Training in progress, epoch 16, checkpoint

checkpoint-1296/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a5c3f3d72e7fb9eb0d334dc78b460bcbdb7eebeebca84632ddd1ff919d0f1834
+oid sha256:a3d708f9758f2c547bdaa08579108a82ce1482bc7bbd3d014e300e74adb7376e
 size 592324828

checkpoint-1296/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:13d074c22b26a45724013c04d65edab21fb8fc7b140c4253920e2916a79fd537
+oid sha256:6697fd4e4fcebbee138a90a840740c584c10e52ecae58923d06d2fac6f5a64d9
 size 1014657786

checkpoint-1296/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59db0ed55574ed3ea0e40aa442c9d9e535333e11d1299a293e3e03d5b4b6e95b
+oid sha256:2007edf5baaf74add4692b6079d83dd32969851bca3e148bc7224db89e12c09f
 size 14244

checkpoint-1296/trainer_state.json

@@ -11,734 +11,734 @@
     {
       "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
     },
     {
       "epoch": 9.0,
       "eval_Claim": {
-        "f1-score": 0.6049033474776048,
-        "precision": 0.6213075060532688,
-        "recall": 0.5893431327514929,
-        "support": 4354.0
+        "f1-score": 0.6284997910572503,
+        "precision": 0.5998005982053839,
+        "recall": 0.6600833881939873,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.823503227242377,
-        "precision": 0.8628731343283582,
-        "recall": 0.787569178373776,
-        "support": 2349.0
+        "f1-score": 0.816712834718375,
+        "precision": 0.857489093553078,
+        "recall": 0.7796386073159982,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9189658498938837,
-        "precision": 0.9027672479150872,
-        "recall": 0.9357563850687622,
-        "support": 10180.0
+        "f1-score": 0.8971996170416466,
+        "precision": 0.9108249301421455,
+        "recall": 0.8839759462327555,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8989367655288193,
-        "precision": 0.8970292606656243,
-        "recall": 0.9008524001794527,
-        "support": 13374.0
+        "f1-score": 0.8942406935938898,
+        "precision": 0.8943022295623452,
+        "recall": 0.8941791660932985,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8589747826949136,
-      "eval_loss": 0.5689579844474792,
+      "eval_accuracy": 0.8468214872155759,
+      "eval_loss": 0.7009620070457458,
       "eval_macro avg": {
-        "f1-score": 0.8115772975356712,
-        "precision": 0.8209942872405847,
-        "recall": 0.803380274093371,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8593,
-      "eval_samples_per_second": 16.669,
-      "eval_steps_per_second": 2.264,
+        "f1-score": 0.8091632341027905,
+        "precision": 0.8156042128657381,
+        "recall": 0.8044692769590099,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8005,
+      "eval_samples_per_second": 16.665,
+      "eval_steps_per_second": 2.083,
       "eval_weighted avg": {
-        "f1-score": 0.8575077142417292,
-        "precision": 0.8566315824374819,
-        "recall": 0.8589747826949136,
-        "support": 30257.0
+        "f1-score": 0.848605581671113,
+        "precision": 0.8512257939173797,
+        "recall": 0.8468214872155759,
+        "support": 29841.0
       },
       "step": 729
     },
     {
       "epoch": 10.0,
       "eval_Claim": {
-        "f1-score": 0.60162224797219,
-        "precision": 0.6071094480823199,
-        "recall": 0.5962333486449242,
-        "support": 4354.0
+        "f1-score": 0.6189745836985101,
+        "precision": 0.6180266900021877,
+        "recall": 0.619925389510643,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8131676477000218,
-        "precision": 0.8333333333333334,
-        "recall": 0.7939548744146445,
-        "support": 2349.0
+        "f1-score": 0.8208921766420085,
+        "precision": 0.8353102189781022,
+        "recall": 0.8069634200088145,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9197247706422018,
-        "precision": 0.9338801134062373,
-        "recall": 0.905992141453831,
-        "support": 10180.0
+        "f1-score": 0.8951902368987796,
