Training in progress, epoch 10, checkpoint

checkpoint-810/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8bfab6dcca559f8e8bcec4cb735066969ef5d718827e72bdca4eba874c6aa313
+oid sha256:aa9d3ff67d83e34c49ed1e06d08d76aed3baa9d5cc1e49f4f7482415d187dfca
 size 592318676

checkpoint-810/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cf8b12394dd1e2159879297504a4f5507318d3c3d813197526dc064ad65a7f23
+oid sha256:bb1073e16f8a72946914f45479a4deed25eb31c5be8b87f5b5f840cfd7a45579
 size 1014645498

checkpoint-810/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05f741fe22f4ca5655fda1f9ef1ecba0015571f3b8afe49bf335811d1e94a138
+oid sha256:516a7ea2b4f58aa1466b3499947a191d75f003f2e001b6d4bbb7a888f04d977f
 size 14244

checkpoint-810/trainer_state.json

@@ -11,397 +11,397 @@
     {
       "epoch": 1.0,
       "eval_B": {
-        "f1-score": 0.8087666523420713,
-        "precision": 0.8189730200174065,
-        "recall": 0.7988115449915111,
-        "support": 1178.0
+        "f1-score": 0.8266764382557715,
+        "precision": 0.7866108786610879,
+        "recall": 0.871042471042471,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9374597034171502,
-        "precision": 0.9521418826739427,
-        "recall": 0.9232234509762421,
-        "support": 18899.0
+        "f1-score": 0.9410658617447671,
+        "precision": 0.9344077040239768,
+        "recall": 0.9478195863443808,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.8881362400419788,
-        "precision": 0.863303347862376,
-        "recall": 0.9144400785854617,
-        "support": 10180.0
+        "f1-score": 0.8562443302086484,
+        "precision": 0.8789421405512788,
+        "recall": 0.8346893055064262,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9154245298608586,
-      "eval_loss": 0.2228638380765915,
+      "eval_accuracy": 0.912335377500754,
+      "eval_loss": 0.2450982630252838,
       "eval_macro avg": {
-        "f1-score": 0.8781208652670669,
-        "precision": 0.8781394168512416,
-        "recall": 0.8788250248510717,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7167,
-      "eval_samples_per_second": 17.173,
-      "eval_steps_per_second": 2.332,
+        "f1-score": 0.8746622100697291,
+        "precision": 0.8666535744121145,
+        "recall": 0.8845171209644261,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6269,
+      "eval_samples_per_second": 17.29,
+      "eval_steps_per_second": 2.161,
       "eval_weighted avg": {
-        "f1-score": 0.9158543469268937,
-        "precision": 0.9170673807540184,
-        "recall": 0.9154245298608586,
-        "support": 30257.0
+        "f1-score": 0.9119949287205363,
+        "precision": 0.9122301519092053,
+        "recall": 0.912335377500754,
+        "support": 29841.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_B": {
-        "f1-score": 0.8401515151515152,
-        "precision": 0.7585499316005472,
-        "recall": 0.9414261460101867,
-        "support": 1178.0
+        "f1-score": 0.8833395452851286,
+        "precision": 0.8537463976945245,
+        "recall": 0.915057915057915,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9568235603470944,
-        "precision": 0.9510219016256338,
-        "recall": 0.9626964389650247,
-        "support": 18899.0
+        "f1-score": 0.9549518227873249,
+        "precision": 0.9516903430183636,
+        "recall": 0.958235733864939,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9185648054827655,
-        "precision": 0.9430877483443708,
-        "recall": 0.8952848722986247,
-        "support": 10180.0
+        "f1-score": 0.88924750463212,
+        "precision": 0.9016969696969697,
+        "recall": 0.877137130055418,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9391876260038999,
-      "eval_loss": 0.16692574322223663,
+      "eval_accuracy": 0.9333132267685399,
+      "eval_loss": 0.20076116919517517,
       "eval_macro avg": {
-        "f1-score": 0.905179960327125,
-        "precision": 0.8842198605235172,
-        "recall": 0.9331358190912787,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7116,
-      "eval_samples_per_second": 17.192,
-      "eval_steps_per_second": 2.335,
+        "f1-score": 0.9091796242348579,
+        "precision": 0.9023779034699526,
+        "recall": 0.916810259659424,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6524,
