Training in progress, epoch 5, checkpoint

checkpoint-405/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:26c3c53af1e11b0efb705437f67aa0454d72b34e66fa350958d60d244368a055
+oid sha256:bdeb7856ca9cdb988d1d25b12eb7d28e9cd92dd844f901d7672a0714474bb78d
 size 592318676

checkpoint-405/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8e99e8f2fb936a72c4972883d1c0d5205081d7b6e728a42df62a4c911282d59f
+oid sha256:9a74676f8c31cc3f20aa46b338d590d7f9ea2577e290b63ea27c69c521a2544e
 size 1014645498

checkpoint-405/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c73756f22c58a684dfffbbc966a96d1e3e061222bcc1371ddea16fc12be01725
+oid sha256:112155cc3e2788408d703d89eaa1d2c9a92b4cf9f31734e313b8f50755aa4e0f
 size 1064

checkpoint-405/trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
+  "best_metric": 0.2252955436706543,
+  "best_model_checkpoint": "longformer-spans/checkpoint-162",
   "epoch": 5.0,
   "eval_steps": 500,
   "global_step": 405,
@@ -11,203 +11,203 @@
     {
       "epoch": 1.0,
       "eval_B": {
-        "f1-score": 0.8331822302810515,
-        "precision": 0.7901977644024075,
-        "recall": 0.8811121764141898,
-        "support": 1043.0
+        "f1-score": 0.816,
+        "precision": 0.7461594732991953,
+        "recall": 0.9002647837599294,
+        "support": 1133.0
       },
       "eval_I": {
-        "f1-score": 0.9471294753175105,
-        "precision": 0.9443616777446711,
-        "recall": 0.9499135446685879,
-        "support": 17350.0
+        "f1-score": 0.9321376763813793,
+        "precision": 0.9024103768767235,
+        "recall": 0.9638902525500463,
+        "support": 18333.0
       },
       "eval_O": {
-        "f1-score": 0.8992868897421833,
-        "precision": 0.9103731674811195,
-        "recall": 0.8884673748103187,
-        "support": 9226.0
+        "f1-score": 0.8560784528570645,
+        "precision": 0.931782945736434,
+        "recall": 0.7917511147142278,
+        "support": 9868.0
       },
-      "eval_accuracy": 0.9267895289474637,
-      "eval_loss": 0.20478281378746033,
+      "eval_accuracy": 0.9035249198881844,
+      "eval_loss": 0.2619660496711731,
       "eval_macro avg": {
-        "f1-score": 0.8931995317802484,
-        "precision": 0.8816442032093993,
-        "recall": 0.9064976986310321,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.6128,
-      "eval_samples_per_second": 17.343,
-      "eval_steps_per_second": 2.168,
+        "f1-score": 0.8680720430794812,
+        "precision": 0.860117598637451,
+        "recall": 0.8853020503414012,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.5032,
+      "eval_samples_per_second": 14.537,
+      "eval_steps_per_second": 1.817,
       "eval_weighted avg": {
-        "f1-score": 0.9268447919078652,
-        "precision": 0.9271861479533134,
-        "recall": 0.9267895289474637,
-        "support": 27619.0
+        "f1-score": 0.9020655278480035,
+        "precision": 0.9062562975065145,
+        "recall": 0.9035249198881844,
+        "support": 29334.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_B": {
-        "f1-score": 0.8682547399124939,
-        "precision": 0.8806706114398422,
-        "recall": 0.8561840843720039,
-        "support": 1043.0
+        "f1-score": 0.8512256973795435,
+        "precision": 0.8167072181670721,
+        "recall": 0.8887908208296558,
+        "support": 1133.0
       },
       "eval_I": {
-        "f1-score": 0.9590983771942466,
-        "precision": 0.9592919333448654,
-        "recall": 0.9589048991354466,
-        "support": 17350.0
+        "f1-score": 0.9416781438711729,
+        "precision": 0.9152551099212274,
+        "recall": 0.9696721758577429,
+        "support": 18333.0
       },
       "eval_O": {
-        "f1-score": 0.9226525313717006,
-        "precision": 0.9208594256100194,
-        "recall": 0.9244526338608281,
-        "support": 9226.0
+        "f1-score": 0.8778173190984578,
+        "precision": 0.9380041484212952,
+        "recall": 0.8248885285772193,
+        "support": 9868.0
       },
-      "eval_accuracy": 0.9435171439950758,
-      "eval_loss": 0.16574698686599731,
