Upload training checkpoints after upsampling dataset

checkpoint-1000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d51378b54e35c00d17c3ac2006f9e9fdbd7cc12e52e1278758b6df39d27a87a
+oid sha256:b58314c6827ac34bcd352ce1fc7950b1d20ca398a3a892afc4dae05a21ce8916
 size 503749208

checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cce9d35e3077c44ed259aba54f709f632f5bf67a59fa794ed03bf3be1f17490a
+oid sha256:8e5bd83a18fef7f081ebc6b4706dc3434cc22b7353d2021ca7d668463d93dc46
 size 1007621899

checkpoint-1000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c9681354556c442279f8d15fdd4995c6180c0aa2fc961cc72bad8bdd7491d3e7
+oid sha256:01417872332d957944438dd9b4b8729804838252520f015df0aea72edce6b540
 size 14645

checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ca762945532cf70776f01dc50c979557aef924ea20456b955387653adde5a87
+oid sha256:e7ae580020b16556dd1a430705dd359df07f40b74863a29d02711598cf5fe41f
 size 1977

checkpoint-1000/trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_global_step": 1000,
-  "best_metric": 0.9613078781032101,
+  "best_metric": 0.9548451024890077,
   "best_model_checkpoint": "taskA-unixcoder-focal/checkpoint-1000",
-  "epoch": 0.064,
+  "epoch": 0.06117704637220115,
   "eval_steps": 500,
   "global_step": 1000,
   "is_hyper_param_search": false,
@@ -10,734 +10,734 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00064,
-      "grad_norm": 3.3732223510742188,
-      "learning_rate": 9.6e-08,
-      "loss": 0.2065,
+      "epoch": 0.0006117704637220115,
+      "grad_norm": 5.2196149826049805,
+      "learning_rate": 9.178990311065784e-08,
+      "loss": 0.2114,
       "step": 10
     },
     {
-      "epoch": 0.00128,
-      "grad_norm": 2.8028833866119385,
-      "learning_rate": 2.0266666666666668e-07,
-      "loss": 0.2003,
+      "epoch": 0.001223540927444023,
+      "grad_norm": 2.4952661991119385,
+      "learning_rate": 1.937786843447221e-07,
+      "loss": 0.179,
       "step": 20
     },
     {
-      "epoch": 0.00192,
-      "grad_norm": 3.137962818145752,
-      "learning_rate": 3.0933333333333335e-07,
-      "loss": 0.2119,
+      "epoch": 0.0018353113911660345,
+      "grad_norm": 3.8488683700561523,
+      "learning_rate": 2.9576746557878637e-07,
+      "loss": 0.1949,
       "step": 30
     },
     {
-      "epoch": 0.00256,
-      "grad_norm": 2.2841057777404785,
-      "learning_rate": 4.16e-07,
-      "loss": 0.1858,
+      "epoch": 0.002447081854888046,
+      "grad_norm": 3.7437973022460938,
+      "learning_rate": 3.9775624681285064e-07,
+      "loss": 0.2007,
       "step": 40
     },
     {
-      "epoch": 0.0032,
-      "grad_norm": 2.4040842056274414,
-      "learning_rate": 5.226666666666667e-07,
-      "loss": 0.1755,
+      "epoch": 0.0030588523186100575,
+      "grad_norm": 3.8342156410217285,
+      "learning_rate": 4.997450280469149e-07,
+      "loss": 0.1828,
       "step": 50
     },
     {
-      "epoch": 0.00384,
-      "grad_norm": 2.1078808307647705,
-      "learning_rate": 6.293333333333334e-07,
-      "loss": 0.1836,
+      "epoch": 0.003670622782332069,
+      "grad_norm": 3.510793685913086,
+      "learning_rate": 6.017338092809792e-07,
+      "loss": 0.1875,
       "step": 60
     },
     {
-      "epoch": 0.00448,
-      "grad_norm": 2.1723551750183105,
-      "learning_rate": 7.36e-07,
-      "loss": 0.182,
+      "epoch": 0.0042823932460540805,
+      "grad_norm": 3.223860740661621,
+      "learning_rate": 7.037225905150434e-07,
+      "loss": 0.1661,
       "step": 70
     },
     {
-      "epoch": 0.00512,
-      "grad_norm": 3.8697516918182373,
-      "learning_rate": 8.426666666666668e-07,
-      "loss": 0.1893,
+      "epoch": 0.004894163709776092,
+      "grad_norm": 3.791808605194092,
+      "learning_rate": 8.057113717491076e-07,
+      "loss": 0.1762,
       "step": 80
     },
     {
-      "epoch": 0.00576,
-      "grad_norm": 2.2711849212646484,
-      "learning_rate": 9.493333333333334e-07,
-      "loss": 0.1645,
+      "epoch": 0.0055059341734981036,
+      "grad_norm": 3.8649113178253174,
+      "learning_rate": 9.077001529831719e-07,
+      "loss": 0.1774,
       "step": 90
     },
     {
-      "epoch": 0.0064,
-      "grad_norm": 2.9138245582580566,
-      "learning_rate": 1.0560000000000001e-06,
-      "loss": 0.1753,
+      "epoch": 0.006117704637220115,
