      "epoch": 33.0,
+      "learning_rate": 9.659999999999999e-05,
+      "loss": 0.5465,
       "step": 165
     },
     {
-      "epoch": 2.72,
-      "learning_rate": 0.000102,
-      "loss": 0.591,
+      "epoch": 34.0,
+      "learning_rate": 9.96e-05,
+      "loss": 0.5191,
       "step": 170
     },
     {
-      "epoch": 2.8,
-      "learning_rate": 0.00010499999999999999,
-      "loss": 0.5885,
+      "epoch": 35.0,
+      "learning_rate": 0.0001026,
+      "loss": 0.4947,
       "step": 175
     },
     {
-      "epoch": 2.88,
-      "learning_rate": 0.00010799999999999998,
-      "loss": 0.5779,
+      "epoch": 36.0,
+      "learning_rate": 0.00010559999999999998,
+      "loss": 0.4681,
       "step": 180
     },
     {
-      "epoch": 2.96,
-      "learning_rate": 0.00011099999999999999,
-      "loss": 0.5663,
+      "epoch": 37.0,
+      "learning_rate": 0.00010859999999999998,
+      "loss": 0.4417,
       "step": 185
     },
     {
-      "epoch": 3.04,
-      "learning_rate": 0.00011399999999999999,
-      "loss": 0.5686,
+      "epoch": 38.0,
+      "learning_rate": 0.00011159999999999999,
+      "loss": 0.4116,
       "step": 190
     },
     {
-      "epoch": 3.12,
-      "learning_rate": 0.000117,
-      "loss": 0.558,
+      "epoch": 39.0,
+      "learning_rate": 0.0001146,
+      "loss": 0.3804,
       "step": 195
     },
     {
-      "epoch": 3.2,
-      "learning_rate": 0.00011999999999999999,
-      "loss": 0.5568,
+      "epoch": 40.0,
+      "learning_rate": 0.0001176,
+      "loss": 0.3544,
+      "step": 200
+    },
+    {
+      "epoch": 40.0,
+      "eval_loss": 1.1081293821334839,
+      "eval_runtime": 1.3764,
+      "eval_samples_per_second": 7.265,
+      "eval_steps_per_second": 1.453,
       "step": 200
     },
     {
-      "epoch": 3.28,
-      "learning_rate": 0.00012299999999999998,
-      "loss": 0.5528,
+      "epoch": 41.0,
+      "learning_rate": 0.00012059999999999999,
+      "loss": 0.3248,
       "step": 205
     },
     {
-      "epoch": 3.36,
-      "learning_rate": 0.00012599999999999997,
-      "loss": 0.5469,
+      "epoch": 42.0,
+      "learning_rate": 0.0001236,
+      "loss": 0.2931,
       "step": 210
     },
     {
-      "epoch": 3.44,
-      "learning_rate": 0.000129,
-      "loss": 0.545,
+      "epoch": 43.0,
+      "learning_rate": 0.0001266,
+      "loss": 0.2677,
       "step": 215
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 0.00013199999999999998,
-      "loss": 0.5373,
+      "epoch": 44.0,
+      "learning_rate": 0.00012959999999999998,
+      "loss": 0.2386,
       "step": 220
     },
     {
-      "epoch": 3.6,
-      "learning_rate": 0.000135,
-      "loss": 0.5376,
+      "epoch": 45.0,
+      "learning_rate": 0.0001326,
+      "loss": 0.2142,
       "step": 225
     },
     {
-      "epoch": 3.68,
-      "learning_rate": 0.000138,
-      "loss": 0.5373,
+      "epoch": 46.0,
+      "learning_rate": 0.0001356,
+      "loss": 0.1932,
       "step": 230
     },
     {
-      "epoch": 3.76,
-      "learning_rate": 0.00014099999999999998,
-      "loss": 0.5241,
+      "epoch": 47.0,
+      "learning_rate": 0.0001386,
+      "loss": 0.1709,
       "step": 235
     },
     {
-      "epoch": 3.84,
-      "learning_rate": 0.00014399999999999998,
-      "loss": 0.5221,
+      "epoch": 48.0,
+      "learning_rate": 0.00014159999999999997,
+      "loss": 0.1571,
       "step": 240
     },
     {
-      "epoch": 3.92,
-      "learning_rate": 0.000147,
-      "loss": 0.5229,
+      "epoch": 49.0,
+      "learning_rate": 0.0001446,
+      "loss": 0.1417,
       "step": 245
     },
     {
-      "epoch": 4.0,