+      "eval_samples_per_second": 17.195,
+      "eval_steps_per_second": 2.149,
       "eval_weighted avg": {
-        "f1-score": 0.9394089523635119,
-        "precision": 0.9408589092241132,
-        "recall": 0.9391876260038999,
-        "support": 30257.0
+        "f1-score": 0.9331705077630417,
+        "precision": 0.9332314707173981,
+        "recall": 0.9333132267685399,
+        "support": 29841.0
       },
       "step": 162
     },
     {
       "epoch": 3.0,
       "eval_B": {
-        "f1-score": 0.8612244897959184,
-        "precision": 0.8294025157232704,
-        "recall": 0.8955857385398981,
-        "support": 1178.0
+        "f1-score": 0.8927335640138409,
+        "precision": 0.8889739663093414,
+        "recall": 0.8965250965250965,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.951106371634231,
-        "precision": 0.9584654236741712,
-        "recall": 0.9438594634636753,
-        "support": 18899.0
+        "f1-score": 0.9590235396687009,
+        "precision": 0.9586653386454184,
+        "recall": 0.9593820084724645,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9106743212999903,
-        "precision": 0.9021592442645074,
-        "recall": 0.9193516699410609,
-        "support": 10180.0
+        "f1-score": 0.9006849315068494,
+        "precision": 0.9020697811945594,
+        "recall": 0.8993043273198915,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9337343424662061,
-      "eval_loss": 0.1742253601551056,
+      "eval_accuracy": 0.9395797727958178,
+      "eval_loss": 0.19863653182983398,
       "eval_macro avg": {
-        "f1-score": 0.9076683942433799,
-        "precision": 0.8966757278873163,
-        "recall": 0.9195989573148781,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7094,
-      "eval_samples_per_second": 17.2,
-      "eval_steps_per_second": 2.336,
+        "f1-score": 0.9174806783964637,
+        "precision": 0.9165696953831063,
+        "recall": 0.9184038107724842,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.655,
+      "eval_samples_per_second": 17.186,
+      "eval_steps_per_second": 2.148,
       "eval_weighted avg": {
-        "f1-score": 0.9340035812317422,
-        "precision": 0.9344963252190521,
-        "recall": 0.9337343424662061,
-        "support": 30257.0
+        "f1-score": 0.9395665759512081,
+        "precision": 0.9395561516236713,
+        "recall": 0.9395797727958178,
+        "support": 29841.0
       },
       "step": 243
     },
     {
       "epoch": 4.0,
       "eval_B": {
-        "f1-score": 0.8717948717948719,
-        "precision": 0.8254931714719271,
-        "recall": 0.9235993208828522,
-        "support": 1178.0
+        "f1-score": 0.8999616711383672,
+        "precision": 0.893455098934551,
+        "recall": 0.9065637065637066,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9561271586371415,
-        "precision": 0.9374586871408959,
-        "recall": 0.9755542621302714,
-        "support": 18899.0
+        "f1-score": 0.9612657113512985,
+        "precision": 0.9436362763454799,
+        "recall": 0.9795664091701969,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9117828500925355,
-        "precision": 0.956427955133736,
-        "recall": 0.8711198428290766,
-        "support": 10180.0
+        "f1-score": 0.8996847765622102,
+        "precision": 0.9454403741231489,
+        "recall": 0.8581535196321188,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9383944211256899,
-      "eval_loss": 0.18792153894901276,
+      "eval_accuracy": 0.9418920277470594,
+      "eval_loss": 0.19743724167346954,
       "eval_macro avg": {
-        "f1-score": 0.9132349601748496,
-        "precision": 0.9064599379155197,
-        "recall": 0.9234244752807333,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7342,
-      "eval_samples_per_second": 17.11,
-      "eval_steps_per_second": 2.324,
+        "f1-score": 0.920304053017292,
+        "precision": 0.9275105831343932,
+        "recall": 0.9147612117886741,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6615,
+      "eval_samples_per_second": 17.162,
+      "eval_steps_per_second": 2.145,
       "eval_weighted avg": {
-        "f1-score": 0.9379241479327001,
-        "precision": 0.9394817486046585,
-        "recall": 0.9383944211256899,
-        "support": 30257.0
