Upload folder using huggingface_hub

training_checkpoints/checkpoint-5144/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ea300bcfa175499d4ec19f323251c0ec04ab53f6fccdfb45becc32f36e94682b
+oid sha256:ff7f93079e95aa5c61d99025b4bf1a1c4773cad4dea7ce3b64cff8b6d0ce1b89
 size 498652812

training_checkpoints/checkpoint-5144/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:321774817049c23659a009fc00e21d4013316cfd8e493a679b9a46fdd4fd8775
+oid sha256:eb0531ba0801cbddadf33dc28120bfc0e875f336bd3a329e0074359b32fe1b2d
 size 997425402

training_checkpoints/checkpoint-5144/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0d90fe545693cad7c5dff78bd2960c30cc079f3f121837e859c9791d09f6eef
+oid sha256:8492805734002a5f78e265565c4e07a3492d7a49a228bc5dc9014ab157685e07
 size 14244

training_checkpoints/checkpoint-5144/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aadc80e92d29a213f8abc69ff587b3ed8f88ce4c228c261ab308dd3144b68b7c
+oid sha256:923ff4843d2ebea772f61352d499b161b5c86d709b968823cf51a662319cffc7
 size 1064

training_checkpoints/checkpoint-5144/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.2703391292878622,
+  "best_metric": 0.7328612504365998,
   "best_model_checkpoint": "roberta-base/checkpoint-5144",
   "epoch": 8.0,
   "eval_steps": 500,
@@ -10,248 +10,248 @@
   "log_history": [
     {
       "epoch": 0.78,
-      "grad_norm": 6.534026145935059,
+      "grad_norm": 16.478424072265625,
       "learning_rate": 4.742871954380509e-05,
-      "loss": 2.2214,
+      "loss": 1.4921,
       "step": 500
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.24554608830364058,
-      "eval_f1_macro": 0.02628524046434494,
-      "eval_f1_micro": 0.24554608830364058,
-      "eval_f1_weighted": 0.09681356964210704,
-      "eval_loss": 2.1982107162475586,
-      "eval_macro_fpr": 0.06666666666666667,
-      "eval_macro_sensitivity": 0.06666666666666667,
-      "eval_macro_specificity": 0.9333333333333333,
-      "eval_precision": 0.06029288148121926,
-      "eval_precision_macro": 0.016369739220242704,
-      "eval_recall": 0.24554608830364058,
-      "eval_recall_macro": 0.06666666666666667,
-      "eval_runtime": 29.4439,
-      "eval_samples_per_second": 43.846,
-      "eval_steps_per_second": 5.502,
-      "eval_weighted_fpr": 0.17997043606799704,
-      "eval_weighted_sensitivity": 0.24554608830364058,
-      "eval_weighted_specificity": 0.7544539116963594,
+      "eval_accuracy": 0.6762199845081333,
+      "eval_f1_macro": 0.36762215914350266,
+      "eval_f1_micro": 0.6762199845081333,
+      "eval_f1_weighted": 0.6381553142782734,
+      "eval_loss": 1.0307424068450928,
+      "eval_macro_fpr": 0.03508857639652692,
+      "eval_macro_sensitivity": 0.4027183655220987,
+      "eval_macro_specificity": 0.9746655126680946,
+      "eval_precision": 0.6361951106730956,
+      "eval_precision_macro": 0.3745780279088578,
+      "eval_recall": 0.6762199845081333,
+      "eval_recall_macro": 0.4027183655220987,
+      "eval_runtime": 65.0146,
+      "eval_samples_per_second": 19.857,
+      "eval_steps_per_second": 2.492,
+      "eval_weighted_fpr": 0.03306962025316456,
+      "eval_weighted_sensitivity": 0.6762199845081333,
+      "eval_weighted_specificity": 0.9437627055132857,
       "step": 643
     },
     {
       "epoch": 1.56,
-      "grad_norm": 6.595059871673584,
+      "grad_norm": 11.052178382873535,
       "learning_rate": 4.483670295489891e-05,
-      "loss": 2.1462,
+      "loss": 0.9019,
       "step": 1000
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.3361735089078234,
-      "eval_f1_macro": 0.06552880630375763,