-      "learning_rate": 0.00015,
-      "loss": 0.5199,
+      "epoch": 50.0,
+      "learning_rate": 0.00014759999999999998,
+      "loss": 0.1184,
       "step": 250
     },
     {
-      "epoch": 4.0,
-      "eval_loss": 0.5244991183280945,
-      "eval_runtime": 265.103,
-      "eval_samples_per_second": 7.548,
-      "eval_steps_per_second": 0.947,
+      "epoch": 50.0,
+      "eval_loss": 1.5212451219558716,
+      "eval_runtime": 1.3827,
+      "eval_samples_per_second": 7.232,
+      "eval_steps_per_second": 1.446,
       "step": 250
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 0.00015299999999999998,
-      "loss": 0.5141,
+      "epoch": 51.0,
+      "learning_rate": 0.00015059999999999997,
+      "loss": 0.1096,
       "step": 255
     },
     {
-      "epoch": 4.16,
-      "learning_rate": 0.000156,
-      "loss": 0.5086,
+      "epoch": 52.0,
+      "learning_rate": 0.0001536,
+      "loss": 0.1037,
       "step": 260
     },
     {
-      "epoch": 4.24,
-      "learning_rate": 0.000159,
-      "loss": 0.5129,
+      "epoch": 53.0,
+      "learning_rate": 0.00015659999999999998,
+      "loss": 0.095,
       "step": 265
     },
     {
-      "epoch": 4.32,
-      "learning_rate": 0.000162,
-      "loss": 0.5156,
+      "epoch": 54.0,
+      "learning_rate": 0.0001596,
+      "loss": 0.0865,
       "step": 270
     },
     {
-      "epoch": 4.4,
-      "learning_rate": 0.000165,
-      "loss": 0.5053,
+      "epoch": 55.0,
+      "learning_rate": 0.0001626,
+      "loss": 0.0808,
       "step": 275
     },
     {
-      "epoch": 4.48,
-      "learning_rate": 0.000168,
-      "loss": 0.5024,
+      "epoch": 56.0,
+      "learning_rate": 0.0001656,
+      "loss": 0.0794,
       "step": 280
     },
     {
-      "epoch": 4.56,
-      "learning_rate": 0.00017099999999999998,
-      "loss": 0.5048,
+      "epoch": 57.0,
+      "learning_rate": 0.0001686,
+      "loss": 0.075,
       "step": 285
     },
     {
-      "epoch": 4.64,
-      "learning_rate": 0.00017399999999999997,
-      "loss": 0.5001,
+      "epoch": 58.0,
+      "learning_rate": 0.00017159999999999997,
+      "loss": 0.0726,
       "step": 290
     },
     {
-      "epoch": 4.72,
-      "learning_rate": 0.00017699999999999997,
-      "loss": 0.497,
+      "epoch": 59.0,
+      "learning_rate": 0.00017459999999999996,
+      "loss": 0.0696,
       "step": 295
     },
     {
-      "epoch": 4.8,
-      "learning_rate": 0.00017999999999999998,
-      "loss": 0.4929,
+      "epoch": 60.0,
+      "learning_rate": 0.00017759999999999998,
+      "loss": 0.0665,
       "step": 300
     },
     {
-      "epoch": 4.88,
-      "learning_rate": 0.00018299999999999998,
-      "loss": 0.4962,
+      "epoch": 60.0,
+      "eval_loss": 1.7048699855804443,
+      "eval_runtime": 1.3753,
+      "eval_samples_per_second": 7.271,
+      "eval_steps_per_second": 1.454,
+      "step": 300
+    },
+    {
+      "epoch": 61.0,
+      "learning_rate": 0.00018059999999999997,
+      "loss": 0.065,
       "step": 305
     },
     {
-      "epoch": 4.96,
-      "learning_rate": 0.000186,
-      "loss": 0.4941,
+      "epoch": 62.0,
+      "learning_rate": 0.0001836,
+      "loss": 0.0623,
       "step": 310
     },
     {
-      "epoch": 5.04,
-      "learning_rate": 0.00018899999999999999,
-      "loss": 0.4927,
+      "epoch": 63.0,
+      "learning_rate": 0.00018659999999999998,
+      "loss": 0.0574,
       "step": 315
     },
     {
-      "epoch": 5.12,
-      "learning_rate": 0.00019199999999999998,
-      "loss": 0.4879,
+      "epoch": 64.0,
+      "learning_rate": 0.00018959999999999997,
+      "loss": 0.0577,
       "step": 320