+        "precision": 0.9085610200364299,
+        "recall": 0.8822072868765476,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.9008479864909512,
-        "precision": 0.884834499170693,
-        "recall": 0.9174517720951099,
-        "support": 13374.0
+        "f1-score": 0.8963474827245805,
+        "precision": 0.8870174493027017,
+        "recall": 0.905875877253337,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8577849753775986,
-      "eval_loss": 0.7088457942008972,
+      "eval_accuracy": 0.8479608592205355,
+      "eval_loss": 0.7416248917579651,
       "eval_macro avg": {
-        "f1-score": 0.8088406632013413,
-        "precision": 0.8147893484981459,
-        "recall": 0.8034080341521274,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.9043,
-      "eval_samples_per_second": 16.516,
-      "eval_steps_per_second": 2.243,
+        "f1-score": 0.8078511199909697,
+        "precision": 0.8122288445798553,
+        "recall": 0.8037429934123356,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7804,
+      "eval_samples_per_second": 16.735,
+      "eval_steps_per_second": 2.092,
       "eval_weighted avg": {
-        "f1-score": 0.8573332851434664,
-        "precision": 0.8573728619273148,
-        "recall": 0.8577849753775986,
-        "support": 30257.0
+        "f1-score": 0.8479238041544329,
+        "precision": 0.8481312332795722,
+        "recall": 0.8479608592205355,
+        "support": 29841.0
       },
       "step": 810
     },
     {
       "epoch": 11.0,
       "eval_Claim": {
-        "f1-score": 0.6253176253176252,
-        "precision": 0.6289498141263941,
-        "recall": 0.6217271474506201,
-        "support": 4354.0
+        "f1-score": 0.6350413962855225,
+        "precision": 0.6477973065510158,
+        "recall": 0.6227781435154707,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7947071414672116,
-        "precision": 0.7318996415770609,
-        "recall": 0.8693060876968923,
-        "support": 2349.0
+        "f1-score": 0.8403837767117314,
+        "precision": 0.8316788951230039,
+        "recall": 0.8492728074041428,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9110249527786062,
-        "precision": 0.9221171261823304,
-        "recall": 0.900196463654224,
-        "support": 10180.0
+        "f1-score": 0.9076665470637484,
+        "precision": 0.920883387938357,
+        "recall": 0.894823723617498,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8996578818752584,
-        "precision": 0.9047258979206049,
-        "recall": 0.8946463286974727,
-        "support": 13374.0
+        "f1-score": 0.9016170675363501,
+        "precision": 0.8904844987250033,
+        "recall": 0.9130315123159488,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8552731599299336,
-      "eval_loss": 0.7312190532684326,
+      "eval_accuracy": 0.8586843604436849,
+      "eval_loss": 0.753268837928772,
       "eval_macro avg": {
-        "f1-score": 0.8076769003596753,
-        "precision": 0.7969231199515976,
-        "recall": 0.8214690068748023,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.9029,
-      "eval_samples_per_second": 16.521,
-      "eval_steps_per_second": 2.244,
+        "f1-score": 0.8211771968993381,
+        "precision": 0.822711022084345,
+        "recall": 0.8199765467132651,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7829,
+      "eval_samples_per_second": 16.726,
+      "eval_steps_per_second": 2.091,
       "eval_weighted avg": {
-        "f1-score": 0.8558567785115953,
-        "precision": 0.857475501606145,
-        "recall": 0.8552731599299336,
-        "support": 30257.0
+        "f1-score": 0.8579718131916827,
+        "precision": 0.8575920866110881,
+        "recall": 0.8586843604436849,
+        "support": 29841.0
       },
       "step": 891
     },
     {
       "epoch": 12.0,
       "eval_Claim": {
-        "f1-score": 0.633677135241627,
-        "precision": 0.6395321637426901,
-        "recall": 0.6279283417547084,
-        "support": 4354.0
