Training in progress, epoch 8, checkpoint

checkpoint-648/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:60c9ae57744dc6358c24fa2e14d4fc6e36499d544ca81303879d2c68cf04c3a0
+oid sha256:b9ff071875cc2b7f405d77faeef86cc07f98c660f1f0a081d26f7ea134218e05
 size 592324828

checkpoint-648/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1f678887af8ae33b6144c8b383ac63501414d4822ec65b790f08d63e4a57193d
+oid sha256:85f47b029daaa5f43c40775a5607a35e88e468c4a9375bd77b6f575323a2579b
 size 1014657786

checkpoint-648/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e241cb4a4c96fc25c055fcd71173708b81cd84562a71b4d7503d9ffa5397d596
+oid sha256:dffdf7f1c15f3344dffce2b259bf63e6131cbf6af7c2108c97e9b326dccb1d62
 size 14244

checkpoint-648/trainer_state.json

@@ -11,367 +11,367 @@
     {
       "epoch": 1.0,
       "eval_Claim": {
-        "f1-score": 0.44734338904742493,
-        "precision": 0.5477537437603993,
-        "recall": 0.3780431786862655,
-        "support": 4354.0
+        "f1-score": 0.38183790485045715,
+        "precision": 0.6497890295358649,
+        "recall": 0.27035330261136714,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.6504965622612681,
-        "precision": 0.5898856944925528,
-        "recall": 0.7249893571732652,
-        "support": 2349.0
+        "f1-score": 0.6266258607498087,
+        "precision": 0.5535653937140926,
+        "recall": 0.7219039224327898,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.8859784283513097,
-        "precision": 0.8689081979599547,
-        "recall": 0.9037328094302554,
-        "support": 10180.0
+        "f1-score": 0.8545803971812939,
+        "precision": 0.8443037974683544,
+        "recall": 0.8651102464332037,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8720120805863504,
-        "precision": 0.859258183929738,
-        "recall": 0.8851502916105877,
-        "support": 13374.0
+        "f1-score": 0.878949075575738,
+        "precision": 0.8314310260186549,
+        "recall": 0.9322278794550709,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8059953068711373,
-      "eval_loss": 0.47828415036201477,
+      "eval_accuracy": 0.7960859220535504,
+      "eval_loss": 0.5359914898872375,
       "eval_macro avg": {
-        "f1-score": 0.7139576150615883,
-        "precision": 0.7164514550356611,
-        "recall": 0.7229789092250934,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8422,
-      "eval_samples_per_second": 16.728,
-      "eval_steps_per_second": 2.272,
+        "f1-score": 0.6854983095893245,
+        "precision": 0.7197723116842417,
+        "recall": 0.697398837733108,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8191,
+      "eval_samples_per_second": 16.601,
+      "eval_steps_per_second": 2.075,
       "eval_weighted avg": {
-        "f1-score": 0.7984036588902531,
-        "precision": 0.7967665566250599,
-        "recall": 0.8059953068711373,
-        "support": 30257.0
+        "f1-score": 0.7769240850962494,
+        "precision": 0.7862232339538372,
+        "recall": 0.7960859220535504,
+        "support": 29841.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_Claim": {
-        "f1-score": 0.5787561000934482,
-        "precision": 0.5281409891984082,
-        "recall": 0.6401010564997703,
-        "support": 4354.0
+        "f1-score": 0.6161300742974051,
+        "precision": 0.6048625792811839,
+        "recall": 0.6278253236778583,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7702650787552824,
-        "precision": 0.7017850892544627,
-        "recall": 0.8535547041294168,
-        "support": 2349.0
+        "f1-score": 0.7920215150156881,
+        "precision": 0.8057455540355677,
+        "recall": 0.7787571617452622,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9095337508698677,
-        "precision": 0.9206077681626081,
-        "recall": 0.8987229862475442,
-        "support": 10180.0
+        "f1-score": 0.8922796352583586,
+        "precision": 0.9209436566695947,
