Training in progress, step 200, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cad8695f9611fd72a1cb7da7c2dc7722ffd94491b413b08931297f19a10ee048
+oid sha256:e5742327392312dd1a7e23ca028cd9e6187add4ba8fb99e3a28ce17d8dbe3b26
 size 234681136

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0f17b41e804bbacf24053241af7cfe65159104d0213019d3237630b978e42eda
+oid sha256:d2a1a873a10876a3541cbda1bda6ef5a52202b10618859918417990065cc8b5c
 size 469516363

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76cd281a3ce995fc434ac757f1094066d955dd6b7188cf148f0dd1523cb9cd80
+oid sha256:78f00cbf2a3a4305cbfbe0a453688d1bb898224da36be2f067d77db0d73462c1
 size 14645

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e63258c783b6bbb756d9067a003d6c6c729901571b495c03dd6472af6b5a7f7b
+oid sha256:4dcd0fcafdf0cac56c6781c9554680e5d2ba9bfee0d823975791d1d648a00172
 size 1465

last-checkpoint/trainer_state.json

@@ -2,409 +2,331 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.07142857142857142,
-  "eval_steps": 100,
-  "global_step": 500,
+  "epoch": 0.06666666666666667,
+  "eval_steps": 50,
+  "global_step": 200,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0014285714285714286,
-      "grad_norm": 0.060791015625,
-      "learning_rate": 8.571428571428572e-07,
-      "loss": 2.637472915649414,
-      "step": 10
+      "epoch": 0.0016666666666666668,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.6374650955200196,
+      "step": 5
     },
     {
-      "epoch": 0.002857142857142857,
-      "grad_norm": 0.053955078125,
-      "learning_rate": 1.8095238095238097e-06,
-      "loss": 2.6134210586547852,
-      "step": 20
-    },
-    {
-      "epoch": 0.004285714285714286,
-      "grad_norm": 0.05126953125,
-      "learning_rate": 2.7619047619047625e-06,
-      "loss": 2.617936134338379,
-      "step": 30
+      "epoch": 0.0033333333333333335,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 3.6e-05,
+      "loss": 2.6133955001831053,
+      "step": 10
     },
     {
-      "epoch": 0.005714285714285714,
-      "grad_norm": 0.052978515625,
-      "learning_rate": 3.7142857142857146e-06,
-      "loss": 2.6247318267822264,
-      "step": 40
+      "epoch": 0.005,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 2.6178274154663086,
+      "step": 15
     },
     {
-      "epoch": 0.007142857142857143,
-      "grad_norm": 0.049560546875,
-      "learning_rate": 4.666666666666667e-06,
-      "loss": 2.6199132919311525,
-      "step": 50
+      "epoch": 0.006666666666666667,
+      "grad_norm": 0.037109375,
+      "learning_rate": 7.6e-05,
+      "loss": 2.624565315246582,
+      "step": 20
     },
     {
-      "epoch": 0.008571428571428572,
-      "grad_norm": 0.049560546875,
-      "learning_rate": 5.619047619047619e-06,
-      "loss": 2.6097543716430662,
-      "step": 60
+      "epoch": 0.008333333333333333,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.6e-05,
+      "loss": 2.619654655456543,
+      "step": 25
     },
     {
       "epoch": 0.01,
-      "grad_norm": 0.10986328125,
-      "learning_rate": 6.571428571428572e-06,
-      "loss": 2.6390161514282227,
-      "step": 70
+      "grad_norm": 0.036376953125,
+      "learning_rate": 0.0001,
+      "loss": 2.609416389465332,
+      "step": 30
     },
     {
-      "epoch": 0.011428571428571429,
-      "grad_norm": 0.3671875,
-      "learning_rate": 7.523809523809524e-06,
-      "loss": 2.6347326278686523,
-      "step": 80
+      "epoch": 0.011666666666666667,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0001,
+      "loss": 2.6385807037353515,
+      "step": 35
     },
     {
-      "epoch": 0.012857142857142857,