+      "grad_norm": 2.9645302295684814,
+      "learning_rate": 1.0096889342172361e-06,
+      "loss": 0.1741,
       "step": 100
     },
     {
-      "epoch": 0.00704,
-      "grad_norm": 2.136301279067993,
-      "learning_rate": 1.1626666666666667e-06,
-      "loss": 0.1639,
+      "epoch": 0.006729475100942127,
+      "grad_norm": 2.9103174209594727,
+      "learning_rate": 1.1116777154513005e-06,
+      "loss": 0.1808,
       "step": 110
     },
     {
-      "epoch": 0.00768,
-      "grad_norm": 3.216343641281128,
-      "learning_rate": 1.2693333333333335e-06,
-      "loss": 0.1623,
+      "epoch": 0.007341245564664138,
+      "grad_norm": 2.3954272270202637,
+      "learning_rate": 1.2136664966853646e-06,
+      "loss": 0.161,
       "step": 120
     },
     {
-      "epoch": 0.00832,
-      "grad_norm": 3.192094326019287,
-      "learning_rate": 1.376e-06,
-      "loss": 0.1607,
+      "epoch": 0.00795301602838615,
+      "grad_norm": 3.030513048171997,
+      "learning_rate": 1.3156552779194289e-06,
+      "loss": 0.1689,
       "step": 130
     },
     {
-      "epoch": 0.00896,
-      "grad_norm": 4.722238540649414,
-      "learning_rate": 1.4826666666666666e-06,
-      "loss": 0.1653,
+      "epoch": 0.008564786492108161,
+      "grad_norm": 1.7593587636947632,
+      "learning_rate": 1.4176440591534932e-06,
+      "loss": 0.1595,
       "step": 140
     },
     {
-      "epoch": 0.0096,
-      "grad_norm": 4.088972568511963,
-      "learning_rate": 1.5893333333333336e-06,
-      "loss": 0.1557,
+      "epoch": 0.009176556955830173,
+      "grad_norm": 1.9787609577178955,
+      "learning_rate": 1.5196328403875573e-06,
+      "loss": 0.1429,
       "step": 150
     },
     {
-      "epoch": 0.01024,
-      "grad_norm": 2.2773237228393555,
-      "learning_rate": 1.6960000000000002e-06,
-      "loss": 0.1634,
+      "epoch": 0.009788327419552184,
+      "grad_norm": 2.549748659133911,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.154,
       "step": 160
     },
     {
-      "epoch": 0.01088,
-      "grad_norm": 2.53987193107605,
-      "learning_rate": 1.8026666666666667e-06,
-      "loss": 0.1435,
+      "epoch": 0.010400097883274196,
+      "grad_norm": 2.4056687355041504,
+      "learning_rate": 1.723610402855686e-06,
+      "loss": 0.1582,
       "step": 170
     },
     {
-      "epoch": 0.01152,
-      "grad_norm": 3.688382387161255,
-      "learning_rate": 1.9093333333333335e-06,
-      "loss": 0.1396,
+      "epoch": 0.011011868346996207,
+      "grad_norm": 3.3082435131073,
+      "learning_rate": 1.8255991840897503e-06,
+      "loss": 0.1502,
       "step": 180
     },
     {
-      "epoch": 0.01216,
-      "grad_norm": 1.8855538368225098,
-      "learning_rate": 2.0160000000000003e-06,
-      "loss": 0.1476,
+      "epoch": 0.011623638810718219,
+      "grad_norm": 3.446185350418091,
+      "learning_rate": 1.9275879653238146e-06,
+      "loss": 0.1452,
       "step": 190
     },
     {
-      "epoch": 0.0128,
-      "grad_norm": 2.2272040843963623,
-      "learning_rate": 2.1226666666666666e-06,
-      "loss": 0.144,
+      "epoch": 0.01223540927444023,
+      "grad_norm": 2.6021993160247803,
+      "learning_rate": 2.029576746557879e-06,
+      "loss": 0.1404,
       "step": 200
     },
     {
-      "epoch": 0.01344,
-      "grad_norm": 3.505934476852417,
-      "learning_rate": 2.229333333333334e-06,
-      "loss": 0.1306,
+      "epoch": 0.012847179738162242,
+      "grad_norm": 2.1967263221740723,
+      "learning_rate": 2.1315655277919432e-06,
+      "loss": 0.1421,
       "step": 210
     },
     {
-      "epoch": 0.01408,
-      "grad_norm": 1.8632491827011108,
-      "learning_rate": 2.336e-06,
-      "loss": 0.1504,
+      "epoch": 0.013458950201884253,
+      "grad_norm": 1.9831513166427612,
+      "learning_rate": 2.233554309026007e-06,
+      "loss": 0.1431,
       "step": 220
     },
     {
-      "epoch": 0.01472,
-      "grad_norm": 2.191713333129883,
-      "learning_rate": 2.442666666666667e-06,
-      "loss": 0.1346,
+      "epoch": 0.014070720665606265,
+      "grad_norm": 1.7445921897888184,
+      "learning_rate": 2.3355430902600715e-06,
+      "loss": 0.1242,
       "step": 230
     },
     {
-      "epoch": 0.01536,
-      "grad_norm": 1.9042261838912964,
-      "learning_rate": 2.5493333333333337e-06,
-      "loss": 0.135,