+        "f1-score": 0.9411036309913238,
+        "precision": 0.9419713163409645,
+        "recall": 0.9418920277470594,
+        "support": 29841.0
       },
       "step": 324
     },
     {
       "epoch": 5.0,
       "eval_B": {
-        "f1-score": 0.8742949234488315,
-        "precision": 0.8320552147239264,
-        "recall": 0.9210526315789473,
-        "support": 1178.0
+        "f1-score": 0.8953846153846153,
+        "precision": 0.8919540229885058,
+        "recall": 0.8988416988416988,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9583957528450158,
-        "precision": 0.9521148825065274,
-        "recall": 0.9647600402137679,
-        "support": 18899.0
+        "f1-score": 0.9602426713742261,
+        "precision": 0.9581224570804803,
+        "recall": 0.9623722900573137,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9198818996146724,
-        "precision": 0.9375701315923697,
-        "recall": 0.9028487229862475,
-        "support": 10180.0
+        "f1-score": 0.9022119433078337,
+        "precision": 0.9075399665950847,
+        "recall": 0.8969461148449476,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9422282447037049,
-      "eval_loss": 0.19990424811840057,
+      "eval_accuracy": 0.9410207432726785,
+      "eval_loss": 0.2340461015701294,
       "eval_macro avg": {
-        "f1-score": 0.9175241919695066,
-        "precision": 0.9072467429409411,
-        "recall": 0.9295537982596542,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7101,
-      "eval_samples_per_second": 17.197,
-      "eval_steps_per_second": 2.335,
+        "f1-score": 0.9192797433555584,
+        "precision": 0.9192054822213569,
+        "recall": 0.9193867012479867,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6719,
+      "eval_samples_per_second": 17.124,
+      "eval_steps_per_second": 2.14,
       "eval_weighted avg": {
-        "f1-score": 0.9421634164298524,
-        "precision": 0.9425469857238316,
-        "recall": 0.9422282447037049,
-        "support": 30257.0
+        "f1-score": 0.9409353496612266,
+        "precision": 0.9408751053176122,
+        "recall": 0.9410207432726785,
+        "support": 29841.0
       },
       "step": 405
     },
     {
       "epoch": 6.0,
       "eval_B": {
-        "f1-score": 0.8808247422680412,
-        "precision": 0.8564554931836408,
-        "recall": 0.9066213921901528,
-        "support": 1178.0
+        "f1-score": 0.8929831438651511,
+        "precision": 0.9068471337579618,
+        "recall": 0.8795366795366796,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.957406827435014,
-        "precision": 0.9446361435855178,
-        "recall": 0.9705275411397428,
-        "support": 18899.0
+        "f1-score": 0.9522293798155866,
+        "precision": 0.9655225409836066,
+        "recall": 0.9392972838275604,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9163000050574014,
-        "precision": 0.9443344105076619,
-        "recall": 0.8898821218074656,
-        "support": 10180.0
+        "f1-score": 0.8885216003647556,
+        "precision": 0.8599007170435742,
+        "recall": 0.919113312109421,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9409062365733549,
-      "eval_loss": 0.20615626871585846,
+      "eval_accuracy": 0.930967460875976,
+      "eval_loss": 0.21774999797344208,
       "eval_macro avg": {
-        "f1-score": 0.9181771915868189,
-        "precision": 0.9151420157589403,
-        "recall": 0.922343685045787,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.6949,
-      "eval_samples_per_second": 17.253,
-      "eval_steps_per_second": 2.343,
+        "f1-score": 0.9112447080151643,
+        "precision": 0.9107567972617142,
+        "recall": 0.9126490918245537,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.647,
+      "eval_samples_per_second": 17.215,
+      "eval_steps_per_second": 2.152,
       "eval_weighted avg": {
-        "f1-score": 0.9405948120953971,
-        "precision": 0.941101475610967,
-        "recall": 0.9409062365733549,
-        "support": 30257.0
+        "f1-score": 0.9315521386682287,
+        "precision": 0.9329578366776978,
+        "recall": 0.930967460875976,
+        "support": 29841.0
       },
       "step": 486
     },
     {
       "epoch": 6.17,
-      "grad_norm": 2.4104158878326416,
+      "grad_norm": 9.310327529907227,
