Training in progress, epoch 9, checkpoint

checkpoint-729/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e4cb71a392cefead6f7be4c24d9b5751ecd3091b01406ccab27107182124ce31
+oid sha256:b941de8ae607041533aa196f81b72b6f39dc290b5345e451132ff1cb1e2159c9
 size 592318676

checkpoint-729/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:484cf65f309c3af6c48483c760039d3010b2f0dfcee4e73bffd25a15a77a4588
+oid sha256:5e131d3c2298937f311923fdd5335558cdc5b55661710b215ea8c787ba81f096
 size 1014645498

checkpoint-729/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:33e201bb4bf305b6dcac90ce17e6d2f74219c833c6d3bc2a1eb84fd26a97a92a
+oid sha256:0968d0d979e679458f54b3ce479d877ab7f754e313e39c8158ce8426e6dbe221
 size 14244

checkpoint-729/trainer_state.json

@@ -11,358 +11,358 @@
     {
       "epoch": 1.0,
       "eval_B": {
-        "f1-score": 0.7905866302864938,
-        "precision": 0.7768096514745308,
-        "recall": 0.8048611111111111,
-        "support": 1440.0
+        "f1-score": 0.8331822302810515,
+        "precision": 0.7901977644024075,
+        "recall": 0.8811121764141898,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9005331227169513,
-        "precision": 0.9637593111099372,
-        "recall": 0.8450919534905267,
-        "support": 21587.0
+        "f1-score": 0.9471294753175105,
+        "precision": 0.9443616777446711,
+        "recall": 0.9499135446685879,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.8288892663043479,
-        "precision": 0.7463108800367001,
-        "recall": 0.9320156593144275,
-        "support": 10473.0
+        "f1-score": 0.8992868897421833,
+        "precision": 0.9103731674811195,
+        "recall": 0.8884673748103187,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.8705373134328358,
-      "eval_loss": 0.32930415868759155,
+      "eval_accuracy": 0.9267895289474637,
+      "eval_loss": 0.20478281378746033,
       "eval_macro avg": {
-        "f1-score": 0.8400030064359308,
-        "precision": 0.8289599475403894,
-        "recall": 0.8606562413053552,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7349,
-      "eval_samples_per_second": 17.107,
-      "eval_steps_per_second": 2.323,
+        "f1-score": 0.8931995317802484,
+        "precision": 0.8816442032093993,
+        "recall": 0.9064976986310321,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6128,
+      "eval_samples_per_second": 17.343,
+      "eval_steps_per_second": 2.168,
       "eval_weighted avg": {
-        "f1-score": 0.8734092702599646,
-        "precision": 0.8877430445874,
-        "recall": 0.8705373134328358,
-        "support": 33500.0
+        "f1-score": 0.9268447919078652,
+        "precision": 0.9271861479533134,
+        "recall": 0.9267895289474637,
+        "support": 27619.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_B": {
-        "f1-score": 0.8656218739579861,
-        "precision": 0.8325849903784477,
-        "recall": 0.9013888888888889,
-        "support": 1440.0
+        "f1-score": 0.8682547399124939,
+        "precision": 0.8806706114398422,
+        "recall": 0.8561840843720039,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9547168948727912,
-        "precision": 0.9423736462093862,
-        "recall": 0.9673877796822161,
-        "support": 21587.0
+        "f1-score": 0.9590983771942466,
+        "precision": 0.9592919333448654,
+        "recall": 0.9589048991354466,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.9013528191962081,
-        "precision": 0.9332379102341274,
-        "recall": 0.8715745249689678,
-        "support": 10473.0
+        "f1-score": 0.9226525313717006,
+        "precision": 0.9208594256100194,
+        "recall": 0.9244526338608281,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9345970149253732,
-      "eval_loss": 0.2162298709154129,
+      "eval_accuracy": 0.9435171439950758,
+      "eval_loss": 0.16574698686599731,
       "eval_macro avg": {