+      "epoch": 0.014682491129328276,
+      "grad_norm": 2.7725071907043457,
+      "learning_rate": 2.4375318714941358e-06,
+      "loss": 0.1346,
       "step": 240
     },
     {
-      "epoch": 0.016,
-      "grad_norm": 2.266148090362549,
-      "learning_rate": 2.656e-06,
-      "loss": 0.1356,
+      "epoch": 0.015294261593050288,
+      "grad_norm": 1.7258802652359009,
+      "learning_rate": 2.5395206527282e-06,
+      "loss": 0.1371,
       "step": 250
     },
     {
-      "epoch": 0.01664,
-      "grad_norm": 2.4776418209075928,
-      "learning_rate": 2.762666666666667e-06,
-      "loss": 0.129,
+      "epoch": 0.0159060320567723,
+      "grad_norm": 2.1158506870269775,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.1271,
       "step": 260
     },
     {
-      "epoch": 0.01728,
-      "grad_norm": 2.340693712234497,
-      "learning_rate": 2.8693333333333336e-06,
-      "loss": 0.1236,
+      "epoch": 0.01651780252049431,
+      "grad_norm": 2.0771985054016113,
+      "learning_rate": 2.7434982151963283e-06,
+      "loss": 0.1214,
       "step": 270
     },
     {
-      "epoch": 0.01792,
-      "grad_norm": 1.6849262714385986,
-      "learning_rate": 2.976e-06,
-      "loss": 0.1152,
+      "epoch": 0.017129572984216322,
+      "grad_norm": 2.924894332885742,
+      "learning_rate": 2.8454869964303926e-06,
+      "loss": 0.1281,
       "step": 280
     },
     {
-      "epoch": 0.01856,
-      "grad_norm": 3.5535154342651367,
-      "learning_rate": 3.082666666666667e-06,
-      "loss": 0.1335,
+      "epoch": 0.017741343447938332,
+      "grad_norm": 2.122687578201294,
+      "learning_rate": 2.947475777664457e-06,
+      "loss": 0.132,
       "step": 290
     },
     {
-      "epoch": 0.0192,
-      "grad_norm": 2.9307031631469727,
-      "learning_rate": 3.1893333333333335e-06,
-      "loss": 0.1136,
+      "epoch": 0.018353113911660345,
+      "grad_norm": 3.2565481662750244,
+      "learning_rate": 3.0494645588985217e-06,
+      "loss": 0.1336,
       "step": 300
     },
     {
-      "epoch": 0.01984,
-      "grad_norm": 3.0100226402282715,
-      "learning_rate": 3.2960000000000003e-06,
-      "loss": 0.1186,
+      "epoch": 0.018964884375382355,
+      "grad_norm": 1.2805263996124268,
+      "learning_rate": 3.151453340132586e-06,
+      "loss": 0.1183,
       "step": 310
     },
     {
-      "epoch": 0.02048,
-      "grad_norm": 2.088911771774292,
-      "learning_rate": 3.402666666666667e-06,
-      "loss": 0.1051,
+      "epoch": 0.019576654839104368,
+      "grad_norm": 2.420628786087036,
+      "learning_rate": 3.25344212136665e-06,
+      "loss": 0.1125,
       "step": 320
     },
     {
-      "epoch": 0.02112,
-      "grad_norm": 4.332556247711182,
-      "learning_rate": 3.5093333333333334e-06,
-      "loss": 0.1075,
+      "epoch": 0.020188425302826378,
+      "grad_norm": 2.887117862701416,
+      "learning_rate": 3.3554309026007142e-06,
+      "loss": 0.1041,
       "step": 330
     },
     {
-      "epoch": 0.02176,
-      "grad_norm": 2.954036235809326,
-      "learning_rate": 3.616e-06,
-      "loss": 0.11,
+      "epoch": 0.02080019576654839,
+      "grad_norm": 1.6021567583084106,
+      "learning_rate": 3.4574196838347786e-06,
+      "loss": 0.1221,
       "step": 340
     },
     {
-      "epoch": 0.0224,
-      "grad_norm": 2.262308359146118,
-      "learning_rate": 3.7226666666666674e-06,
-      "loss": 0.1057,
+      "epoch": 0.0214119662302704,
+      "grad_norm": 2.9132304191589355,
+      "learning_rate": 3.5594084650688425e-06,
+      "loss": 0.1094,
       "step": 350
     },
     {
-      "epoch": 0.02304,
-      "grad_norm": 1.5145989656448364,
-      "learning_rate": 3.829333333333334e-06,
-      "loss": 0.1049,
+      "epoch": 0.022023736693992414,
+      "grad_norm": 2.5460691452026367,
+      "learning_rate": 3.6613972463029068e-06,
+      "loss": 0.1096,
       "step": 360
     },
     {
-      "epoch": 0.02368,
-      "grad_norm": 1.740663766860962,
-      "learning_rate": 3.936e-06,
-      "loss": 0.0964,
+      "epoch": 0.022635507157714424,
+      "grad_norm": 2.2170896530151367,
+      "learning_rate": 3.763386027536971e-06,
+      "loss": 0.0933,
       "step": 370
     },
     {
-      "epoch": 0.02432,
-      "grad_norm": 1.7438851594924927,
-      "learning_rate": 4.042666666666667e-06,