     },
     {
-      "epoch": 5.2,
-      "learning_rate": 0.000195,
-      "loss": 0.4825,
+      "epoch": 65.0,
+      "learning_rate": 0.0001926,
+      "loss": 0.0597,
       "step": 325
     },
     {
-      "epoch": 5.28,
-      "learning_rate": 0.000198,
-      "loss": 0.4841,
+      "epoch": 66.0,
+      "learning_rate": 0.00019559999999999998,
+      "loss": 0.0546,
       "step": 330
     },
     {
-      "epoch": 5.36,
-      "learning_rate": 0.000201,
-      "loss": 0.484,
+      "epoch": 67.0,
+      "learning_rate": 0.0001986,
+      "loss": 0.0603,
       "step": 335
     },
     {
-      "epoch": 5.44,
-      "learning_rate": 0.000204,
-      "loss": 0.4777,
+      "epoch": 68.0,
+      "learning_rate": 0.0002016,
+      "loss": 0.0555,
       "step": 340
     },
     {
-      "epoch": 5.52,
-      "learning_rate": 0.00020699999999999996,
-      "loss": 0.4777,
+      "epoch": 69.0,
+      "learning_rate": 0.00020459999999999999,
+      "loss": 0.0551,
       "step": 345
     },
     {
-      "epoch": 5.6,
-      "learning_rate": 0.00020999999999999998,
-      "loss": 0.4834,
+      "epoch": 70.0,
+      "learning_rate": 0.00020759999999999998,
+      "loss": 0.0529,
+      "step": 350
+    },
+    {
+      "epoch": 70.0,
+      "eval_loss": 1.901442527770996,
+      "eval_runtime": 1.3785,
+      "eval_samples_per_second": 7.254,
+      "eval_steps_per_second": 1.451,
       "step": 350
     },
     {
-      "epoch": 5.68,
-      "learning_rate": 0.00021299999999999997,
-      "loss": 0.4724,
+      "epoch": 71.0,
+      "learning_rate": 0.00021059999999999997,
+      "loss": 0.051,
       "step": 355
     },
     {
-      "epoch": 5.76,
-      "learning_rate": 0.00021599999999999996,
-      "loss": 0.4777,
+      "epoch": 72.0,
+      "learning_rate": 0.00021359999999999996,
+      "loss": 0.05,
       "step": 360
     },
     {
-      "epoch": 5.84,
-      "learning_rate": 0.00021899999999999998,
-      "loss": 0.4799,
+      "epoch": 73.0,
+      "learning_rate": 0.00021659999999999998,
+      "loss": 0.0493,
       "step": 365
     },
     {
-      "epoch": 5.92,
-      "learning_rate": 0.00022199999999999998,
-      "loss": 0.4771,
+      "epoch": 74.0,
+      "learning_rate": 0.00021959999999999997,
+      "loss": 0.0465,
       "step": 370
     },
     {
-      "epoch": 6.0,
-      "learning_rate": 0.000225,
-      "loss": 0.4771,
+      "epoch": 75.0,
+      "learning_rate": 0.0002226,
+      "loss": 0.0504,
       "step": 375
     },
     {
-      "epoch": 6.08,
-      "learning_rate": 0.00022799999999999999,
-      "loss": 0.4671,
+      "epoch": 76.0,
+      "learning_rate": 0.00022559999999999998,
+      "loss": 0.0491,
       "step": 380
     },
     {
-      "epoch": 6.16,
-      "learning_rate": 0.00023099999999999998,
-      "loss": 0.465,
+      "epoch": 77.0,
+      "learning_rate": 0.00022859999999999997,
+      "loss": 0.0485,
       "step": 385
     },
     {
-      "epoch": 6.24,
-      "learning_rate": 0.000234,
-      "loss": 0.4634,
+      "epoch": 78.0,
+      "learning_rate": 0.0002316,
+      "loss": 0.0451,
       "step": 390
     },
     {
-      "epoch": 6.32,
-      "learning_rate": 0.000237,
-      "loss": 0.4656,
+      "epoch": 79.0,
+      "learning_rate": 0.00023459999999999998,
+      "loss": 0.0478,
       "step": 395
     },
     {
-      "epoch": 6.4,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 0.4726,
+      "epoch": 80.0,
+      "learning_rate": 0.0002376,
+      "loss": 0.0435,
       "step": 400
     },
     {
-      "epoch": 6.48,
-      "learning_rate": 0.000243,
-      "loss": 0.4679,
+      "epoch": 80.0,