-      "grad_norm": 0.796875,
-      "learning_rate": 8.476190476190477e-06,
-      "loss": 2.6170269012451173,
-      "step": 90
-    },
-    {
-      "epoch": 0.014285714285714285,
-      "grad_norm": 5.34375,
-      "learning_rate": 9.42857142857143e-06,
-      "loss": 2.62125186920166,
-      "step": 100
+      "epoch": 0.013333333333333334,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 0.0001,
+      "loss": 2.6341808319091795,
+      "step": 40
     },
     {
-      "epoch": 0.014285714285714285,
-      "eval_loss": 3.02799129486084,
-      "eval_runtime": 4.4776,
-      "eval_samples_per_second": 66.554,
-      "eval_steps_per_second": 3.797,
-      "step": 100
+      "epoch": 0.015,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 0.0001,
+      "loss": 2.6163944244384765,
+      "step": 45
     },
     {
-      "epoch": 0.015714285714285715,
-      "grad_norm": 0.4140625,
-      "learning_rate": 1.0380952380952383e-05,
-      "loss": 2.6266862869262697,
-      "step": 110
+      "epoch": 0.016666666666666666,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001,
+      "loss": 2.620637893676758,
+      "step": 50
     },
     {
-      "epoch": 0.017142857142857144,
-      "grad_norm": 0.0634765625,
-      "learning_rate": 1.1333333333333334e-05,
-      "loss": 2.646818161010742,
-      "step": 120
+      "epoch": 0.016666666666666666,
+      "eval_loss": 3.0298774242401123,
+      "eval_runtime": 4.5343,
+      "eval_samples_per_second": 65.722,
+      "eval_steps_per_second": 3.749,
+      "step": 50
     },
     {
-      "epoch": 0.018571428571428572,
-      "grad_norm": 0.3671875,
-      "learning_rate": 1.2285714285714288e-05,
-      "loss": 2.616552543640137,
-      "step": 130
+      "epoch": 0.018333333333333333,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001,
+      "loss": 2.6260648727416993,
+      "step": 55
     },
     {
       "epoch": 0.02,
-      "grad_norm": 0.04638671875,
-      "learning_rate": 1.3238095238095238e-05,
-      "loss": 2.619216728210449,
-      "step": 140
-    },
-    {
-      "epoch": 0.02142857142857143,
-      "grad_norm": 0.0498046875,
-      "learning_rate": 1.4190476190476192e-05,
-      "loss": 2.652239990234375,
-      "step": 150
-    },
-    {
-      "epoch": 0.022857142857142857,
-      "grad_norm": 0.057373046875,
-      "learning_rate": 1.5142857142857144e-05,
-      "loss": 2.5990116119384767,
-      "step": 160
+      "grad_norm": 0.047607421875,
+      "learning_rate": 0.0001,
+      "loss": 2.6460687637329103,
+      "step": 60
     },
     {
-      "epoch": 0.024285714285714285,
-      "grad_norm": 0.4140625,
-      "learning_rate": 1.6095238095238096e-05,
-      "loss": 2.635771369934082,
-      "step": 170
+      "epoch": 0.021666666666666667,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 0.0001,
+      "loss": 2.6157953262329103,
+      "step": 65
     },
     {
-      "epoch": 0.025714285714285714,
-      "grad_norm": 0.0517578125,
-      "learning_rate": 1.704761904761905e-05,
-      "loss": 2.617422103881836,
-      "step": 180
+      "epoch": 0.023333333333333334,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0001,
+      "loss": 2.61843318939209,
+      "step": 70
     },
     {
-      "epoch": 0.027142857142857142,
-      "grad_norm": 0.064453125,
-      "learning_rate": 1.8e-05,
-      "loss": 2.64025764465332,
-      "step": 190
+      "epoch": 0.025,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0001,
+      "loss": 2.651448059082031,
+      "step": 75
     },
     {
-      "epoch": 0.02857142857142857,
-      "grad_norm": 0.046875,