-      "loss": 0.0936,
+      "epoch": 0.023247277621436437,
+      "grad_norm": 3.799591541290283,
+      "learning_rate": 3.865374808771036e-06,
+      "loss": 0.1142,
       "step": 380
     },
     {
-      "epoch": 0.02496,
-      "grad_norm": 3.946115493774414,
-      "learning_rate": 4.149333333333334e-06,
-      "loss": 0.1038,
+      "epoch": 0.023859048085158447,
+      "grad_norm": 4.609161376953125,
+      "learning_rate": 3.9673635900051e-06,
+      "loss": 0.0948,
       "step": 390
     },
     {
-      "epoch": 0.0256,
-      "grad_norm": 5.289428234100342,
-      "learning_rate": 4.256e-06,
-      "loss": 0.1015,
+      "epoch": 0.02447081854888046,
+      "grad_norm": 3.329650402069092,
+      "learning_rate": 4.0693523712391645e-06,
+      "loss": 0.0944,
       "step": 400
     },
     {
-      "epoch": 0.02624,
-      "grad_norm": 1.9344863891601562,
-      "learning_rate": 4.362666666666667e-06,
-      "loss": 0.0855,
+      "epoch": 0.02508258901260247,
+      "grad_norm": 2.135805368423462,
+      "learning_rate": 4.171341152473228e-06,
+      "loss": 0.0847,
       "step": 410
     },
     {
-      "epoch": 0.02688,
-      "grad_norm": 7.482412338256836,
-      "learning_rate": 4.469333333333334e-06,
-      "loss": 0.1031,
+      "epoch": 0.025694359476324483,
+      "grad_norm": 2.006049156188965,
+      "learning_rate": 4.273329933707292e-06,
+      "loss": 0.1077,
       "step": 420
     },
     {
-      "epoch": 0.02752,
-      "grad_norm": 2.5646328926086426,
-      "learning_rate": 4.576000000000001e-06,
-      "loss": 0.0777,
+      "epoch": 0.026306129940046493,
+      "grad_norm": 5.709030628204346,
+      "learning_rate": 4.375318714941357e-06,
+      "loss": 0.099,
       "step": 430
     },
     {
-      "epoch": 0.02816,
-      "grad_norm": 1.473496437072754,
-      "learning_rate": 4.682666666666667e-06,
-      "loss": 0.0854,
+      "epoch": 0.026917900403768506,
+      "grad_norm": 2.466867446899414,
+      "learning_rate": 4.477307496175421e-06,
+      "loss": 0.0771,
       "step": 440
     },
     {
-      "epoch": 0.0288,
-      "grad_norm": 2.204341411590576,
-      "learning_rate": 4.7893333333333334e-06,
-      "loss": 0.0739,
+      "epoch": 0.027529670867490516,
+      "grad_norm": 2.3829548358917236,
+      "learning_rate": 4.579296277409485e-06,
+      "loss": 0.0802,
       "step": 450
     },
     {
-      "epoch": 0.02944,
-      "grad_norm": 3.1363987922668457,
-      "learning_rate": 4.896e-06,
-      "loss": 0.0886,
+      "epoch": 0.02814144133121253,
+      "grad_norm": 2.990241050720215,
+      "learning_rate": 4.6812850586435496e-06,
+      "loss": 0.1048,
       "step": 460
     },
     {
-      "epoch": 0.03008,
-      "grad_norm": 2.0473029613494873,
-      "learning_rate": 5.002666666666667e-06,
-      "loss": 0.0794,
+      "epoch": 0.02875321179493454,
+      "grad_norm": 3.7126529216766357,
+      "learning_rate": 4.783273839877614e-06,
+      "loss": 0.0879,
       "step": 470
     },
     {
-      "epoch": 0.03072,
-      "grad_norm": 1.6088207960128784,
-      "learning_rate": 5.109333333333334e-06,
-      "loss": 0.0604,
+      "epoch": 0.029364982258656552,
+      "grad_norm": 2.6238865852355957,
+      "learning_rate": 4.885262621111678e-06,
+      "loss": 0.0825,
       "step": 480
     },
     {
-      "epoch": 0.03136,
-      "grad_norm": 2.974717617034912,
-      "learning_rate": 5.216e-06,
-      "loss": 0.0827,
+      "epoch": 0.029976752722378562,
+      "grad_norm": 2.1882681846618652,
+      "learning_rate": 4.987251402345742e-06,
+      "loss": 0.0821,
       "step": 490
     },
     {
-      "epoch": 0.032,
-      "grad_norm": 6.114630699157715,
-      "learning_rate": 5.322666666666667e-06,
-      "loss": 0.0631,
+      "epoch": 0.030588523186100575,
+      "grad_norm": 2.1399929523468018,
+      "learning_rate": 5.089240183579807e-06,
+      "loss": 0.0926,
       "step": 500
     },
     {
-      "epoch": 0.032,
-      "eval_accuracy": 0.88603,
-      "eval_f1_weighted": 0.8857328199768308,
-      "eval_loss": 0.08366190642118454,
-      "eval_macro_f1": 0.8859061279979327,
-      "eval_precision": 0.8966428429175579,
-      "eval_recall": 0.88603,
-      "eval_runtime": 1629.6546,
-      "eval_samples_per_second": 61.363,
-      "eval_steps_per_second": 0.959,