+      "eval_loss": 1.9840269088745117,
+      "eval_runtime": 1.3787,
+      "eval_samples_per_second": 7.253,
+      "eval_steps_per_second": 1.451,
+      "step": 400
+    },
+    {
+      "epoch": 81.0,
+      "learning_rate": 0.0002406,
+      "loss": 0.0429,
       "step": 405
     },
     {
-      "epoch": 6.56,
-      "learning_rate": 0.00024599999999999996,
-      "loss": 0.467,
+      "epoch": 82.0,
+      "learning_rate": 0.00024359999999999999,
+      "loss": 0.0464,
       "step": 410
     },
     {
-      "epoch": 6.64,
-      "learning_rate": 0.000249,
-      "loss": 0.4642,
+      "epoch": 83.0,
+      "learning_rate": 0.0002466,
+      "loss": 0.0458,
       "step": 415
     },
     {
-      "epoch": 6.72,
-      "learning_rate": 0.00025199999999999995,
-      "loss": 0.462,
+      "epoch": 84.0,
+      "learning_rate": 0.00024959999999999994,
+      "loss": 0.0441,
       "step": 420
     },
     {
-      "epoch": 6.8,
-      "learning_rate": 0.00025499999999999996,
-      "loss": 0.4599,
+      "epoch": 85.0,
+      "learning_rate": 0.00025259999999999996,
+      "loss": 0.0421,
       "step": 425
     },
     {
-      "epoch": 6.88,
-      "learning_rate": 0.000258,
-      "loss": 0.4569,
+      "epoch": 86.0,
+      "learning_rate": 0.0002556,
+      "loss": 0.0433,
       "step": 430
     },
     {
-      "epoch": 6.96,
-      "learning_rate": 0.000261,
-      "loss": 0.4644,
+      "epoch": 87.0,
+      "learning_rate": 0.0002586,
+      "loss": 0.0444,
       "step": 435
     },
     {
-      "epoch": 7.04,
-      "learning_rate": 0.00026399999999999997,
-      "loss": 0.4594,
+      "epoch": 88.0,
+      "learning_rate": 0.00026159999999999996,
+      "loss": 0.0472,
       "step": 440
     },
     {
-      "epoch": 7.12,
-      "learning_rate": 0.000267,
-      "loss": 0.4524,
+      "epoch": 89.0,
+      "learning_rate": 0.0002646,
+      "loss": 0.0442,
       "step": 445
     },
     {
-      "epoch": 7.2,
-      "learning_rate": 0.00027,
-      "loss": 0.4547,
+      "epoch": 90.0,
+      "learning_rate": 0.0002676,
+      "loss": 0.0431,
+      "step": 450
+    },
+    {
+      "epoch": 90.0,
+      "eval_loss": 1.839582085609436,
+      "eval_runtime": 1.3848,
+      "eval_samples_per_second": 7.221,
+      "eval_steps_per_second": 1.444,
       "step": 450
     },
     {
-      "epoch": 7.28,
-      "learning_rate": 0.00027299999999999997,
-      "loss": 0.4565,
+      "epoch": 91.0,
+      "learning_rate": 0.00027059999999999996,
+      "loss": 0.0431,
       "step": 455
     },
     {
-      "epoch": 7.36,
-      "learning_rate": 0.000276,
-      "loss": 0.4521,
+      "epoch": 92.0,
+      "learning_rate": 0.0002736,
+      "loss": 0.044,
       "step": 460
     },
     {
-      "epoch": 7.44,
-      "learning_rate": 0.000279,
-      "loss": 0.4557,
+      "epoch": 93.0,
+      "learning_rate": 0.0002766,
+      "loss": 0.0429,
       "step": 465
     },
     {
-      "epoch": 7.52,
-      "learning_rate": 0.00028199999999999997,
-      "loss": 0.4541,
+      "epoch": 94.0,
+      "learning_rate": 0.00027959999999999997,
+      "loss": 0.042,
       "step": 470
     },
     {
-      "epoch": 7.6,
-      "learning_rate": 0.000285,
-      "loss": 0.4537,
+      "epoch": 95.0,
+      "learning_rate": 0.0002826,
+      "loss": 0.0415,
       "step": 475
     },
     {
-      "epoch": 7.68,
-      "learning_rate": 0.00028799999999999995,
-      "loss": 0.4516,
+      "epoch": 96.0,
+      "learning_rate": 0.00028559999999999995,
+      "loss": 0.0433,
       "step": 480
     },
     {
-      "epoch": 7.76,
-      "learning_rate": 0.00029099999999999997,