+        "f1-score": 0.6230640668523677,
+        "precision": 0.6328655500226347,
+        "recall": 0.6135615536537196,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8290997764681975,
-        "precision": 0.7931570762052877,
-        "recall": 0.8684546615581098,
-        "support": 2349.0
+        "f1-score": 0.837270341207349,
+        "precision": 0.8310898827616153,
+        "recall": 0.8435434111943587,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.916678984971668,
-        "precision": 0.9196243203163619,
-        "recall": 0.9137524557956778,
-        "support": 10180.0
+        "f1-score": 0.8982022214588343,
+        "precision": 0.8730105731775181,
+        "recall": 0.9248909326730338,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.9029209944129889,
-        "precision": 0.905603610379842,
-        "recall": 0.9002542246149244,
-        "support": 13374.0
+        "f1-score": 0.8888346297394397,
+        "precision": 0.9013795542978422,
+        "recall": 0.8766340993532407,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8631391083055161,
-      "eval_loss": 0.7092207074165344,
+      "eval_accuracy": 0.8476592607486344,
+      "eval_loss": 0.8125314712524414,
       "eval_macro avg": {
-        "f1-score": 0.8205942227736204,
-        "precision": 0.8144792926610454,
-        "recall": 0.8275974209308551,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8734,
-      "eval_samples_per_second": 16.621,
-      "eval_steps_per_second": 2.257,
+        "f1-score": 0.8118428148144977,
+        "precision": 0.8095863900649026,
+        "recall": 0.8146574992185882,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7824,
+      "eval_samples_per_second": 16.728,
+      "eval_steps_per_second": 2.091,
       "eval_weighted avg": {
-        "f1-score": 0.8630744313070275,
-        "precision": 0.8633032778855295,
-        "recall": 0.8631391083055161,
-        "support": 30257.0
+        "f1-score": 0.846990546753497,
+        "precision": 0.8469677413197487,
+        "recall": 0.8476592607486344,
+        "support": 29841.0
       },
       "step": 972
     },
     {
       "epoch": 12.35,
-      "grad_norm": 4.426708698272705,
+      "grad_norm": 15.335172653198242,
       "learning_rate": 1.506172839506173e-05,
-      "loss": 0.063,
+      "loss": 0.061,
       "step": 1000
     },
     {
       "epoch": 13.0,
       "eval_Claim": {
-        "f1-score": 0.5936837143530521,
-        "precision": 0.6096321393998064,
-        "recall": 0.5785484611851172,
-        "support": 4354.0
+        "f1-score": 0.6158673240699238,
+        "precision": 0.6292649416075109,
+        "recall": 0.6030283080974326,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8065005417118094,
-        "precision": 0.821270962047661,
-        "recall": 0.7922520221370796,
-        "support": 2349.0
+        "f1-score": 0.8147180492053124,
+        "precision": 0.8050774526678142,
+        "recall": 0.8245923314235346,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9133826964170352,
-        "precision": 0.9242683294780247,
-        "recall": 0.9027504911591355,
-        "support": 10180.0
+        "f1-score": 0.9060346875742457,
+        "precision": 0.9128665469778575,
+        "recall": 0.8993043273198915,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.901282521069989,
-        "precision": 0.883730957171601,
-        "recall": 0.9195453865709585,
-        "support": 13374.0
+        "f1-score": 0.9008830850011933,
+        "precision": 0.8929368029739777,
+        "recall": 0.9089720655015825,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.854942657897346,
-      "eval_loss": 0.7807008624076843,
+      "eval_accuracy": 0.8530880332428538,
+      "eval_loss": 0.8537381887435913,
       "eval_macro avg": {
-        "f1-score": 0.8037123683879714,
-        "precision": 0.8097255970242733,
-        "recall": 0.7982740902630727,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8517,
-      "eval_samples_per_second": 16.695,
-      "eval_steps_per_second": 2.267,