+      "epoch": 0.030588523186100575,
+      "eval_accuracy": 0.8732,
+      "eval_f1_weighted": 0.8728301347090768,
+      "eval_loss": 0.08081420511007309,
+      "eval_macro_f1": 0.8730387325792941,
+      "eval_precision": 0.8842823474890878,
+      "eval_recall": 0.8732,
+      "eval_runtime": 1731.362,
+      "eval_samples_per_second": 57.758,
+      "eval_steps_per_second": 0.903,
       "step": 500
     },
     {
-      "epoch": 0.03264,
-      "grad_norm": 1.4639947414398193,
-      "learning_rate": 5.429333333333334e-06,
-      "loss": 0.0724,
+      "epoch": 0.031200293649822585,
+      "grad_norm": 4.9644694328308105,
+      "learning_rate": 5.191228964813872e-06,
+      "loss": 0.0846,
       "step": 510
     },
     {
-      "epoch": 0.03328,
-      "grad_norm": 4.234114646911621,
-      "learning_rate": 5.536e-06,
-      "loss": 0.0737,
+      "epoch": 0.0318120641135446,
+      "grad_norm": 3.2999346256256104,
+      "learning_rate": 5.293217746047935e-06,
+      "loss": 0.0774,
       "step": 520
     },
     {
-      "epoch": 0.03392,
-      "grad_norm": 1.9391822814941406,
-      "learning_rate": 5.642666666666668e-06,
-      "loss": 0.0875,
+      "epoch": 0.03242383457726661,
+      "grad_norm": 1.9823625087738037,
+      "learning_rate": 5.395206527281999e-06,
+      "loss": 0.0781,
       "step": 530
     },
     {
-      "epoch": 0.03456,
-      "grad_norm": 2.0388095378875732,
-      "learning_rate": 5.7493333333333336e-06,
-      "loss": 0.0661,
+      "epoch": 0.03303560504098862,
+      "grad_norm": 2.010012149810791,
+      "learning_rate": 5.497195308516063e-06,
+      "loss": 0.0748,
       "step": 540
     },
     {
-      "epoch": 0.0352,
-      "grad_norm": 1.9697074890136719,
-      "learning_rate": 5.856e-06,
-      "loss": 0.0647,
+      "epoch": 0.033647375504710635,
+      "grad_norm": 1.7753310203552246,
+      "learning_rate": 5.599184089750128e-06,
+      "loss": 0.0687,
       "step": 550
     },
     {
-      "epoch": 0.03584,
-      "grad_norm": 3.9906229972839355,
-      "learning_rate": 5.962666666666666e-06,
-      "loss": 0.0696,
+      "epoch": 0.034259145968432644,
+      "grad_norm": 1.4343310594558716,
+      "learning_rate": 5.701172870984192e-06,
+      "loss": 0.0729,
       "step": 560
     },
     {
-      "epoch": 0.03648,
-      "grad_norm": 1.5516153573989868,
-      "learning_rate": 6.069333333333334e-06,
-      "loss": 0.0576,
+      "epoch": 0.034870916432154654,
+      "grad_norm": 2.562145948410034,
+      "learning_rate": 5.803161652218257e-06,
+      "loss": 0.0712,
       "step": 570
     },
     {
-      "epoch": 0.03712,
-      "grad_norm": 3.5216825008392334,
-      "learning_rate": 6.176000000000001e-06,
-      "loss": 0.0793,
+      "epoch": 0.035482686895876664,
+      "grad_norm": 1.727373480796814,
+      "learning_rate": 5.9051504334523206e-06,
+      "loss": 0.0605,
       "step": 580
     },
     {
-      "epoch": 0.03776,
-      "grad_norm": 6.22791051864624,
-      "learning_rate": 6.282666666666667e-06,
-      "loss": 0.0757,
+      "epoch": 0.03609445735959868,
+      "grad_norm": 1.8711422681808472,
+      "learning_rate": 6.007139214686385e-06,
+      "loss": 0.071,
       "step": 590
     },
     {
-      "epoch": 0.0384,
-      "grad_norm": 1.8255369663238525,
-      "learning_rate": 6.389333333333334e-06,
-      "loss": 0.0602,
+      "epoch": 0.03670622782332069,
+      "grad_norm": 7.535566329956055,
+      "learning_rate": 6.10912799592045e-06,
+      "loss": 0.067,
       "step": 600
     },
     {
-      "epoch": 0.03904,
-      "grad_norm": 2.506188154220581,
-      "learning_rate": 6.496e-06,
-      "loss": 0.0643,
+      "epoch": 0.0373179982870427,
+      "grad_norm": 3.9871439933776855,
+      "learning_rate": 6.211116777154513e-06,
+      "loss": 0.063,
       "step": 610
     },
     {
-      "epoch": 0.03968,
-      "grad_norm": 1.5877528190612793,
-      "learning_rate": 6.602666666666667e-06,
-      "loss": 0.0501,
+      "epoch": 0.03792976875076471,
+      "grad_norm": 2.457789659500122,
+      "learning_rate": 6.313105558388578e-06,
+      "loss": 0.0905,
       "step": 620
     },
     {
-      "epoch": 0.04032,
-      "grad_norm": 2.1468632221221924,
-      "learning_rate": 6.7093333333333345e-06,
-      "loss": 0.0521,
+      "epoch": 0.03854153921448673,