+        "recall": 0.865346067680698,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8736648538675224,
-        "precision": 0.9162905211325401,
-        "recall": 0.8348287722446538,
-        "support": 13374.0
+        "f1-score": 0.8914289590611538,
+        "precision": 0.8790554552143957,
+        "recall": 0.9041557726709784,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8297584030141786,
-      "eval_loss": 0.4241589605808258,
+      "eval_accuracy": 0.8413927147213566,
+      "eval_loss": 0.443877637386322,
       "eval_macro avg": {
-        "f1-score": 0.7830549458965301,
-        "precision": 0.7667060919370048,
-        "recall": 0.8068018797803463,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8636,
-      "eval_samples_per_second": 16.654,
-      "eval_steps_per_second": 2.262,
+        "f1-score": 0.7979650459081514,
+        "precision": 0.8026518113001855,
+        "recall": 0.7940210814436992,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8087,
+      "eval_samples_per_second": 16.637,
+      "eval_steps_per_second": 2.08,
       "eval_weighted avg": {
-        "f1-score": 0.8352680063880269,
-        "precision": 0.8452350051608073,
-        "recall": 0.8297584030141786,
-        "support": 30257.0
+        "f1-score": 0.842071433389123,
+        "precision": 0.8435143116581857,
+        "recall": 0.8413927147213566,
+        "support": 29841.0
       },
       "step": 162
     },
     {
       "epoch": 3.0,
       "eval_Claim": {
-        "f1-score": 0.5728465083303793,
-        "precision": 0.5899245558530056,
-        "recall": 0.5567294441892513,
-        "support": 4354.0
+        "f1-score": 0.6042799305957202,
+        "precision": 0.6389432485322897,
+        "recall": 0.5731841123546193,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7862876979822088,
-        "precision": 0.8017699115044248,
-        "recall": 0.7713920817369093,
-        "support": 2349.0
+        "f1-score": 0.8349557522123894,
+        "precision": 0.8382940915148823,
+        "recall": 0.8316438959894227,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.914532821534425,
-        "precision": 0.9048168445341794,
-        "recall": 0.9244597249508841,
-        "support": 10180.0
+        "f1-score": 0.8941106484235574,
+        "precision": 0.9022691799735862,
+        "recall": 0.8860983374602052,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8880533114924986,
-        "precision": 0.8843330614665975,
-        "recall": 0.8918049947659639,
-        "support": 13374.0
+        "f1-score": 0.8917763490086511,
+        "precision": 0.8729980887102089,
+        "recall": 0.9113802119168846,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8452258981392735,
-      "eval_loss": 0.41067537665367126,
+      "eval_accuracy": 0.8464863778023525,
+      "eval_loss": 0.4555923044681549,
       "eval_macro avg": {
-        "f1-score": 0.7904300848348779,
-        "precision": 0.7952110933395518,
-        "recall": 0.7860965614107521,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8684,
-      "eval_samples_per_second": 16.638,
-      "eval_steps_per_second": 2.259,
+        "f1-score": 0.8062806700600795,
+        "precision": 0.8131261521827418,
+        "recall": 0.8005766394302829,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7811,
+      "eval_samples_per_second": 16.733,
+      "eval_steps_per_second": 2.092,
       "eval_weighted avg": {
-        "f1-score": 0.8437033615676307,
-        "precision": 0.8424495118392472,
-        "recall": 0.8452258981392735,
-        "support": 30257.0
+        "f1-score": 0.8442159482345275,
+        "precision": 0.8429359878649869,
+        "recall": 0.8464863778023525,
+        "support": 29841.0
       },
       "step": 243
     },
     {
       "epoch": 4.0,
       "eval_Claim": {
-        "f1-score": 0.5865384615384616,
-        "precision": 0.5714596949891068,
-        "recall": 0.6024345429490124,
-        "support": 4354.0
+        "f1-score": 0.6278026905829596,