-      "loss": 0.4423,
+      "epoch": 97.0,
+      "learning_rate": 0.00028859999999999997,
+      "loss": 0.0405,
       "step": 485
     },
     {
-      "epoch": 7.84,
-      "learning_rate": 0.000294,
-      "loss": 0.4479,
+      "epoch": 98.0,
+      "learning_rate": 0.0002916,
+      "loss": 0.0416,
       "step": 490
     },
     {
-      "epoch": 7.92,
-      "learning_rate": 0.00029699999999999996,
-      "loss": 0.4484,
+      "epoch": 99.0,
+      "learning_rate": 0.00029459999999999995,
+      "loss": 0.0378,
       "step": 495
     },
     {
-      "epoch": 8.0,
-      "learning_rate": 0.0003,
-      "loss": 0.4589,
+      "epoch": 100.0,
+      "learning_rate": 0.00029759999999999997,
+      "loss": 0.0379,
       "step": 500
     },
     {
-      "epoch": 8.0,
-      "eval_loss": 0.4642273485660553,
-      "eval_runtime": 264.6645,
-      "eval_samples_per_second": 7.561,
-      "eval_steps_per_second": 0.948,
+      "epoch": 100.0,
+      "eval_loss": 1.9192386865615845,
+      "eval_runtime": 1.3756,
+      "eval_samples_per_second": 7.269,
+      "eval_steps_per_second": 1.454,
       "step": 500
-    },
-    {
-      "epoch": 8.08,
-      "learning_rate": 0.00029984210526315787,
-      "loss": 0.441,
-      "step": 505
-    },
-    {
-      "epoch": 8.16,
-      "learning_rate": 0.00029968421052631577,
-      "loss": 0.4404,
-      "step": 510
-    },
-    {
-      "epoch": 8.24,
-      "learning_rate": 0.00029952631578947366,
-      "loss": 0.4452,
-      "step": 515
-    },
-    {
-      "epoch": 8.32,
-      "learning_rate": 0.00029936842105263156,
-      "loss": 0.4385,
-      "step": 520
-    },
-    {
-      "epoch": 8.4,
-      "learning_rate": 0.00029921052631578946,
-      "loss": 0.4457,
-      "step": 525
-    },
-    {
-      "epoch": 8.48,
-      "learning_rate": 0.00029905263157894735,
-      "loss": 0.4407,
-      "step": 530
-    },
-    {
-      "epoch": 8.56,
-      "learning_rate": 0.00029889473684210525,
-      "loss": 0.4432,
-      "step": 535
-    },
-    {
-      "epoch": 8.64,
-      "learning_rate": 0.00029873684210526315,
-      "loss": 0.4408,
-      "step": 540
-    },
-    {
-      "epoch": 8.72,
-      "learning_rate": 0.00029857894736842104,
-      "loss": 0.4431,
-      "step": 545
-    },
-    {
-      "epoch": 8.8,
-      "learning_rate": 0.00029842105263157894,
-      "loss": 0.4403,
-      "step": 550
-    },
-    {
-      "epoch": 8.88,
-      "learning_rate": 0.0002982631578947368,
-      "loss": 0.4357,
-      "step": 555
-    },
-    {
-      "epoch": 8.96,
-      "learning_rate": 0.00029810526315789473,
-      "loss": 0.4411,
-      "step": 560
-    },
-    {
-      "epoch": 9.04,
-      "learning_rate": 0.00029794736842105263,
-      "loss": 0.4347,
-      "step": 565
-    },
-    {
-      "epoch": 9.12,
-      "learning_rate": 0.0002977894736842105,
-      "loss": 0.4317,
-      "step": 570
-    },
-    {
-      "epoch": 9.2,
-      "learning_rate": 0.00029763157894736837,
-      "loss": 0.4332,
-      "step": 575
-    },
-    {
-      "epoch": 9.28,
-      "learning_rate": 0.00029747368421052627,
-      "loss": 0.4261,
-      "step": 580
-    },
-    {
-      "epoch": 9.36,
-      "learning_rate": 0.0002973157894736842,
-      "loss": 0.4348,
-      "step": 585
-    },
-    {
-      "epoch": 9.44,
-      "learning_rate": 0.00029715789473684206,
-      "loss": 0.434,
-      "step": 590
-    },
-    {
-      "epoch": 9.52,
-      "learning_rate": 0.00029699999999999996,
-      "loss": 0.4334,
-      "step": 595
-    },
-    {
-      "epoch": 9.6,
-      "learning_rate": 0.00029684210526315785,
-      "loss": 0.4348,