+      "grad_norm": 4.795208930969238,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.0611,
       "step": 630
     },
     {
-      "epoch": 0.04096,
-      "grad_norm": 2.6225597858428955,
-      "learning_rate": 6.8160000000000005e-06,
-      "loss": 0.068,
+      "epoch": 0.039153309678208736,
+      "grad_norm": 3.688422441482544,
+      "learning_rate": 6.5170831208567065e-06,
+      "loss": 0.0657,
       "step": 640
     },
     {
-      "epoch": 0.0416,
-      "grad_norm": 3.1171772480010986,
-      "learning_rate": 6.922666666666667e-06,
-      "loss": 0.0644,
+      "epoch": 0.039765080141930746,
+      "grad_norm": 4.9787116050720215,
+      "learning_rate": 6.61907190209077e-06,
+      "loss": 0.0593,
       "step": 650
     },
     {
-      "epoch": 0.04224,
-      "grad_norm": 2.7589240074157715,
-      "learning_rate": 7.029333333333333e-06,
-      "loss": 0.0763,
+      "epoch": 0.040376850605652756,
+      "grad_norm": 6.09706974029541,
+      "learning_rate": 6.721060683324835e-06,
+      "loss": 0.0803,
       "step": 660
     },
     {
-      "epoch": 0.04288,
-      "grad_norm": 2.3155124187469482,
-      "learning_rate": 7.136000000000001e-06,
-      "loss": 0.0451,
+      "epoch": 0.04098862106937477,
+      "grad_norm": 1.9047552347183228,
+      "learning_rate": 6.8230494645589e-06,
+      "loss": 0.0486,
       "step": 670
     },
     {
-      "epoch": 0.04352,
-      "grad_norm": 0.9795209169387817,
-      "learning_rate": 7.242666666666667e-06,
-      "loss": 0.0552,
+      "epoch": 0.04160039153309678,
+      "grad_norm": 3.514389753341675,
+      "learning_rate": 6.925038245792963e-06,
+      "loss": 0.058,
       "step": 680
     },
     {
-      "epoch": 0.04416,
-      "grad_norm": 6.224135398864746,
-      "learning_rate": 7.3493333333333335e-06,
-      "loss": 0.0614,
+      "epoch": 0.04221216199681879,
+      "grad_norm": 3.366957187652588,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.0651,
       "step": 690
     },
     {
-      "epoch": 0.0448,
-      "grad_norm": 1.9188823699951172,
-      "learning_rate": 7.456000000000001e-06,
-      "loss": 0.0539,
+      "epoch": 0.0428239324605408,
+      "grad_norm": 4.780582904815674,
+      "learning_rate": 7.1290158082610916e-06,
+      "loss": 0.0561,
       "step": 700
     },
     {
-      "epoch": 0.04544,
-      "grad_norm": 2.0834364891052246,
-      "learning_rate": 7.562666666666667e-06,
-      "loss": 0.0545,
+      "epoch": 0.04343570292426282,
+      "grad_norm": 2.457082748413086,
+      "learning_rate": 7.231004589495156e-06,
+      "loss": 0.0611,
       "step": 710
     },
     {
-      "epoch": 0.04608,
-      "grad_norm": 2.081609010696411,
-      "learning_rate": 7.669333333333335e-06,
-      "loss": 0.0563,
+      "epoch": 0.04404747338798483,
+      "grad_norm": 1.283014178276062,
+      "learning_rate": 7.33299337072922e-06,
+      "loss": 0.0601,
       "step": 720
     },
     {
-      "epoch": 0.04672,
-      "grad_norm": 1.461228370666504,
-      "learning_rate": 7.776e-06,
-      "loss": 0.0495,
+      "epoch": 0.04465924385170684,
+      "grad_norm": 3.092107057571411,
+      "learning_rate": 7.434982151963285e-06,
+      "loss": 0.0598,
       "step": 730
     },
     {
-      "epoch": 0.04736,
-      "grad_norm": 2.7746951580047607,
-      "learning_rate": 7.882666666666667e-06,
-      "loss": 0.0517,
+      "epoch": 0.04527101431542885,
+      "grad_norm": 1.26460599899292,
+      "learning_rate": 7.536970933197348e-06,
+      "loss": 0.0503,
       "step": 740
     },
     {
-      "epoch": 0.048,
-      "grad_norm": 1.2801299095153809,
-      "learning_rate": 7.989333333333335e-06,
-      "loss": 0.034,
+      "epoch": 0.045882784779150865,
+      "grad_norm": 2.9157862663269043,
+      "learning_rate": 7.638959714431414e-06,
+      "loss": 0.0507,
       "step": 750
     },
     {
-      "epoch": 0.04864,
-      "grad_norm": 3.921491861343384,
-      "learning_rate": 8.096e-06,
-      "loss": 0.0537,
+      "epoch": 0.046494555242872875,
+      "grad_norm": 5.487971782684326,
+      "learning_rate": 7.740948495665478e-06,
+      "loss": 0.0525,
       "step": 760
     },
     {
-      "epoch": 0.04928,
-      "grad_norm": 1.9986109733581543,
-      "learning_rate": 8.202666666666667e-06,
-      "loss": 0.0444,
+      "epoch": 0.047106325706594884,