+        "precision": 0.5981717011128775,
+        "recall": 0.6605222734254992,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7783630470016208,
-        "precision": 0.7425589485890993,
-        "recall": 0.8177948063005535,
-        "support": 2349.0
+        "f1-score": 0.8182207014542344,
+        "precision": 0.7947652679684254,
+        "recall": 0.8431026884089907,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9108591535529275,
-        "precision": 0.9063411787589963,
-        "recall": 0.9154223968565816,
-        "support": 10180.0
+        "f1-score": 0.9069328626209752,
+        "precision": 0.9373427277302466,
+        "recall": 0.8784341469166372,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8868256151612411,
-        "precision": 0.9067823097358962,
-        "recall": 0.8677284282937042,
-        "support": 13374.0
+        "f1-score": 0.888919552918449,
+        "precision": 0.8913795489137954,
+        "recall": 0.8864730975643319,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8417225765938461,
-      "eval_loss": 0.4545813202857971,
+      "eval_accuracy": 0.8463858449783854,
+      "eval_loss": 0.4377317428588867,
       "eval_macro avg": {
-        "f1-score": 0.7906465693135627,
-        "precision": 0.7817855330182747,
-        "recall": 0.8008450435999629,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.9112,
-      "eval_samples_per_second": 16.493,
-      "eval_steps_per_second": 2.24,
+        "f1-score": 0.8104689518941546,
+        "precision": 0.8054148114313363,
+        "recall": 0.8171330515788647,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8109,
+      "eval_samples_per_second": 16.629,
+      "eval_steps_per_second": 2.079,
       "eval_weighted avg": {
-        "f1-score": 0.8432797441676475,
-        "precision": 0.8456313015960876,
-        "recall": 0.8417225765938461,
-        "support": 30257.0
+        "f1-score": 0.8487883791626096,
+        "precision": 0.8523207959781866,
+        "recall": 0.8463858449783854,
+        "support": 29841.0
       },
       "step": 324
     },
     {
       "epoch": 5.0,
       "eval_Claim": {
-        "f1-score": 0.597399608790703,
-        "precision": 0.5985704403965875,
-        "recall": 0.5962333486449242,
-        "support": 4354.0
+        "f1-score": 0.6247990569070839,
+        "precision": 0.6105990783410138,
+        "recall": 0.6396752249286811,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7933532168726033,
-        "precision": 0.7940298507462686,
-        "recall": 0.7926777352064709,
-        "support": 2349.0
+        "f1-score": 0.7998289136013687,
+        "precision": 0.7769007062733694,
+        "recall": 0.8241516086381666,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9101112629318758,
-        "precision": 0.9042862684251358,
-        "recall": 0.9160117878192534,
-        "support": 10180.0
+        "f1-score": 0.9028039585296891,
+        "precision": 0.9020600353148911,
+        "recall": 0.9035491097747907,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8923677591320345,
-        "precision": 0.8961019377214808,
-        "recall": 0.8886645730521908,
-        "support": 13374.0
+        "f1-score": 0.8885327014878567,
+        "precision": 0.9001058948111542,
+        "recall": 0.8772533370028898,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.848332617245596,
-      "eval_loss": 0.48704993724823,
+      "eval_accuracy": 0.8444086994403672,
+      "eval_loss": 0.5236712694168091,
       "eval_macro avg": {
-        "f1-score": 0.7983079619318041,
-        "precision": 0.7982471243223681,
-        "recall": 0.7983968611807098,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8769,
-      "eval_samples_per_second": 16.609,
-      "eval_steps_per_second": 2.256,
+        "f1-score": 0.8039911576314996,
+        "precision": 0.7974164286851071,
+        "recall": 0.8111573200861321,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8153,
+      "eval_samples_per_second": 16.614,