-      "step": 600
-    },
-    {
-      "epoch": 9.68,
-      "learning_rate": 0.0002966842105263158,
-      "loss": 0.4292,
-      "step": 605
-    },
-    {
-      "epoch": 9.76,
-      "learning_rate": 0.00029652631578947364,
-      "loss": 0.4332,
-      "step": 610
-    },
-    {
-      "epoch": 9.84,
-      "learning_rate": 0.00029636842105263154,
-      "loss": 0.4302,
-      "step": 615
-    },
-    {
-      "epoch": 9.92,
-      "learning_rate": 0.00029621052631578944,
-      "loss": 0.4303,
-      "step": 620
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 0.00029605263157894733,
-      "loss": 0.427,
-      "step": 625
-    },
-    {
-      "epoch": 10.08,
-      "learning_rate": 0.00029589473684210523,
-      "loss": 0.4234,
-      "step": 630
-    },
-    {
-      "epoch": 10.16,
-      "learning_rate": 0.00029573684210526313,
-      "loss": 0.4251,
-      "step": 635
-    },
-    {
-      "epoch": 10.24,
-      "learning_rate": 0.000295578947368421,
-      "loss": 0.4208,
-      "step": 640
-    },
-    {
-      "epoch": 10.32,
-      "learning_rate": 0.0002954210526315789,
-      "loss": 0.4252,
-      "step": 645
-    },
-    {
-      "epoch": 10.4,
-      "learning_rate": 0.0002952631578947368,
-      "loss": 0.4263,
-      "step": 650
-    },
-    {
-      "epoch": 10.48,
-      "learning_rate": 0.0002951052631578947,
-      "loss": 0.4221,
-      "step": 655
-    },
-    {
-      "epoch": 10.56,
-      "learning_rate": 0.0002949473684210526,
-      "loss": 0.4169,
-      "step": 660
-    },
-    {
-      "epoch": 10.64,
-      "learning_rate": 0.0002947894736842105,
-      "loss": 0.4282,
-      "step": 665
-    },
-    {
-      "epoch": 10.72,
-      "learning_rate": 0.0002946315789473684,
-      "loss": 0.4207,
-      "step": 670
-    },
-    {
-      "epoch": 10.8,
-      "learning_rate": 0.0002944736842105263,
-      "loss": 0.4257,
-      "step": 675
-    },
-    {
-      "epoch": 10.88,
-      "learning_rate": 0.0002943157894736842,
-      "loss": 0.4207,
-      "step": 680
-    },
-    {
-      "epoch": 10.96,
-      "learning_rate": 0.0002941578947368421,
-      "loss": 0.4171,
-      "step": 685
-    },
-    {
-      "epoch": 11.04,
-      "learning_rate": 0.000294,
-      "loss": 0.4215,
-      "step": 690
-    },
-    {
-      "epoch": 11.12,
-      "learning_rate": 0.00029384210526315783,
-      "loss": 0.4112,
-      "step": 695
-    },
-    {
-      "epoch": 11.2,
-      "learning_rate": 0.0002936842105263158,
-      "loss": 0.4127,
-      "step": 700
-    },
-    {
-      "epoch": 11.28,
-      "learning_rate": 0.0002935263157894737,
-      "loss": 0.4125,
-      "step": 705
-    },
-    {
-      "epoch": 11.36,
-      "learning_rate": 0.0002933684210526316,
-      "loss": 0.4149,
-      "step": 710
-    },
-    {
-      "epoch": 11.44,
-      "learning_rate": 0.0002932105263157894,
-      "loss": 0.4144,
-      "step": 715
-    },
-    {
-      "epoch": 11.52,
-      "learning_rate": 0.0002930526315789473,
-      "loss": 0.4157,
-      "step": 720
-    },
-    {
-      "epoch": 11.6,
-      "learning_rate": 0.00029289473684210527,
-      "loss": 0.4087,
-      "step": 725
-    },
-    {
-      "epoch": 11.68,
-      "learning_rate": 0.00029273684210526316,
-      "loss": 0.4162,
-      "step": 730
-    },
-    {
-      "epoch": 11.76,
-      "learning_rate": 0.000292578947368421,
-      "loss": 0.4104,
-      "step": 735
-    },
-    {
-      "epoch": 11.84,
-      "learning_rate": 0.0002924210526315789,
-      "loss": 0.414,
-      "step": 740
-    },
-    {
-      "epoch": 11.92,
-      "learning_rate": 0.00029226315789473685,