-        "f1-score": 0.9072305293423284,
-        "precision": 0.902732182273987,
-        "recall": 0.9134503978466909,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.753,
-      "eval_samples_per_second": 17.042,
-      "eval_steps_per_second": 2.314,
+        "f1-score": 0.9166685494928136,
+        "precision": 0.9202739901315756,
+        "recall": 0.9131805391227594,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6238,
+      "eval_samples_per_second": 17.302,
+      "eval_steps_per_second": 2.163,
       "eval_weighted avg": {
-        "f1-score": 0.9342040950316517,
-        "precision": 0.9347982961417612,
-        "recall": 0.9345970149253732,
-        "support": 33500.0
+        "f1-score": 0.9434932036816766,
+        "precision": 0.9434846863370583,
+        "recall": 0.9435171439950758,
+        "support": 27619.0
       },
       "step": 162
     },
     {
       "epoch": 3.0,
       "eval_B": {
-        "f1-score": 0.8603587759409076,
-        "precision": 0.8717034925160371,
-        "recall": 0.8493055555555555,
-        "support": 1440.0
+        "f1-score": 0.8791725434884815,
+        "precision": 0.8625461254612546,
+        "recall": 0.8964525407478428,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9434195218082959,
-        "precision": 0.9677057963955188,
-        "recall": 0.9203224162690509,
-        "support": 21587.0
+        "f1-score": 0.9588372159825352,
+        "precision": 0.9557349825345015,
+        "recall": 0.9619596541786744,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.8879310344827586,
-        "precision": 0.8459410391631365,
-        "recall": 0.9343072663038289,
-        "support": 10473.0
+        "f1-score": 0.9204284621270084,
+        "precision": 0.9282407407407407,
+        "recall": 0.9127465857359636,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9216417910447762,
-      "eval_loss": 0.24082203209400177,
+      "eval_accuracy": 0.9430464535283681,
+      "eval_loss": 0.15394224226474762,
       "eval_macro avg": {
-        "f1-score": 0.8972364440773206,
-        "precision": 0.8951167760248975,
-        "recall": 0.9013117460428117,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7554,
-      "eval_samples_per_second": 17.033,
-      "eval_steps_per_second": 2.313,
+        "f1-score": 0.9194794071993417,
+        "precision": 0.9155072829121655,
+        "recall": 0.9237195935541602,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6233,
+      "eval_samples_per_second": 17.303,
+      "eval_steps_per_second": 2.163,
       "eval_weighted avg": {
-        "f1-score": 0.9225019575751797,
-        "precision": 0.9255121957960801,
-        "recall": 0.9216417910447762,
-        "support": 33500.0
+        "f1-score": 0.9429985029052194,
+        "precision": 0.9430314866542512,
+        "recall": 0.9430464535283681,
+        "support": 27619.0
       },
       "step": 243
     },
     {
       "epoch": 4.0,
       "eval_B": {
-        "f1-score": 0.8804020100502512,
-        "precision": 0.8504854368932039,
-        "recall": 0.9125,
-        "support": 1440.0
+        "f1-score": 0.8800746616892207,
+        "precision": 0.8572727272727273,
+        "recall": 0.9041227229146692,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9589184992350025,
-        "precision": 0.945595388218339,
-        "recall": 0.9726224116366332,
-        "support": 21587.0
+        "f1-score": 0.9537844036697248,
+        "precision": 0.9488876212207644,
+        "recall": 0.9587319884726225,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.9091178797468354,
-        "precision": 0.9427750999897446,
-        "recall": 0.877780960565263,
-        "support": 10473.0
+        "f1-score": 0.9100192149327477,
+        "precision": 0.9220157970853265,
+        "recall": 0.8983308042488619,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9403880597014925,
-      "eval_loss": 0.22083505988121033,