+      "eval_steps_per_second": 2.077,
       "eval_weighted avg": {
-        "f1-score": 0.8482045037639816,
-        "precision": 0.8481162489521332,
-        "recall": 0.848332617245596,
-        "support": 30257.0
+        "f1-score": 0.8455693764619782,
+        "precision": 0.8470828034323642,
+        "recall": 0.8444086994403672,
+        "support": 29841.0
       },
       "step": 405
     },
     {
       "epoch": 6.0,
       "eval_Claim": {
-        "f1-score": 0.6060809300245921,
-        "precision": 0.5903745644599303,
-        "recall": 0.6226458429030777,
-        "support": 4354.0
+        "f1-score": 0.6166529266281945,
+        "precision": 0.5813095006800078,
+        "recall": 0.6565723063418916,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7858529221479404,
-        "precision": 0.7963286713286714,
-        "recall": 0.7756492124308216,
-        "support": 2349.0
+        "f1-score": 0.8074107959743824,
+        "precision": 0.8392772230147408,
+        "recall": 0.7778757161745262,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9135615470059578,
-        "precision": 0.9398504051527115,
-        "recall": 0.8887033398821218,
-        "support": 10180.0
+        "f1-score": 0.8913107165875667,
+        "precision": 0.8858607034707664,
+        "recall": 0.8968282042212004,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.895557603686636,
-        "precision": 0.88328121591157,
-        "recall": 0.908180050844923,
-        "support": 13374.0
+        "f1-score": 0.8815305371596762,
+        "precision": 0.8981792217065334,
+        "recall": 0.865487821659557,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8502495290346036,
-      "eval_loss": 0.5423789024353027,
+      "eval_accuracy": 0.8358298984618477,
+      "eval_loss": 0.5537309050559998,
       "eval_macro avg": {
-        "f1-score": 0.8002632507162816,
-        "precision": 0.8024587142132208,
-        "recall": 0.7987946115152361,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.867,
-      "eval_samples_per_second": 16.643,
-      "eval_steps_per_second": 2.26,
+        "f1-score": 0.7992262440874549,
+        "precision": 0.8011566622180121,
+        "recall": 0.7991910120992938,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7996,
+      "eval_samples_per_second": 16.668,
+      "eval_steps_per_second": 2.084,
       "eval_weighted avg": {
-        "f1-score": 0.8514422719925407,
-        "precision": 0.8534139871324166,
-        "recall": 0.8502495290346036,
-        "support": 30257.0
+        "f1-score": 0.838225042631562,
+        "precision": 0.8418105575563007,
+        "recall": 0.8358298984618477,
+        "support": 29841.0
       },
       "step": 486
     },
     {
       "epoch": 6.17,
-      "grad_norm": 4.494631767272949,
+      "grad_norm": 6.804128646850586,
       "learning_rate": 1.7530864197530865e-05,
-      "loss": 0.35,
+      "loss": 0.3426,
       "step": 500
     },
     {
       "epoch": 7.0,
       "eval_Claim": {
-        "f1-score": 0.6040515653775324,
-        "precision": 0.6054453161052146,
-        "recall": 0.6026642168121268,
-        "support": 4354.0
+        "f1-score": 0.6206536493682782,
+        "precision": 0.5920318725099601,
+        "recall": 0.652183454026772,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8052777186635455,
-        "precision": 0.8051063829787234,
-        "recall": 0.8054491272882077,
-        "support": 2349.0
+        "f1-score": 0.8364370652905541,
+        "precision": 0.8519195612431444,
+        "recall": 0.8215072719259586,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9184720354599332,
-        "precision": 0.9315082331548641,
-        "recall": 0.9057956777996071,
-        "support": 10180.0
+        "f1-score": 0.9008658773634918,
+        "precision": 0.9000706214689266,
+        "recall": 0.9016625397948355,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8908606921029282,
-        "precision": 0.8810881965774463,