+        "f1-score": 0.8093757864626687,
+        "precision": 0.8100364360567901,
+        "recall": 0.8089742580856103,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7933,
+      "eval_samples_per_second": 16.69,
+      "eval_steps_per_second": 2.086,
       "eval_weighted avg": {
-        "f1-score": 0.8537315976828396,
-        "precision": 0.8530777420132861,
-        "recall": 0.854942657897346,
-        "support": 30257.0
+        "f1-score": 0.8522709558043637,
+        "precision": 0.8516552983596964,
+        "recall": 0.8530880332428538,
+        "support": 29841.0
       },
       "step": 1053
     },
     {
       "epoch": 14.0,
       "eval_Claim": {
-        "f1-score": 0.6335702820454024,
-        "precision": 0.632554945054945,
-        "recall": 0.6345888837850253,
-        "support": 4354.0
+        "f1-score": 0.6346033728919426,
+        "precision": 0.6036838978015449,
+        "recall": 0.6688610928242265,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8183556405353729,
-        "precision": 0.816793893129771,
-        "recall": 0.8199233716475096,
-        "support": 2349.0
+        "f1-score": 0.8067737733391228,
+        "precision": 0.7950363714163458,
+        "recall": 0.8188629352137505,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9172396976347231,
-        "precision": 0.9107990314769976,
-        "recall": 0.9237721021611002,
-        "support": 10180.0
+        "f1-score": 0.9004597838418822,
+        "precision": 0.9121703363174449,
+        "recall": 0.8890461030538852,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8981941309255079,
-        "precision": 0.9039073148568832,
-        "recall": 0.8925527142216241,
-        "support": 13374.0
+        "f1-score": 0.8921770010096438,
+        "precision": 0.903023468884347,
+        "recall": 0.8815880005504335,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8602967908252636,
-      "eval_loss": 0.7879041433334351,
+      "eval_accuracy": 0.8464528668610302,
+      "eval_loss": 0.9449928998947144,
       "eval_macro avg": {
-        "f1-score": 0.8168399377852517,
-        "precision": 0.8160137961296492,
-        "recall": 0.8177092679538147,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.864,
-      "eval_samples_per_second": 16.653,
-      "eval_steps_per_second": 2.262,
+        "f1-score": 0.8085034827706478,
+        "precision": 0.8034785186049206,
+        "recall": 0.8145895329105739,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.8154,
+      "eval_samples_per_second": 16.613,
+      "eval_steps_per_second": 2.077,
       "eval_weighted avg": {
-        "f1-score": 0.8603242501425289,
-        "precision": 0.8604152313535001,
-        "recall": 0.8602967908252636,
-        "support": 30257.0
+        "f1-score": 0.8487033685671466,
+        "precision": 0.8517001698367573,
+        "recall": 0.8464528668610302,
+        "support": 29841.0
       },
       "step": 1134
     },
     {
       "epoch": 15.0,
       "eval_Claim": {
-        "f1-score": 0.6212592513139547,
-        "precision": 0.5828134433487623,
-        "recall": 0.6651355075792375,
-        "support": 4354.0
+        "f1-score": 0.6062717770034843,
+        "precision": 0.5687367813881946,
+        "recall": 0.6491112574061882,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7988422575976845,
-        "precision": 0.7765273311897106,
-        "recall": 0.822477650063857,
-        "support": 2349.0
+        "f1-score": 0.8176843948347561,
+        "precision": 0.8121739130434783,
+        "recall": 0.8232701630674306,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9143625765954267,
-        "precision": 0.9276255938542404,
-        "recall": 0.9014734774066797,
-        "support": 10180.0
+        "f1-score": 0.8942823803967327,
+        "precision": 0.8850906571197598,
+        "recall": 0.9036670203985379,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8928883984627678,
-        "precision": 0.9090416053304409,
-        "recall": 0.8772992373261552,