       "learning_rate": 1.7530864197530865e-05,
-      "loss": 0.1662,
+      "loss": 0.1627,
       "step": 500
     },
     {
       "epoch": 7.0,
       "eval_B": {
-        "f1-score": 0.8836820083682009,
-        "precision": 0.8712871287128713,
-        "recall": 0.8964346349745331,
-        "support": 1178.0
+        "f1-score": 0.8936490041337843,
+        "precision": 0.8704245973645681,
+        "recall": 0.9181467181467181,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9614174269469787,
-        "precision": 0.9609854091774159,
-        "recall": 0.9618498333245146,
-        "support": 18899.0
+        "f1-score": 0.9599271850630982,
+        "precision": 0.9477800446905664,
+        "recall": 0.9723897333665587,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9271751440248166,
-        "precision": 0.9295093296475466,
-        "recall": 0.924852652259332,
-        "support": 10180.0
+        "f1-score": 0.8979841172877214,
+        "precision": 0.9316770186335404,
+        "recall": 0.8666430845419172,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9468552731599299,
-      "eval_loss": 0.22925782203674316,
+      "eval_accuracy": 0.9399819040916859,
+      "eval_loss": 0.27376389503479004,
       "eval_macro avg": {
-        "f1-score": 0.9240915264466655,
-        "precision": 0.9205939558459445,
-        "recall": 0.9277123735194599,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7288,
-      "eval_samples_per_second": 17.129,
-      "eval_steps_per_second": 2.326,
+        "f1-score": 0.9171867688282013,
+        "precision": 0.9166272202295582,
+        "recall": 0.9190598453517312,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6671,
+      "eval_samples_per_second": 17.141,
+      "eval_steps_per_second": 2.143,
       "eval_weighted avg": {
-        "f1-score": 0.9468700903559945,
-        "precision": 0.9469030129054358,
-        "recall": 0.9468552731599299,
-        "support": 30257.0
+        "f1-score": 0.9394463230910989,
+        "precision": 0.9398464945991887,
+        "recall": 0.9399819040916859,
+        "support": 29841.0
       },
       "step": 567
     },
     {
       "epoch": 8.0,
       "eval_B": {
-        "f1-score": 0.8803630363036303,
-        "precision": 0.8563402889245586,
-        "recall": 0.9057724957555179,
-        "support": 1178.0
+        "f1-score": 0.8948545861297539,
+        "precision": 0.8651766402307137,
+        "recall": 0.9266409266409267,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9580651967507121,
-        "precision": 0.9550928116947994,
-        "recall": 0.9610561405365363,
-        "support": 18899.0
+        "f1-score": 0.9590139880661254,
+        "precision": 0.9414702069429106,
+        "recall": 0.9772240219287316,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9208882720333101,
-        "precision": 0.9294576746047628,
-        "recall": 0.912475442043222,
-        "support": 10180.0
+        "f1-score": 0.8927241122423641,
+        "precision": 0.9427035531663825,
+        "recall": 0.8477773847423653,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9425587467362925,
-      "eval_loss": 0.22973769903182983,
+      "eval_accuracy": 0.9382393351429241,
+      "eval_loss": 0.3103167414665222,
       "eval_macro avg": {
-        "f1-score": 0.9197721683625509,
-        "precision": 0.9136302584080402,
-        "recall": 0.9264346927784253,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.7191,
-      "eval_samples_per_second": 17.164,
-      "eval_steps_per_second": 2.331,
+        "f1-score": 0.9155308954794145,
+        "precision": 0.916450133446669,
+        "recall": 0.9172141111040079,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6339,
+      "eval_samples_per_second": 17.264,
+      "eval_steps_per_second": 2.158,
       "eval_weighted avg": {
-        "f1-score": 0.9425317916335554,
-        "precision": 0.942623096673485,
-        "recall": 0.9425587467362925,
-        "support": 30257.0
+        "f1-score": 0.9373896838414373,
+        "precision": 0.9385098450391195,
+        "recall": 0.9382393351429241,
+        "support": 29841.0
       },
       "step": 648
     },
     {
       "epoch": 9.0,
       "eval_B": {
-        "f1-score": 0.8885191347753745,
-        "precision": 0.8711256117455138,
-        "recall": 0.9066213921901528,