-        "recall": 0.9008524001794527,
-        "support": 13374.0
+        "f1-score": 0.8854242963272995,
+        "precision": 0.8978566881233642,
+        "recall": 0.8733314985551122,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8521994910268698,
-      "eval_loss": 0.5764245986938477,
+      "eval_accuracy": 0.8436714587312758,
+      "eval_loss": 0.5738794207572937,
       "eval_macro avg": {
-        "f1-score": 0.8046655029009848,
-        "precision": 0.8057870322040621,
-        "recall": 0.8036903555198486,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8398,
-      "eval_samples_per_second": 16.736,
-      "eval_steps_per_second": 2.273,
+        "f1-score": 0.8108452220874058,
+        "precision": 0.8104696858363488,
+        "recall": 0.8121711910756696,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8167,
+      "eval_samples_per_second": 16.609,
+      "eval_steps_per_second": 2.076,
       "eval_weighted avg": {
-        "f1-score": 0.8522343290465387,
-        "precision": 0.8524880574571969,
-        "recall": 0.8521994910268698,
-        "support": 30257.0
+        "f1-score": 0.8456551258354699,
+        "precision": 0.8482906998207675,
+        "recall": 0.8436714587312758,
+        "support": 29841.0
       },
       "step": 567
     },
     {
       "epoch": 8.0,
       "eval_Claim": {
-        "f1-score": 0.597346719254213,
-        "precision": 0.6227261400448543,
-        "recall": 0.5739549839228296,
-        "support": 4354.0
+        "f1-score": 0.6410105757931845,
+        "precision": 0.5787798408488064,
+        "recall": 0.7182356813693219,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8047001620745543,
-        "precision": 0.7676845767298028,
-        "recall": 0.8454661558109834,
-        "support": 2349.0
+        "f1-score": 0.8115942028985507,
+        "precision": 0.8804123711340206,
+        "recall": 0.75275451740855,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9099175618286286,
-        "precision": 0.9258769700050838,
-        "recall": 0.8944990176817289,
-        "support": 10180.0
+        "f1-score": 0.8960937967338637,
+        "precision": 0.9094220495386109,
+        "recall": 0.8831505718665251,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8942491542873952,
-        "precision": 0.8797569092750688,
-        "recall": 0.9092268580828473,
-        "support": 13374.0
+        "f1-score": 0.891816143497758,
+        "precision": 0.9084939329050679,
+        "recall": 0.8757396449704142,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8510757841160723,
-      "eval_loss": 0.5954988598823547,
+      "eval_accuracy": 0.8444422103816897,
+      "eval_loss": 0.5891872644424438,
       "eval_macro avg": {
-        "f1-score": 0.8015533993611977,
-        "precision": 0.7990111490137024,
-        "recall": 0.8057867538745973,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8609,
-      "eval_samples_per_second": 16.663,
-      "eval_steps_per_second": 2.263,
+        "f1-score": 0.8101286797308392,
+        "precision": 0.8192770486066265,
+        "recall": 0.8074701039037029,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.799,
+      "eval_samples_per_second": 16.67,
+      "eval_steps_per_second": 2.084,
       "eval_weighted avg": {
-        "f1-score": 0.8498442431569896,
-        "precision": 0.8495864475589161,
-        "recall": 0.8510757841160723,
-        "support": 30257.0
+        "f1-score": 0.8486317402219317,
+        "precision": 0.8562720628608399,
+        "recall": 0.8444422103816897,
+        "support": 29841.0
       },
       "step": 648
     }
@@ -381,7 +381,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 1146737470939200.0,
+  "total_flos": 1150309861814400.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

checkpoint-648/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ce83aa27c1a93c128c1dae6a1646206ab5d85d935efad925ac4e4beb0be29a0
+oid sha256:19f05e33d3889e7c1ce1bf18bf9e16659f52e74109a304d618b5d09955736add
 size 4856