-        "support": 13374.0
+        "f1-score": 0.877618288144604,
+        "precision": 0.9049777063080184,
+        "recall": 0.8518645933672767,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8506461314737086,
-      "eval_loss": 0.8185116052627563,
+      "eval_accuracy": 0.8334506216279616,
+      "eval_loss": 1.0531803369522095,
       "eval_macro avg": {
-        "f1-score": 0.8068381209924584,
-        "precision": 0.7990019934307886,
-        "recall": 0.8165964680939823,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.8575,
-      "eval_samples_per_second": 16.675,
-      "eval_steps_per_second": 2.265,
+        "f1-score": 0.7989642100948944,
+        "precision": 0.7927447644648628,
+        "recall": 0.8069782585598584,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.774,
+      "eval_samples_per_second": 16.757,
+      "eval_steps_per_second": 2.095,
       "eval_weighted avg": {
-        "f1-score": 0.8537245501570023,
-        "precision": 0.8580620487302317,
-        "recall": 0.8506461314737086,
-        "support": 30257.0
+        "f1-score": 0.8363600230462553,
+        "precision": 0.8409220859889104,
+        "recall": 0.8334506216279616,
+        "support": 29841.0
       },
       "step": 1215
     },
     {
       "epoch": 16.0,
       "eval_Claim": {
-        "f1-score": 0.620457750298297,
-        "precision": 0.5878725590955807,
-        "recall": 0.6568672485071199,
-        "support": 4354.0
+        "f1-score": 0.6349479110729246,
+        "precision": 0.6217921750105174,
+        "recall": 0.6486723721746763,
+        "support": 4557.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8217720482913852,
-        "precision": 0.7911741528762806,
-        "recall": 0.8548318433375904,
-        "support": 2349.0
+        "f1-score": 0.8260292164674634,
+        "precision": 0.8297020898176968,
+        "recall": 0.8223887174966946,
+        "support": 2269.0
       },
       "eval_O": {
-        "f1-score": 0.9171081176529942,
-        "precision": 0.9427445285758739,
-        "recall": 0.8928290766208251,
-        "support": 10180.0
+        "f1-score": 0.9045516104339634,
+        "precision": 0.9034344860032933,
+        "recall": 0.9056715010022403,
+        "support": 8481.0
       },
       "eval_Premise": {
-        "f1-score": 0.8952495580546884,
-        "precision": 0.900703852266707,
-        "recall": 0.8898609241812472,
-        "support": 13374.0
+        "f1-score": 0.8956702459300312,
+        "precision": 0.90185546875,
+        "recall": 0.8895692858125774,
+        "support": 14534.0
       },
-      "eval_accuracy": 0.8546121558647586,
-      "eval_loss": 0.8933361768722534,
+      "eval_accuracy": 0.8522502597097953,
+      "eval_loss": 0.9694080352783203,
       "eval_macro avg": {
-        "f1-score": 0.8136468685743412,
-        "precision": 0.8056237732036106,
-        "recall": 0.8235972731616956,
-        "support": 30257.0
-      },
-      "eval_runtime": 4.839,
-      "eval_samples_per_second": 16.739,
-      "eval_steps_per_second": 2.273,
+        "f1-score": 0.8152997459760956,
+        "precision": 0.8141960548953769,
+        "recall": 0.8165754691215471,
+        "support": 29841.0
+      },
+      "eval_runtime": 4.7829,
+      "eval_samples_per_second": 16.726,
+      "eval_steps_per_second": 2.091,
       "eval_weighted avg": {
-        "f1-score": 0.8573567707758909,
-        "precision": 0.8613285464033407,
-        "recall": 0.8546121558647586,
-        "support": 30257.0
+        "f1-score": 0.8530843968151707,
+        "precision": 0.854049671322198,
+        "recall": 0.8522502597097953,
+        "support": 29841.0
       },
       "step": 1296
     }
@@ -748,7 +748,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 50,
   "save_steps": 500,
-  "total_flos": 2293474941878400.0,
+  "total_flos": 2300619723628800.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

checkpoint-1296/training_args.bin

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