-        "support": 1178.0
+        "f1-score": 0.9011235955056179,
+        "precision": 0.8749090909090909,
+        "recall": 0.9289575289575289,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.9620326487625066,
-        "precision": 0.9574445783763954,
-        "recall": 0.9666649029049156,
-        "support": 18899.0
+        "f1-score": 0.9615119937082186,
+        "precision": 0.9485040973670175,
+        "recall": 0.9748816346872664,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9275707898658718,
-        "precision": 0.9382914572864322,
-        "recall": 0.9170923379174852,
-        "support": 10180.0
+        "f1-score": 0.9013722126929674,
+        "precision": 0.9380339155935229,
+        "recall": 0.8674684589081476,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9476484780381399,
-      "eval_loss": 0.24401217699050903,
+      "eval_accuracy": 0.9423611809255722,
+      "eval_loss": 0.32883673906326294,
       "eval_macro avg": {
-        "f1-score": 0.9260408578012509,
-        "precision": 0.9222872158027805,
-        "recall": 0.9301262110041845,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.701,
-      "eval_samples_per_second": 17.23,
-      "eval_steps_per_second": 2.34,
+        "f1-score": 0.9213359339689347,
+        "precision": 0.9204823679565438,
+        "recall": 0.9237692075176476,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6714,
+      "eval_samples_per_second": 17.126,
+      "eval_steps_per_second": 2.141,
       "eval_weighted avg": {
-        "f1-score": 0.94757580760031,
-        "precision": 0.9476398219436029,
-        "recall": 0.9476484780381399,
-        "support": 30257.0
+        "f1-score": 0.9417992341337168,
+        "precision": 0.9423346276781992,
+        "recall": 0.9423611809255722,
+        "support": 29841.0
       },
       "step": 729
     },
     {
       "epoch": 10.0,
       "eval_B": {
-        "f1-score": 0.875959595959596,
-        "precision": 0.8357748650732459,
-        "recall": 0.9202037351443124,
-        "support": 1178.0
+        "f1-score": 0.8935361216730039,
+        "precision": 0.8801498127340824,
+        "recall": 0.9073359073359073,
+        "support": 1295.0
       },
       "eval_I": {
-        "f1-score": 0.955652945166171,
-        "precision": 0.9424749163879599,
-        "recall": 0.9692047198264458,
-        "support": 18899.0
+        "f1-score": 0.9632332861646931,
+        "precision": 0.9596359319351009,
+        "recall": 0.9668577124345876,
+        "support": 20065.0
       },
       "eval_O": {
-        "f1-score": 0.9127632580563309,
-        "precision": 0.9441469816272966,
-        "recall": 0.8833988212180747,
-        "support": 10180.0
+        "f1-score": 0.9095462405342556,
+        "precision": 0.9200241254523522,
+        "recall": 0.8993043273198915,
+        "support": 8481.0
       },
-      "eval_accuracy": 0.9384274713289487,
-      "eval_loss": 0.29471203684806824,
+      "eval_accuracy": 0.9450755671726819,
+      "eval_loss": 0.3041447699069977,
       "eval_macro avg": {
-        "f1-score": 0.9147919330606994,
-        "precision": 0.9074655876961675,
-        "recall": 0.9242690920629443,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.6985,
-      "eval_samples_per_second": 17.24,
-      "eval_steps_per_second": 2.341,
+        "f1-score": 0.9221052161239841,
+        "precision": 0.9199366233738452,
+        "recall": 0.9244993156967954,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.6532,
+      "eval_samples_per_second": 17.193,
+      "eval_steps_per_second": 2.149,
       "eval_weighted avg": {
-        "f1-score": 0.9381199518045186,
-        "precision": 0.9388833165494999,
-        "recall": 0.9384274713289487,
-        "support": 30257.0
+        "f1-score": 0.9449504651463465,
+        "precision": 0.9449285744355027,
+        "recall": 0.9450755671726819,
+        "support": 29841.0
       },
       "step": 810
     }
@@ -411,7 +411,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 1433401103358000.0,
+  "total_flos": 1437866527356000.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

checkpoint-810/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a9f81281de988dbf2616f1c96e725c00f6e4415b289b670a7cb21136eb66e95f
+oid sha256:d229d7bc9127d222c880edfd3fa9418a9ccede656fe27f7f0aa04990c9605be4
 size 4856