-      "loss": 0.4132,
-      "step": 745
-    },
-    {
-      "epoch": 12.0,
-      "learning_rate": 0.0002921052631578947,
-      "loss": 0.4223,
-      "step": 750
-    },
-    {
-      "epoch": 12.0,
-      "eval_loss": 0.44360578060150146,
-      "eval_runtime": 265.1184,
-      "eval_samples_per_second": 7.548,
-      "eval_steps_per_second": 0.947,
-      "step": 750
-    },
-    {
-      "epoch": 12.08,
-      "learning_rate": 0.0002919473684210526,
-      "loss": 0.4084,
-      "step": 755
-    },
-    {
-      "epoch": 12.16,
-      "learning_rate": 0.0002917894736842105,
-      "loss": 0.4045,
-      "step": 760
-    },
-    {
-      "epoch": 12.24,
-      "learning_rate": 0.0002916315789473684,
-      "loss": 0.4084,
-      "step": 765
-    },
-    {
-      "epoch": 12.32,
-      "learning_rate": 0.0002914736842105263,
-      "loss": 0.4053,
-      "step": 770
-    },
-    {
-      "epoch": 12.4,
-      "learning_rate": 0.0002913157894736842,
-      "loss": 0.4079,
-      "step": 775
-    },
-    {
-      "epoch": 12.48,
-      "learning_rate": 0.0002911578947368421,
-      "loss": 0.4045,
-      "step": 780
-    },
-    {
-      "epoch": 12.56,
-      "learning_rate": 0.00029099999999999997,
-      "loss": 0.4009,
-      "step": 785
-    },
-    {
-      "epoch": 12.64,
-      "learning_rate": 0.00029084210526315787,
-      "loss": 0.4064,
-      "step": 790
-    },
-    {
-      "epoch": 12.72,
-      "learning_rate": 0.00029068421052631577,
-      "loss": 0.4104,
-      "step": 795
-    },
-    {
-      "epoch": 12.8,
-      "learning_rate": 0.00029052631578947366,
-      "loss": 0.4121,
-      "step": 800
-    },
-    {
-      "epoch": 12.88,
-      "learning_rate": 0.00029036842105263156,
-      "loss": 0.4064,
-      "step": 805
-    },
-    {
-      "epoch": 12.96,
-      "learning_rate": 0.00029021052631578945,
-      "loss": 0.4126,
-      "step": 810
-    },
-    {
-      "epoch": 13.04,
-      "learning_rate": 0.00029005263157894735,
-      "loss": 0.4028,
-      "step": 815
-    },
-    {
-      "epoch": 13.12,
-      "learning_rate": 0.00028989473684210525,
-      "loss": 0.3973,
-      "step": 820
-    },
-    {
-      "epoch": 13.2,
-      "learning_rate": 0.00028973684210526314,
-      "loss": 0.3977,
-      "step": 825
-    },
-    {
-      "epoch": 13.28,
-      "learning_rate": 0.00028957894736842104,
-      "loss": 0.3989,
-      "step": 830
-    },
-    {
-      "epoch": 13.36,
-      "learning_rate": 0.0002894210526315789,
-      "loss": 0.4025,
-      "step": 835
-    },
-    {
-      "epoch": 13.44,
-      "learning_rate": 0.00028926315789473683,
-      "loss": 0.4013,
-      "step": 840
-    },
-    {
-      "epoch": 13.52,
-      "learning_rate": 0.00028910526315789473,
-      "loss": 0.4044,
-      "step": 845
-    },
-    {
-      "epoch": 13.6,
-      "learning_rate": 0.00028894736842105263,
-      "loss": 0.4037,
-      "step": 850
-    },
-    {
-      "epoch": 13.68,
-      "learning_rate": 0.00028878947368421047,
-      "loss": 0.4,
-      "step": 855
-    },
-    {
-      "epoch": 13.76,
-      "learning_rate": 0.00028863157894736837,
-      "loss": 0.4023,
-      "step": 860
-    },
-    {
-      "epoch": 13.84,
-      "learning_rate": 0.0002884736842105263,
-      "loss": 0.402,
-      "step": 865
-    },
-    {
-      "epoch": 13.92,
-      "learning_rate": 0.0002883157894736842,
-      "loss": 0.4033,
-      "step": 870
-    },
-    {
-      "epoch": 14.0,
-      "learning_rate": 0.00028815789473684206,
-      "loss": 0.4009,
-      "step": 875
-    },
-    {
-      "epoch": 14.08,
-      "learning_rate": 0.00028799999999999995,