+      "eval_accuracy": 0.9364929939534379,
+      "eval_loss": 0.19801080226898193,
       "eval_macro avg": {
-        "f1-score": 0.9161461296773631,
-        "precision": 0.9129519750337626,
-        "recall": 0.9209677907339654,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7449,
-      "eval_samples_per_second": 17.071,
-      "eval_steps_per_second": 2.318,
+        "f1-score": 0.9146260934305644,
+        "precision": 0.9093920485262728,
+        "recall": 0.9203951718787179,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6132,
+      "eval_samples_per_second": 17.342,
+      "eval_steps_per_second": 2.168,
       "eval_weighted avg": {
-        "f1-score": 0.9399744505088647,
-        "precision": 0.9406253819936745,
-        "recall": 0.9403880597014925,
-        "support": 33500.0
+        "f1-score": 0.9363812792925563,
+        "precision": 0.9364514800186444,
+        "recall": 0.9364929939534379,
+        "support": 27619.0
       },
       "step": 324
     },
     {
       "epoch": 5.0,
       "eval_B": {
-        "f1-score": 0.8879781420765027,
-        "precision": 0.8736559139784946,
-        "recall": 0.9027777777777778,
-        "support": 1440.0
+        "f1-score": 0.883877159309021,
+        "precision": 0.8847262247838616,
+        "recall": 0.8830297219558965,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9613201259451634,
-        "precision": 0.9539317642765919,
-        "recall": 0.9688238291564367,
-        "support": 21587.0
+        "f1-score": 0.9591500482927107,
+        "precision": 0.9456643513331839,
+        "recall": 0.9730259365994236,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.916006030835076,
-        "precision": 0.9334853291038858,
-        "recall": 0.899169292466342,
-        "support": 10473.0
+        "f1-score": 0.9191176470588236,
+        "precision": 0.9454503781801513,
+        "recall": 0.8942120095382614,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9442089552238806,
-      "eval_loss": 0.2257838249206543,
+      "eval_accuracy": 0.9432999022412107,
+      "eval_loss": 0.20074255764484406,
       "eval_macro avg": {
-        "f1-score": 0.921768099618914,
-        "precision": 0.9203576691196576,
-        "recall": 0.9235902998001855,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7845,
-      "eval_samples_per_second": 16.93,
-      "eval_steps_per_second": 2.299,
+        "f1-score": 0.9207149515535183,
+        "precision": 0.9252803180990656,
+        "recall": 0.9167558893645271,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6092,
+      "eval_samples_per_second": 17.356,
+      "eval_steps_per_second": 2.17,
       "eval_weighted avg": {
-        "f1-score": 0.9440011117702735,
-        "precision": 0.9440889959185916,
-        "recall": 0.9442089552238806,
-        "support": 33500.0
+        "f1-score": 0.9429348139614955,
+        "precision": 0.9432916158141273,
+        "recall": 0.9432999022412107,
+        "support": 27619.0
       },
       "step": 405
     },
     {
       "epoch": 6.0,
       "eval_B": {
-        "f1-score": 0.8873579056148813,
-        "precision": 0.8803827751196173,
-        "recall": 0.8944444444444445,
-        "support": 1440.0
+        "f1-score": 0.8897338403041826,
+        "precision": 0.882186616399623,
+        "recall": 0.8974113135186961,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9622060984570168,
-        "precision": 0.9539267015706806,
-        "recall": 0.9706304720433594,
-        "support": 21587.0
+        "f1-score": 0.9624160683661308,
+        "precision": 0.9707969272268809,
+        "recall": 0.954178674351585,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.9182769530299342,
-        "precision": 0.9365567911040509,
-        "recall": 0.9006970304592762,
-        "support": 10473.0
+        "f1-score": 0.9297955261331482,
+        "precision": 0.9161493950552341,