+      "grad_norm": 2.9894769191741943,
+      "learning_rate": 7.842937276899541e-06,
+      "loss": 0.0656,
       "step": 770
     },
     {
-      "epoch": 0.04992,
-      "grad_norm": 4.678001403808594,
-      "learning_rate": 8.309333333333334e-06,
-      "loss": 0.0589,
+      "epoch": 0.047718096170316894,
+      "grad_norm": 2.7786455154418945,
+      "learning_rate": 7.944926058133606e-06,
+      "loss": 0.0521,
       "step": 780
     },
     {
-      "epoch": 0.05056,
-      "grad_norm": 1.7302327156066895,
-      "learning_rate": 8.416e-06,
-      "loss": 0.043,
+      "epoch": 0.04832986663403891,
+      "grad_norm": 1.3629153966903687,
+      "learning_rate": 8.04691483936767e-06,
+      "loss": 0.0529,
       "step": 790
     },
     {
-      "epoch": 0.0512,
-      "grad_norm": 2.5643601417541504,
-      "learning_rate": 8.522666666666667e-06,
-      "loss": 0.0671,
+      "epoch": 0.04894163709776092,
+      "grad_norm": 4.066152572631836,
+      "learning_rate": 8.148903620601734e-06,
+      "loss": 0.047,
       "step": 800
     },
     {
-      "epoch": 0.05184,
-      "grad_norm": 2.8869082927703857,
-      "learning_rate": 8.629333333333334e-06,
-      "loss": 0.0565,
+      "epoch": 0.04955340756148293,
+      "grad_norm": 2.162338972091675,
+      "learning_rate": 8.250892401835799e-06,
+      "loss": 0.0583,
       "step": 810
     },
     {
-      "epoch": 0.05248,
-      "grad_norm": 3.723088264465332,
-      "learning_rate": 8.736e-06,
-      "loss": 0.0428,
+      "epoch": 0.05016517802520494,
+      "grad_norm": 1.5794885158538818,
+      "learning_rate": 8.352881183069863e-06,
+      "loss": 0.0415,
       "step": 820
     },
     {
-      "epoch": 0.05312,
-      "grad_norm": 4.387404441833496,
-      "learning_rate": 8.842666666666667e-06,
-      "loss": 0.0559,
+      "epoch": 0.05077694848892696,
+      "grad_norm": 5.697793960571289,
+      "learning_rate": 8.454869964303928e-06,
+      "loss": 0.0614,
       "step": 830
     },
     {
-      "epoch": 0.05376,
-      "grad_norm": 1.889413833618164,
-      "learning_rate": 8.949333333333334e-06,
-      "loss": 0.0405,
+      "epoch": 0.05138871895264897,
+      "grad_norm": 2.243929386138916,
+      "learning_rate": 8.556858745537991e-06,
+      "loss": 0.0836,
       "step": 840
     },
     {
-      "epoch": 0.0544,
-      "grad_norm": 1.4773402214050293,
-      "learning_rate": 9.056000000000001e-06,
-      "loss": 0.0414,
+      "epoch": 0.052000489416370976,
+      "grad_norm": 1.272515892982483,
+      "learning_rate": 8.658847526772056e-06,
+      "loss": 0.05,
       "step": 850
     },
     {
-      "epoch": 0.05504,
-      "grad_norm": 0.4495583176612854,
-      "learning_rate": 9.162666666666668e-06,
-      "loss": 0.0446,
+      "epoch": 0.052612259880092986,
+      "grad_norm": 1.4770886898040771,
+      "learning_rate": 8.760836308006119e-06,
+      "loss": 0.0403,
       "step": 860
     },
     {
-      "epoch": 0.05568,
-      "grad_norm": 1.158181071281433,
-      "learning_rate": 9.269333333333335e-06,
-      "loss": 0.0432,
+      "epoch": 0.053224030343815,
+      "grad_norm": 3.411848306655884,
+      "learning_rate": 8.862825089240184e-06,
+      "loss": 0.0478,
       "step": 870
     },
     {
-      "epoch": 0.05632,
-      "grad_norm": 2.434063196182251,
-      "learning_rate": 9.376000000000001e-06,
-      "loss": 0.0502,
+      "epoch": 0.05383580080753701,
+      "grad_norm": 8.478028297424316,
+      "learning_rate": 8.964813870474248e-06,
+      "loss": 0.0855,
       "step": 880
     },
     {
-      "epoch": 0.05696,
-      "grad_norm": 1.2009385824203491,
-      "learning_rate": 9.482666666666668e-06,
-      "loss": 0.0473,
+      "epoch": 0.05444757127125902,
+      "grad_norm": 0.6067395806312561,
+      "learning_rate": 9.066802651708313e-06,
+      "loss": 0.0366,
       "step": 890
     },
     {
-      "epoch": 0.0576,
-      "grad_norm": 3.9361023902893066,
-      "learning_rate": 9.589333333333333e-06,
-      "loss": 0.046,
+      "epoch": 0.05505934173498103,
+      "grad_norm": 2.6501152515411377,
+      "learning_rate": 9.168791432942376e-06,
+      "loss": 0.0358,
       "step": 900
     },
     {
-      "epoch": 0.05824,
-      "grad_norm": 2.926988124847412,
-      "learning_rate": 9.696000000000002e-06,
-      "loss": 0.0464,