-      "eval_f1_micro": 0.3361735089078234,
-      "eval_f1_weighted": 0.2010480796875688,
-      "eval_loss": 1.9634978771209717,
-      "eval_macro_fpr": 0.09092506634165738,
-      "eval_macro_sensitivity": 0.11547020741602194,
-      "eval_macro_specificity": 0.9457066851291582,
-      "eval_precision": 0.1591168191834871,
-      "eval_precision_macro": 0.05308920125936356,
-      "eval_recall": 0.3361735089078234,
-      "eval_recall_macro": 0.11547020741602194,
-      "eval_runtime": 57.2389,
-      "eval_samples_per_second": 22.555,
-      "eval_steps_per_second": 2.83,
-      "eval_weighted_fpr": 0.12361171210154334,
-      "eval_weighted_sensitivity": 0.3361735089078234,
-      "eval_weighted_specificity": 0.84942676802955,
+      "eval_accuracy": 0.737412858249419,
+      "eval_f1_macro": 0.4767493392755858,
+      "eval_f1_micro": 0.737412858249419,
+      "eval_f1_weighted": 0.7197263095885438,
+      "eval_loss": 0.8135831952095032,
+      "eval_macro_fpr": 0.025093207469991694,
+      "eval_macro_sensitivity": 0.5148741493607558,
+      "eval_macro_specificity": 0.9804978566683986,
+      "eval_precision": 0.7158327481302543,
+      "eval_precision_macro": 0.4625056436667857,
+      "eval_recall": 0.737412858249419,
+      "eval_recall_macro": 0.5148741493607558,
+      "eval_runtime": 78.3833,
+      "eval_samples_per_second": 16.47,
+      "eval_steps_per_second": 2.067,
+      "eval_weighted_fpr": 0.024804273066510573,
+      "eval_weighted_sensitivity": 0.737412858249419,
+      "eval_weighted_specificity": 0.9700549917765612,
       "step": 1286
     },
     {
       "epoch": 2.33,
-      "grad_norm": 41.05898666381836,
-      "learning_rate": 4.224987039917055e-05,
-      "loss": 1.995,
+      "grad_norm": 16.875791549682617,
+      "learning_rate": 4.224468636599274e-05,
+      "loss": 0.7714,
       "step": 1500
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.3555383423702556,
-      "eval_f1_macro": 0.09312335075016234,
-      "eval_f1_micro": 0.3555383423702556,
-      "eval_f1_weighted": 0.2541975590745503,
-      "eval_loss": 1.932247281074524,
-      "eval_macro_fpr": 0.09537141294157879,
-      "eval_macro_sensitivity": 0.14351969127084646,
-      "eval_macro_specificity": 0.9497795632717652,
-      "eval_precision": 0.23412920314698393,
-      "eval_precision_macro": 0.09260367628400626,
-      "eval_recall": 0.3555383423702556,
-      "eval_recall_macro": 0.14351969127084646,
-      "eval_runtime": 59.8887,
-      "eval_samples_per_second": 21.557,
-      "eval_steps_per_second": 2.705,
-      "eval_weighted_fpr": 0.11463213006337834,
-      "eval_weighted_sensitivity": 0.3555383423702556,
-      "eval_weighted_specificity": 0.8911551067062222,
+      "eval_accuracy": 0.7621998450813323,
+      "eval_f1_macro": 0.5064301619528891,
+      "eval_f1_micro": 0.7621998450813323,
+      "eval_f1_weighted": 0.7440787580787277,
+      "eval_loss": 0.8493772149085999,
+      "eval_macro_fpr": 0.022439395825223803,
+      "eval_macro_sensitivity": 0.5353612181578932,
+      "eval_macro_specificity": 0.9822327546765126,
+      "eval_precision": 0.7385639759374297,
+      "eval_precision_macro": 0.505572707108006,
+      "eval_recall": 0.7621998450813323,
+      "eval_recall_macro": 0.5353612181578932,
+      "eval_runtime": 62.8149,
+      "eval_samples_per_second": 20.552,
+      "eval_steps_per_second": 2.579,
+      "eval_weighted_fpr": 0.02179933252858056,
+      "eval_weighted_sensitivity": 0.7621998450813323,
+      "eval_weighted_specificity": 0.9712914750663563,
       "step": 1929