+        "recall": 0.9438543247344461,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9454925373134329,
-      "eval_loss": 0.2242102175951004,
+      "eval_accuracy": 0.9485861182519281,
+      "eval_loss": 0.16565027832984924,
       "eval_macro avg": {
-        "f1-score": 0.9226136523672773,
-        "precision": 0.923622089264783,
-        "recall": 0.9219239823156933,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7715,
-      "eval_samples_per_second": 16.976,
-      "eval_steps_per_second": 2.305,
+        "f1-score": 0.9273151449344872,
+        "precision": 0.9230443128939126,
+        "recall": 0.9318147708682424,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6219,
+      "eval_samples_per_second": 17.309,
+      "eval_steps_per_second": 2.164,
       "eval_weighted avg": {
-        "f1-score": 0.945255312255509,
-        "precision": 0.9453351097376493,
-        "recall": 0.9454925373134329,
-        "support": 33500.0
+        "f1-score": 0.9487745648174827,
+        "precision": 0.9491959030765336,
+        "recall": 0.9485861182519281,
+        "support": 27619.0
       },
       "step": 486
     },
     {
       "epoch": 6.17,
-      "grad_norm": 1.8705081939697266,
+      "grad_norm": 4.767655849456787,
       "learning_rate": 1.7530864197530865e-05,
-      "loss": 0.1601,
+      "loss": 0.1585,
       "step": 500
     },
     {
       "epoch": 7.0,
       "eval_B": {
-        "f1-score": 0.8876749743939911,
-        "precision": 0.8730691739422431,
-        "recall": 0.9027777777777778,
-        "support": 1440.0
+        "f1-score": 0.8802588996763754,
+        "precision": 0.85,
+        "recall": 0.912751677852349,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9611415113943341,
-        "precision": 0.9581108451482232,
-        "recall": 0.9641914114976606,
-        "support": 21587.0
+        "f1-score": 0.958423523458803,
+        "precision": 0.9354661691269275,
+        "recall": 0.9825360230547551,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.9164739884393064,
-        "precision": 0.9247594050743657,
-        "recall": 0.9083357204239473,
-        "support": 10473.0
+        "f1-score": 0.9139526911210146,
+        "precision": 0.9664088931851136,
+        "recall": 0.8668978972469109,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.944089552238806,
-      "eval_loss": 0.22517083585262299,
+      "eval_accuracy": 0.9412723125384699,
+      "eval_loss": 0.2591131329536438,
       "eval_macro avg": {
-        "f1-score": 0.9217634914092105,
-        "precision": 0.9186464747216107,
-        "recall": 0.9251016365664619,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7926,
-      "eval_samples_per_second": 16.901,
-      "eval_steps_per_second": 2.295,
+        "f1-score": 0.9175450380853977,
+        "precision": 0.917291687437347,
+        "recall": 0.9207285327180049,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.617,
+      "eval_samples_per_second": 17.327,
+      "eval_steps_per_second": 2.166,
       "eval_weighted avg": {
-        "f1-score": 0.9440192791200506,
-        "precision": 0.9440287663891153,
-        "recall": 0.944089552238806,
-        "support": 33500.0
+        "f1-score": 0.9406164485555297,
+        "precision": 0.9425749115781907,
+        "recall": 0.9412723125384699,
+        "support": 27619.0
       },
       "step": 567
     },
     {
       "epoch": 8.0,
       "eval_B": {
-        "f1-score": 0.894540522210919,
-        "precision": 0.8740888005301524,
-        "recall": 0.9159722222222222,
-        "support": 1440.0
+        "f1-score": 0.8687910028116214,
+        "precision": 0.84967919340055,
+        "recall": 0.8887823585810163,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9638018385639943,
-        "precision": 0.9540255968049378,
-        "recall": 0.9737805160513272,
-        "support": 21587.0
+        "f1-score": 0.9595010203788336,
+        "precision": 0.956997878562009,