-      "loss": 0.3952,
-      "step": 880
-    },
-    {
-      "epoch": 14.16,
-      "learning_rate": 0.0002878421052631579,
-      "loss": 0.3911,
-      "step": 885
-    },
-    {
-      "epoch": 14.24,
-      "learning_rate": 0.00028768421052631575,
-      "loss": 0.392,
-      "step": 890
-    },
-    {
-      "epoch": 14.32,
-      "learning_rate": 0.00028752631578947364,
-      "loss": 0.3912,
-      "step": 895
-    },
-    {
-      "epoch": 14.4,
-      "learning_rate": 0.00028736842105263154,
-      "loss": 0.3918,
-      "step": 900
-    },
-    {
-      "epoch": 14.48,
-      "learning_rate": 0.00028721052631578944,
-      "loss": 0.3955,
-      "step": 905
-    },
-    {
-      "epoch": 14.56,
-      "learning_rate": 0.00028705263157894733,
-      "loss": 0.3979,
-      "step": 910
-    },
-    {
-      "epoch": 14.64,
-      "learning_rate": 0.00028689473684210523,
-      "loss": 0.396,
-      "step": 915
-    },
-    {
-      "epoch": 14.72,
-      "learning_rate": 0.0002867368421052631,
-      "loss": 0.3957,
-      "step": 920
-    },
-    {
-      "epoch": 14.8,
-      "learning_rate": 0.000286578947368421,
-      "loss": 0.3975,
-      "step": 925
-    },
-    {
-      "epoch": 14.88,
-      "learning_rate": 0.0002864210526315789,
-      "loss": 0.398,
-      "step": 930
-    },
-    {
-      "epoch": 14.96,
-      "learning_rate": 0.0002862631578947368,
-      "loss": 0.3965,
-      "step": 935
-    },
-    {
-      "epoch": 15.04,
-      "learning_rate": 0.0002861052631578947,
-      "loss": 0.3906,
-      "step": 940
-    },
-    {
-      "epoch": 15.12,
-      "learning_rate": 0.0002859473684210526,
-      "loss": 0.3852,
-      "step": 945
-    },
-    {
-      "epoch": 15.2,
-      "learning_rate": 0.0002857894736842105,
-      "loss": 0.3933,
-      "step": 950
-    },
-    {
-      "epoch": 15.28,
-      "learning_rate": 0.0002856315789473684,
-      "loss": 0.3857,
-      "step": 955
-    },
-    {
-      "epoch": 15.36,
-      "learning_rate": 0.0002854736842105263,
-      "loss": 0.3897,
-      "step": 960
-    },
-    {
-      "epoch": 15.44,
-      "learning_rate": 0.0002853157894736842,
-      "loss": 0.3885,
-      "step": 965
-    },
-    {
-      "epoch": 15.52,
-      "learning_rate": 0.0002851578947368421,
-      "loss": 0.3882,
-      "step": 970
-    },
-    {
-      "epoch": 15.6,
-      "learning_rate": 0.000285,
-      "loss": 0.3961,
-      "step": 975
-    },
-    {
-      "epoch": 15.68,
-      "learning_rate": 0.0002848421052631579,
-      "loss": 0.3886,
-      "step": 980
-    },
-    {
-      "epoch": 15.76,
-      "learning_rate": 0.0002846842105263158,
-      "loss": 0.391,
-      "step": 985
-    },
-    {
-      "epoch": 15.84,
-      "learning_rate": 0.0002845263157894737,
-      "loss": 0.391,
-      "step": 990
-    },
-    {
-      "epoch": 15.92,
-      "learning_rate": 0.0002843684210526315,
-      "loss": 0.3901,
-      "step": 995
-    },
-    {
-      "epoch": 16.0,
-      "learning_rate": 0.0002842105263157894,
-      "loss": 0.3869,
-      "step": 1000
-    },
-    {
-      "epoch": 16.0,
-      "eval_loss": 0.43608614802360535,
-      "eval_runtime": 265.3444,
-      "eval_samples_per_second": 7.541,
-      "eval_steps_per_second": 0.946,
-      "step": 1000
     }
   ],
   "max_steps": 10000,
-  "num_train_epochs": 162,
-  "total_flos": 2.599744722936791e+18,
+  "num_train_epochs": 2000,
+  "total_flos": 8.0956959522816e+16,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:938e3a85e1584e1669ed7b89d0c0c72ab3ee7f45a24587d91ef691e943ee42d6
+oid sha256:ccf3cbed56dd6bc8e51a2fe8aa833170620c235dac2584c4369f03afd0d2c443
 size 3963