-      "learning_rate": 1.8952380952380953e-05,
-      "loss": 2.6089372634887695,
-      "step": 200
+      "epoch": 0.02666666666666667,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 0.0001,
+      "loss": 2.59826545715332,
+      "step": 80
     },
     {
-      "epoch": 0.02857142857142857,
-      "eval_loss": 3.0282089710235596,
-      "eval_runtime": 4.3641,
-      "eval_samples_per_second": 68.284,
-      "eval_steps_per_second": 3.895,
-      "step": 200
+      "epoch": 0.028333333333333332,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0001,
+      "loss": 2.6349016189575196,
+      "step": 85
     },
     {
       "epoch": 0.03,
-      "grad_norm": 0.0537109375,
-      "learning_rate": 1.9904761904761908e-05,
-      "loss": 2.6311504364013674,
-      "step": 210
+      "grad_norm": 0.036376953125,
+      "learning_rate": 0.0001,
+      "loss": 2.6166595458984374,
+      "step": 90
     },
     {
-      "epoch": 0.03142857142857143,
-      "grad_norm": 0.05078125,
-      "learning_rate": 2e-05,
-      "loss": 2.6472368240356445,
-      "step": 220
+      "epoch": 0.03166666666666667,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001,
+      "loss": 2.6394269943237303,
+      "step": 95
     },
     {
-      "epoch": 0.032857142857142856,
-      "grad_norm": 0.05322265625,
-      "learning_rate": 2e-05,
-      "loss": 2.6085578918457033,
-      "step": 230
+      "epoch": 0.03333333333333333,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 0.0001,
+      "loss": 2.6080867767333986,
+      "step": 100
     },
     {
-      "epoch": 0.03428571428571429,
-      "grad_norm": 0.046630859375,
-      "learning_rate": 2e-05,
-      "loss": 2.6177978515625,
-      "step": 240
+      "epoch": 0.03333333333333333,
+      "eval_loss": 3.0300662517547607,
+      "eval_runtime": 4.4357,
+      "eval_samples_per_second": 67.181,
+      "eval_steps_per_second": 3.832,
+      "step": 100
     },
     {
-      "epoch": 0.03571428571428571,
-      "grad_norm": 0.06298828125,
-      "learning_rate": 2e-05,
-      "loss": 2.6038957595825196,
-      "step": 250
+      "epoch": 0.035,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 0.0001,
+      "loss": 2.6302499771118164,
+      "step": 105
     },
     {
-      "epoch": 0.037142857142857144,
-      "grad_norm": 0.048095703125,
-      "learning_rate": 2e-05,
-      "loss": 2.6111982345581053,
-      "step": 260
+      "epoch": 0.03666666666666667,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 0.0001,
+      "loss": 2.646461296081543,
+      "step": 110
     },
     {
-      "epoch": 0.03857142857142857,
-      "grad_norm": 0.04833984375,
-      "learning_rate": 2e-05,
-      "loss": 2.600232696533203,
-      "step": 270
+      "epoch": 0.03833333333333333,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0001,
+      "loss": 2.6076818466186524,
+      "step": 115
     },
     {
       "epoch": 0.04,
-      "grad_norm": 0.046630859375,
-      "learning_rate": 2e-05,
-      "loss": 2.591439056396484,
-      "step": 280
-    },
-    {
-      "epoch": 0.041428571428571426,
-      "grad_norm": 0.04833984375,
-      "learning_rate": 2e-05,
-      "loss": 2.6285400390625,
-      "step": 290
-    },
-    {
-      "epoch": 0.04285714285714286,
-      "grad_norm": 40.0,
-      "learning_rate": 2e-05,
-      "loss": 2.6076997756958007,
-      "step": 300
+      "grad_norm": 0.032470703125,
+      "learning_rate": 0.0001,
+      "loss": 2.616958236694336,
+      "step": 120
     },
     {
-      "epoch": 0.04285714285714286,
-      "eval_loss": 3.02835750579834,
-      "eval_runtime": 4.3773,