+        "recall": 0.9620172910662824,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.9203132648066569,
-        "precision": 0.9441598875163202,
-        "recall": 0.8976415544734078,
-        "support": 10473.0
+        "f1-score": 0.9212035166275324,
+        "precision": 0.9282491471332673,
+        "recall": 0.9142640364188164,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9474925373134329,
-      "eval_loss": 0.24869142472743988,
+      "eval_accuracy": 0.9432999022412107,
+      "eval_loss": 0.2693409323692322,
       "eval_macro avg": {
-        "f1-score": 0.9262185418605234,
-        "precision": 0.9240914282838034,
-        "recall": 0.9291314309156524,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7743,
-      "eval_samples_per_second": 16.966,
-      "eval_steps_per_second": 2.304,
+        "f1-score": 0.9164985132726624,
+        "precision": 0.9116420730319422,
+        "recall": 0.9216878953553717,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6153,
+      "eval_samples_per_second": 17.333,
+      "eval_steps_per_second": 2.167,
       "eval_weighted avg": {
-        "f1-score": 0.947228939205516,
-        "precision": 0.9475052218791054,
-        "recall": 0.9474925373134329,
-        "support": 33500.0
+        "f1-score": 0.9432823550422136,
+        "precision": 0.9433417293609165,
+        "recall": 0.9432999022412107,
+        "support": 27619.0
       },
       "step": 648
     },
     {
       "epoch": 9.0,
       "eval_B": {
-        "f1-score": 0.8851236868858015,
-        "precision": 0.8643282594308405,
-        "recall": 0.9069444444444444,
-        "support": 1440.0
+        "f1-score": 0.8645343367826904,
+        "precision": 0.8485687903970452,
+        "recall": 0.8811121764141898,
+        "support": 1043.0
       },
       "eval_I": {
-        "f1-score": 0.9617001723147618,
-        "precision": 0.9540067462849849,
-        "recall": 0.9695186918052532,
-        "support": 21587.0
+        "f1-score": 0.9567003348057282,
+        "precision": 0.934535260814599,
+        "recall": 0.9799423631123919,
+        "support": 17350.0
       },
       "eval_O": {
-        "f1-score": 0.9165854609237966,
-        "precision": 0.9358272808675754,
-        "recall": 0.8981189725961998,
-        "support": 10473.0
+        "f1-score": 0.9104673003585861,
+        "precision": 0.9586479683567062,
+        "recall": 0.8668978972469109,
+        "support": 9226.0
       },
-      "eval_accuracy": 0.9445074626865672,
-      "eval_loss": 0.3045330345630646,
+      "eval_accuracy": 0.9384481697382236,
+      "eval_loss": 0.3087053894996643,
       "eval_macro avg": {
-        "f1-score": 0.9211364400414532,
-        "precision": 0.9180540955278002,
-        "recall": 0.9248607029486324,
-        "support": 33500.0
-      },
-      "eval_runtime": 4.7599,
-      "eval_samples_per_second": 17.017,
-      "eval_steps_per_second": 2.311,
+        "f1-score": 0.910567323982335,
+        "precision": 0.9139173398561168,
+        "recall": 0.9093174789244975,
+        "support": 27619.0
+      },
+      "eval_runtime": 4.6156,
+      "eval_samples_per_second": 17.333,
+      "eval_steps_per_second": 2.167,
       "eval_weighted avg": {
-        "f1-score": 0.9443044555560967,
-        "precision": 0.9444685205421044,
-        "recall": 0.9445074626865672,
-        "support": 33500.0
+        "f1-score": 0.9377758584761231,
+        "precision": 0.9393435743356523,
+        "recall": 0.9384481697382236,
+        "support": 27619.0
       },
       "step": 729
     }
@@ -372,7 +372,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 1290060993022200.0,
+  "total_flos": 1294079874620400.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

checkpoint-729/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b8ca363922308e3892528dd031742ffedc372734b464a042d09444af9700901e
+oid sha256:821ca433db9ddb761291ab61fc06d651ad248cd9d560cdb25f7f905e5aaff14e
 size 4856