     },
     {
       "epoch": 3.11,
-      "grad_norm": 4.287881851196289,
-      "learning_rate": 3.96630378434422e-05,
-      "loss": 2.0064,
+      "grad_norm": 59.48842239379883,
+      "learning_rate": 3.9652669777086575e-05,
+      "loss": 0.7018,
       "step": 2000
     },
     {
       "epoch": 3.89,
-      "grad_norm": 5.713494777679443,
-      "learning_rate": 3.7076205287713846e-05,
-      "loss": 1.8909,
+      "grad_norm": 1.6855363845825195,
+      "learning_rate": 3.706583722135822e-05,
+      "loss": 0.5351,
       "step": 2500
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.3787761425251743,
-      "eval_f1_macro": 0.14496386130680605,
-      "eval_f1_micro": 0.3787761425251743,
-      "eval_f1_weighted": 0.29773714484044667,
-      "eval_loss": 1.8874640464782715,
-      "eval_macro_fpr": 0.08706956494211866,
-      "eval_macro_sensitivity": 0.2191470222337379,
-      "eval_macro_specificity": 0.9525422186332462,
-      "eval_precision": 0.28749388015966415,
-      "eval_precision_macro": 0.12703293957809556,
-      "eval_recall": 0.3787761425251743,
-      "eval_recall_macro": 0.2191470222337379,
-      "eval_runtime": 69.4305,
-      "eval_samples_per_second": 18.594,
-      "eval_steps_per_second": 2.333,
-      "eval_weighted_fpr": 0.10486401673640168,
-      "eval_weighted_sensitivity": 0.3787761425251743,
-      "eval_weighted_specificity": 0.9093571369735185,
+      "eval_accuracy": 0.7645236250968241,
+      "eval_f1_macro": 0.544445143585482,
+      "eval_f1_micro": 0.7645236250968241,
+      "eval_f1_weighted": 0.7545581478665218,
+      "eval_loss": 0.9904704093933105,
+      "eval_macro_fpr": 0.022368675533492623,
+      "eval_macro_sensitivity": 0.5653277909563001,
+      "eval_macro_specificity": 0.9821237739107975,
+      "eval_precision": 0.7656722240785093,
+      "eval_precision_macro": 0.5495765557734748,
+      "eval_recall": 0.7645236250968241,
+      "eval_recall_macro": 0.5653277909563001,
+      "eval_runtime": 80.7211,
+      "eval_samples_per_second": 15.993,
+      "eval_steps_per_second": 2.007,
+      "eval_weighted_fpr": 0.021526695935419914,
+      "eval_weighted_sensitivity": 0.7645236250968241,
+      "eval_weighted_specificity": 0.9673329835651372,
       "step": 2572
     },
     {
       "epoch": 4.67,
-      "grad_norm": 18.91859245300293,
-      "learning_rate": 3.4484188698807675e-05,
-      "loss": 1.7621,
+      "grad_norm": 41.18514633178711,
+      "learning_rate": 3.447382063245205e-05,
+      "loss": 0.4521,
       "step": 3000
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.42835011618900076,
-      "eval_f1_macro": 0.1745447316852942,
-      "eval_f1_micro": 0.4283501161890007,
-      "eval_f1_weighted": 0.3593798293323463,
-      "eval_loss": 1.7168936729431152,
-      "eval_macro_fpr": 0.07431772683182448,
-      "eval_macro_sensitivity": 0.2270221309090987,
-      "eval_macro_specificity": 0.95583829077833,
-      "eval_precision": 0.374728097525222,
-      "eval_precision_macro": 0.18202607308331148,
-      "eval_recall": 0.42835011618900076,
-      "eval_recall_macro": 0.2270221309090987,
-      "eval_runtime": 58.7135,
-      "eval_samples_per_second": 21.988,
-      "eval_steps_per_second": 2.759,
-      "eval_weighted_fpr": 0.08702830188679245,
-      "eval_weighted_sensitivity": 0.42835011618900076,
-      "eval_weighted_specificity": 0.9092242454859507,
+      "eval_accuracy": 0.7900852052672347,
+      "eval_f1_macro": 0.6116525431724917,
+      "eval_f1_micro": 0.7900852052672347,
+      "eval_f1_weighted": 0.7794301250682801,