-      "eval_samples_per_second": 68.079,
-      "eval_steps_per_second": 3.884,
-      "step": 300
+      "epoch": 0.041666666666666664,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0001,
+      "loss": 2.603002166748047,
+      "step": 125
     },
     {
-      "epoch": 0.04428571428571428,
-      "grad_norm": 0.049072265625,
-      "learning_rate": 2e-05,
-      "loss": 2.618265914916992,
-      "step": 310
+      "epoch": 0.043333333333333335,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 0.0001,
+      "loss": 2.6103355407714846,
+      "step": 130
     },
     {
-      "epoch": 0.045714285714285714,
-      "grad_norm": 0.044677734375,
-      "learning_rate": 2e-05,
-      "loss": 2.6028215408325197,
-      "step": 320
+      "epoch": 0.045,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001,
+      "loss": 2.599276542663574,
+      "step": 135
     },
     {
-      "epoch": 0.047142857142857146,
-      "grad_norm": 0.048828125,
-      "learning_rate": 2e-05,
-      "loss": 2.6143060684204102,
-      "step": 330
+      "epoch": 0.04666666666666667,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 0.0001,
+      "loss": 2.590505599975586,
+      "step": 140
     },
     {
-      "epoch": 0.04857142857142857,
-      "grad_norm": 0.0498046875,
-      "learning_rate": 2e-05,
-      "loss": 2.602345275878906,
-      "step": 340
+      "epoch": 0.04833333333333333,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 0.0001,
+      "loss": 2.627636528015137,
+      "step": 145
     },
     {
       "epoch": 0.05,
-      "grad_norm": 0.06494140625,
-      "learning_rate": 2e-05,
-      "loss": 2.597218704223633,
-      "step": 350
-    },
-    {
-      "epoch": 0.05142857142857143,
-      "grad_norm": 0.07568359375,
-      "learning_rate": 2e-05,
-      "loss": 2.6153676986694334,
-      "step": 360
+      "grad_norm": 262.0,
+      "learning_rate": 0.0001,
+      "loss": 2.6068201065063477,
+      "step": 150
     },
     {
-      "epoch": 0.05285714285714286,
-      "grad_norm": 0.06787109375,
-      "learning_rate": 2e-05,
-      "loss": 2.610495948791504,
-      "step": 370
+      "epoch": 0.05,
+      "eval_loss": 3.0302553176879883,
+      "eval_runtime": 4.3841,
+      "eval_samples_per_second": 67.972,
+      "eval_steps_per_second": 3.878,
+      "step": 150
     },
     {
-      "epoch": 0.054285714285714284,
-      "grad_norm": 0.05224609375,
-      "learning_rate": 2e-05,
-      "loss": 2.6116031646728515,
-      "step": 380
+      "epoch": 0.051666666666666666,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0001,
+      "loss": 2.617295265197754,
+      "step": 155
     },
     {
-      "epoch": 0.055714285714285716,
-      "grad_norm": 0.08837890625,
-      "learning_rate": 2e-05,
-      "loss": 2.6053627014160154,
-      "step": 390
+      "epoch": 0.05333333333333334,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0001,
+      "loss": 2.601869010925293,
+      "step": 160
     },
     {
-      "epoch": 0.05714285714285714,
-      "grad_norm": 0.08447265625,
-      "learning_rate": 2e-05,
-      "loss": 2.599230194091797,
-      "step": 400
+      "epoch": 0.055,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0001,
+      "loss": 2.613315391540527,
+      "step": 165
     },
     {
-      "epoch": 0.05714285714285714,
-      "eval_loss": 3.0280556678771973,
-      "eval_runtime": 4.3932,
-      "eval_samples_per_second": 67.831,
-      "eval_steps_per_second": 3.87,
-      "step": 400
+      "epoch": 0.056666666666666664,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 0.0001,
+      "loss": 2.601426124572754,
+      "step": 170