+      "epoch": 0.05567111219870305,
+      "grad_norm": 1.1650569438934326,
+      "learning_rate": 9.270780214176441e-06,
+      "loss": 0.0733,
       "step": 910
     },
     {
-      "epoch": 0.05888,
-      "grad_norm": 2.215726375579834,
-      "learning_rate": 9.802666666666667e-06,
-      "loss": 0.0471,
+      "epoch": 0.05628288266242506,
+      "grad_norm": 2.4513559341430664,
+      "learning_rate": 9.372768995410506e-06,
+      "loss": 0.0535,
       "step": 920
     },
     {
-      "epoch": 0.05952,
-      "grad_norm": 2.4104645252227783,
-      "learning_rate": 9.909333333333334e-06,
-      "loss": 0.0482,
+      "epoch": 0.05689465312614707,
+      "grad_norm": 0.8360533714294434,
+      "learning_rate": 9.47475777664457e-06,
+      "loss": 0.0453,
       "step": 930
     },
     {
-      "epoch": 0.06016,
-      "grad_norm": 5.664385795593262,
-      "learning_rate": 1.0016000000000002e-05,
-      "loss": 0.044,
+      "epoch": 0.05750642358986908,
+      "grad_norm": 2.2937448024749756,
+      "learning_rate": 9.576746557878635e-06,
+      "loss": 0.0655,
       "step": 940
     },
     {
-      "epoch": 0.0608,
-      "grad_norm": 1.7977851629257202,
-      "learning_rate": 1.0122666666666667e-05,
-      "loss": 0.0429,
+      "epoch": 0.058118194053591095,
+      "grad_norm": 2.025761127471924,
+      "learning_rate": 9.678735339112698e-06,
+      "loss": 0.06,
       "step": 950
     },
     {
-      "epoch": 0.06144,
-      "grad_norm": 2.953789472579956,
-      "learning_rate": 1.0229333333333334e-05,
-      "loss": 0.0533,
+      "epoch": 0.058729964517313105,
+      "grad_norm": 2.6468706130981445,
+      "learning_rate": 9.780724120346763e-06,
+      "loss": 0.0416,
       "step": 960
     },
     {
-      "epoch": 0.06208,
-      "grad_norm": 1.6272608041763306,
-      "learning_rate": 1.0336000000000002e-05,
-      "loss": 0.0307,
+      "epoch": 0.059341734981035114,
+      "grad_norm": 3.626215934753418,
+      "learning_rate": 9.882712901580826e-06,
+      "loss": 0.0239,
       "step": 970
     },
     {
-      "epoch": 0.06272,
-      "grad_norm": 2.8621678352355957,
-      "learning_rate": 1.0442666666666667e-05,
-      "loss": 0.0461,
+      "epoch": 0.059953505444757124,
+      "grad_norm": 2.599356174468994,
+      "learning_rate": 9.984701682814891e-06,
+      "loss": 0.0395,
       "step": 980
     },
     {
-      "epoch": 0.06336,
-      "grad_norm": 3.70019793510437,
-      "learning_rate": 1.0549333333333334e-05,
-      "loss": 0.0546,
+      "epoch": 0.06056527590847914,
+      "grad_norm": 2.4896769523620605,
+      "learning_rate": 1.0086690464048956e-05,
+      "loss": 0.0703,
       "step": 990
     },
     {
-      "epoch": 0.064,
-      "grad_norm": 1.1777608394622803,
-      "learning_rate": 1.0656000000000003e-05,
-      "loss": 0.0508,
+      "epoch": 0.06117704637220115,
+      "grad_norm": 2.18449068069458,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 0.0306,
       "step": 1000
     },
     {
-      "epoch": 0.064,
-      "eval_accuracy": 0.96139,
-      "eval_f1_weighted": 0.9613900534763816,
-      "eval_loss": 0.031560152769088745,
-      "eval_macro_f1": 0.9613078781032101,
-      "eval_precision": 0.9613901086866451,
-      "eval_recall": 0.96139,
-      "eval_runtime": 1630.3285,
-      "eval_samples_per_second": 61.337,
-      "eval_steps_per_second": 0.959,
+      "epoch": 0.06117704637220115,
+      "eval_accuracy": 0.95488,
+      "eval_f1_weighted": 0.9549029721025599,
+      "eval_loss": 0.03926030173897743,
+      "eval_macro_f1": 0.9548451024890077,
+      "eval_precision": 0.955565530093091,
+      "eval_recall": 0.95488,
+      "eval_runtime": 1718.0831,
+      "eval_samples_per_second": 58.204,
+      "eval_steps_per_second": 0.91,
       "step": 1000
     }
   ],
   "logging_steps": 10,
-  "max_steps": 15625,
+  "max_steps": 16346,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
   "save_steps": 500,
@@ -762,7 +762,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 8415459105720960.0,
+  "total_flos": 8415771550099200.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null

checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0256e15114b0df6f165bb3e6d75f279805d6f125574ac927fca7a602f9449f4a
+oid sha256:4a51c9ce1080123075e6ab993dc495e2fe9920836854bf1888707e1cdafdc082
 size 5777