+      "eval_loss": 1.0259578227996826,
+      "eval_macro_fpr": 0.019043445988564375,
+      "eval_macro_sensitivity": 0.6315170743424566,
+      "eval_macro_specificity": 0.984305159354804,
+      "eval_precision": 0.7882093327742286,
+      "eval_precision_macro": 0.6420940275483907,
+      "eval_recall": 0.7900852052672347,
+      "eval_recall_macro": 0.6315170743424566,
+      "eval_runtime": 56.9964,
+      "eval_samples_per_second": 22.651,
+      "eval_steps_per_second": 2.842,
+      "eval_weighted_fpr": 0.018624149542986735,
+      "eval_weighted_sensitivity": 0.7900852052672347,
+      "eval_weighted_specificity": 0.9744921850548264,
       "step": 3215
     },
     {
       "epoch": 5.44,
-      "grad_norm": 7.408999919891357,
-      "learning_rate": 3.1892172109901504e-05,
-      "loss": 1.6254,
+      "grad_norm": 92.02596282958984,
+      "learning_rate": 3.1881804043545884e-05,
+      "loss": 0.3466,
       "step": 3500
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.4500387296669249,
-      "eval_f1_macro": 0.18588331301714964,
-      "eval_f1_micro": 0.4500387296669249,
-      "eval_f1_weighted": 0.39697994452590485,
-      "eval_loss": 1.8831233978271484,
-      "eval_macro_fpr": 0.07944022523045574,
-      "eval_macro_sensitivity": 0.2544695464493438,
-      "eval_macro_specificity": 0.9582711519830998,
-      "eval_precision": 0.3947542941359455,
-      "eval_precision_macro": 0.1877780146030796,
-      "eval_recall": 0.4500387296669249,
-      "eval_recall_macro": 0.2544695464493438,
-      "eval_runtime": 65.7966,
-      "eval_samples_per_second": 19.621,
-      "eval_steps_per_second": 2.462,
-      "eval_weighted_fpr": 0.08028041610131162,
-      "eval_weighted_sensitivity": 0.4500387296669249,
-      "eval_weighted_specificity": 0.92402855007957,
+      "eval_accuracy": 0.7869868319132456,
+      "eval_f1_macro": 0.6118575336738351,
+      "eval_f1_micro": 0.7869868319132455,
+      "eval_f1_weighted": 0.7889880056842663,
+      "eval_loss": 1.0384949445724487,
+      "eval_macro_fpr": 0.019220806323432427,
+      "eval_macro_sensitivity": 0.633518783056355,
+      "eval_macro_specificity": 0.984312476194111,
+      "eval_precision": 0.8134105657387156,
+      "eval_precision_macro": 0.6722188161248251,
+      "eval_recall": 0.7869868319132456,
+      "eval_recall_macro": 0.633518783056355,
+      "eval_runtime": 58.1615,
+      "eval_samples_per_second": 22.197,
+      "eval_steps_per_second": 2.785,
+      "eval_weighted_fpr": 0.018966825298296436,
+      "eval_weighted_sensitivity": 0.7869868319132456,
+      "eval_weighted_specificity": 0.9777003109984174,
       "step": 3858
     },
     {
       "epoch": 6.22,
-      "grad_norm": 8.139199256896973,
-      "learning_rate": 2.9300155520995337e-05,
-      "loss": 1.4996,
+      "grad_norm": 80.3629379272461,
+      "learning_rate": 2.9294971487817523e-05,
+      "loss": 0.3321,
       "step": 4000
     },
     {
       "epoch": 7.0,
-      "grad_norm": 40.22551727294922,
-      "learning_rate": 2.671332296526698e-05,
-      "loss": 1.4139,
+      "grad_norm": 0.3377046585083008,
+      "learning_rate": 2.6702954898911353e-05,
+      "loss": 0.2333,
       "step": 4500
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.46398140975987606,
-      "eval_f1_macro": 0.22845278849292336,
-      "eval_f1_micro": 0.4639814097598761,
-      "eval_f1_weighted": 0.4297793996222424,
-      "eval_loss": 1.662539005279541,
-      "eval_macro_fpr": 0.07171513504664777,
-      "eval_macro_sensitivity": 0.2949397340388095,