     },
     {
-      "epoch": 0.05857142857142857,
-      "grad_norm": 0.0712890625,
-      "learning_rate": 2e-05,
-      "loss": 2.6664922714233397,
-      "step": 410
+      "epoch": 0.058333333333333334,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 0.0001,
+      "loss": 2.5963130950927735,
+      "step": 175
     },
     {
       "epoch": 0.06,
-      "grad_norm": 0.07373046875,
-      "learning_rate": 2e-05,
-      "loss": 2.6177946090698243,
-      "step": 420
-    },
-    {
-      "epoch": 0.06142857142857143,
-      "grad_norm": 0.064453125,
-      "learning_rate": 2e-05,
-      "loss": 2.6283496856689452,
-      "step": 430
-    },
-    {
-      "epoch": 0.06285714285714286,
-      "grad_norm": 0.04736328125,
-      "learning_rate": 2e-05,
-      "loss": 2.6089170455932615,
-      "step": 440
-    },
-    {
-      "epoch": 0.06428571428571428,
-      "grad_norm": 0.06298828125,
-      "learning_rate": 2e-05,
-      "loss": 2.625172233581543,
-      "step": 450
-    },
-    {
-      "epoch": 0.06571428571428571,
-      "grad_norm": 0.109375,
-      "learning_rate": 2e-05,
-      "loss": 2.6342344284057617,
-      "step": 460
+      "grad_norm": 0.036376953125,
+      "learning_rate": 0.0001,
+      "loss": 2.614262008666992,
+      "step": 180
     },
     {
-      "epoch": 0.06714285714285714,
-      "grad_norm": 0.048583984375,
-      "learning_rate": 2e-05,
-      "loss": 2.6153539657592773,
-      "step": 470
+      "epoch": 0.06166666666666667,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 0.0001,
+      "loss": 2.6094554901123046,
+      "step": 185
     },
     {
-      "epoch": 0.06857142857142857,
-      "grad_norm": 0.087890625,
-      "learning_rate": 2e-05,
-      "loss": 2.5881879806518553,
-      "step": 480
+      "epoch": 0.06333333333333334,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0001,
+      "loss": 2.6106592178344727,
+      "step": 190
     },
     {
-      "epoch": 0.07,
-      "grad_norm": 0.047607421875,
-      "learning_rate": 2e-05,
-      "loss": 2.6207881927490235,
-      "step": 490
+      "epoch": 0.065,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 0.0001,
+      "loss": 2.604322814941406,
+      "step": 195
     },
     {
-      "epoch": 0.07142857142857142,
-      "grad_norm": 0.054931640625,
-      "learning_rate": 2e-05,
-      "loss": 2.632369041442871,
-      "step": 500
+      "epoch": 0.06666666666666667,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 0.0001,
+      "loss": 2.5981882095336912,
+      "step": 200
     },
     {
-      "epoch": 0.07142857142857142,
-      "eval_loss": 3.028033494949341,
-      "eval_runtime": 4.349,
-      "eval_samples_per_second": 68.521,
-      "eval_steps_per_second": 3.909,
-      "step": 500
+      "epoch": 0.06666666666666667,
+      "eval_loss": 3.0292723178863525,
+      "eval_runtime": 4.3967,
+      "eval_samples_per_second": 67.777,
+      "eval_steps_per_second": 3.866,
+      "step": 200
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 7000,
+  "logging_steps": 5,
+  "max_steps": 3000,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
-  "save_steps": 500,
+  "save_steps": 200,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -417,7 +339,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.63077257428992e+17,
+  "total_flos": 1.304618059431936e+17,
   "train_batch_size": 18,
   "trial_name": null,
   "trial_params": null

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b1ff39bd552f28407a4692ecdd5eb017e018aa993cdeeb1b9fe03f5f01952f94
+oid sha256:2fa13de9d6de96ae7af42b5d8ce1dda570123e506d560d8d7900c3522fff7dfe
 size 5265