-      "eval_macro_specificity": 0.9601261188979237,
-      "eval_precision": 0.451602882474783,
-      "eval_precision_macro": 0.232765042661752,
-      "eval_recall": 0.46398140975987606,
-      "eval_recall_macro": 0.2949397340388095,
-      "eval_runtime": 60.1854,
-      "eval_samples_per_second": 21.45,
-      "eval_steps_per_second": 2.692,
-      "eval_weighted_fpr": 0.07622824410663141,
-      "eval_weighted_sensitivity": 0.46398140975987606,
-      "eval_weighted_specificity": 0.9379103737089811,
+      "eval_accuracy": 0.8117738187451587,
+      "eval_f1_macro": 0.665659747836216,
+      "eval_f1_micro": 0.8117738187451587,
+      "eval_f1_weighted": 0.8057624075524522,
+      "eval_loss": 1.1464879512786865,
+      "eval_macro_fpr": 0.01693769060060091,
+      "eval_macro_sensitivity": 0.6784424177464591,
+      "eval_macro_specificity": 0.9857572234742,
+      "eval_precision": 0.8064435087918365,
+      "eval_precision_macro": 0.6794791733053616,
+      "eval_recall": 0.8117738187451587,
+      "eval_recall_macro": 0.6784424177464591,
+      "eval_runtime": 54.6308,
+      "eval_samples_per_second": 23.631,
+      "eval_steps_per_second": 2.965,
+      "eval_weighted_fpr": 0.016292323164599398,
+      "eval_weighted_sensitivity": 0.8117738187451587,
+      "eval_weighted_specificity": 0.9745845333678407,
       "step": 4501
     },
     {
       "epoch": 7.78,
-      "grad_norm": 5.4409565925598145,
-      "learning_rate": 2.412130637636081e-05,
-      "loss": 1.2681,
+      "grad_norm": 0.19802448153495789,
+      "learning_rate": 2.4110938310005186e-05,
+      "loss": 0.1658,
       "step": 5000
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.5383423702556158,
-      "eval_f1_macro": 0.2703391292878622,
-      "eval_f1_micro": 0.5383423702556158,
-      "eval_f1_weighted": 0.48876940771692845,
-      "eval_loss": 1.4461586475372314,
-      "eval_macro_fpr": 0.056020043825060734,
-      "eval_macro_sensitivity": 0.3381295733874737,
-      "eval_macro_specificity": 0.9655641856187219,
-      "eval_precision": 0.5269280210096271,
-      "eval_precision_macro": 0.2765942082966614,
-      "eval_recall": 0.5383423702556158,
-      "eval_recall_macro": 0.3381295733874737,
-      "eval_runtime": 64.8556,
-      "eval_samples_per_second": 19.906,
-      "eval_steps_per_second": 2.498,
-      "eval_weighted_fpr": 0.057718380786364515,
-      "eval_weighted_sensitivity": 0.5383423702556158,
-      "eval_weighted_specificity": 0.9451204140252133,
+      "eval_accuracy": 0.814872192099148,
+      "eval_f1_macro": 0.7328612504365998,
+      "eval_f1_micro": 0.814872192099148,
+      "eval_f1_weighted": 0.8120704640541792,
+      "eval_loss": 1.241939902305603,
+      "eval_macro_fpr": 0.016489901852411035,
+      "eval_macro_sensitivity": 0.7440371076362254,
+      "eval_macro_specificity": 0.9861286253461167,
+      "eval_precision": 0.8148683401292235,
+      "eval_precision_macro": 0.7318970148814324,
+      "eval_recall": 0.814872192099148,
+      "eval_recall_macro": 0.7440371076362254,
+      "eval_runtime": 57.719,
+      "eval_samples_per_second": 22.367,
+      "eval_steps_per_second": 2.807,
+      "eval_weighted_fpr": 0.015968463954032203,
+      "eval_weighted_sensitivity": 0.814872192099148,
+      "eval_weighted_specificity": 0.9770571880926026,
       "step": 5144
     }
   ],

training_checkpoints/checkpoint-5144/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6da235c4e1e89b590bb6c4dd543edc36c67b04a23baad1858c6c64a331ed4bd1
+oid sha256:beb8f128167a5d05587c32e6e3bff806f690d3ab0996792b1d08193bcb3299f2
 size 4920