+      "eval_accuracy": 0.9178427763005387,
+      "eval_loss": 0.2252955436706543,
       "eval_macro avg": {
-        "f1-score": 0.9166685494928136,
-        "precision": 0.9202739901315756,
-        "recall": 0.9131805391227594,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.6238,
-      "eval_samples_per_second": 17.302,
-      "eval_steps_per_second": 2.163,
+        "f1-score": 0.8902403867830581,
+        "precision": 0.8899888255031981,
+        "recall": 0.8944505084215394,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.4972,
+      "eval_samples_per_second": 14.553,
+      "eval_steps_per_second": 1.819,
       "eval_weighted avg": {
-        "f1-score": 0.9434932036816766,
-        "precision": 0.9434846863370583,
-        "recall": 0.9435171439950758,
-        "support": 27619.0
+        "f1-score": 0.9167016237671239,
+        "precision": 0.9191015935430046,
+        "recall": 0.9178427763005387,
+        "support": 29334.0
       },
       "step": 162
     },
     {
       "epoch": 3.0,
       "eval_B": {
-        "f1-score": 0.8791725434884815,
-        "precision": 0.8625461254612546,
-        "recall": 0.8964525407478428,
-        "support": 1043.0
+        "f1-score": 0.8531120331950207,
+        "precision": 0.8050117462803446,
+        "recall": 0.9073256840247131,
+        "support": 1133.0
       },
       "eval_I": {
-        "f1-score": 0.9588372159825352,
-        "precision": 0.9557349825345015,
-        "recall": 0.9619596541786744,
-        "support": 17350.0
+        "f1-score": 0.9469915571230095,
+        "precision": 0.9280963603037444,
+        "recall": 0.9666721213112965,
+        "support": 18333.0
       },
       "eval_O": {
-        "f1-score": 0.9204284621270084,
-        "precision": 0.9282407407407407,
-        "recall": 0.9127465857359636,
-        "support": 9226.0
+        "f1-score": 0.8903876792352629,
+        "precision": 0.9353938852934612,
+        "recall": 0.8495135792460479,
+        "support": 9868.0
       },
-      "eval_accuracy": 0.9430464535283681,
-      "eval_loss": 0.15394224226474762,
+      "eval_accuracy": 0.9249676143724006,
+      "eval_loss": 0.22786012291908264,
       "eval_macro avg": {
-        "f1-score": 0.9194794071993417,
-        "precision": 0.9155072829121655,
-        "recall": 0.9237195935541602,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.6233,
-      "eval_samples_per_second": 17.303,
-      "eval_steps_per_second": 2.163,
+        "f1-score": 0.8968304231844311,
+        "precision": 0.8895006639591835,
+        "recall": 0.9078371281940192,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.5036,
+      "eval_samples_per_second": 14.536,
+      "eval_steps_per_second": 1.817,
       "eval_weighted avg": {
-        "f1-score": 0.9429985029052194,
-        "precision": 0.9430314866542512,
-        "recall": 0.9430464535283681,
-        "support": 27619.0
+        "f1-score": 0.9243239165827936,
+        "precision": 0.9257972230878861,
+        "recall": 0.9249676143724006,
+        "support": 29334.0
       },
       "step": 243
     },
     {
       "epoch": 4.0,
       "eval_B": {
-        "f1-score": 0.8800746616892207,
-        "precision": 0.8572727272727273,
-        "recall": 0.9041227229146692,
-        "support": 1043.0
+        "f1-score": 0.8567807351077312,
+        "precision": 0.8217179902755267,
+        "recall": 0.8949691085613416,
+        "support": 1133.0
       },
       "eval_I": {
-        "f1-score": 0.9537844036697248,
-        "precision": 0.9488876212207644,
-        "recall": 0.9587319884726225,
-        "support": 17350.0
+        "f1-score": 0.9472597903427299,
+        "precision": 0.9432635621180161,
+        "recall": 0.9512900234549719,
+        "support": 18333.0
       },
       "eval_O": {
-        "f1-score": 0.9100192149327477,
-        "precision": 0.9220157970853265,
-        "recall": 0.8983308042488619,
-        "support": 9226.0
+        "f1-score": 0.8979927100980543,
+        "precision": 0.9099989595255437,
+        "recall": 0.8862991487636805,
+        "support": 9868.0
       },
-      "eval_accuracy": 0.9364929939534379,
-      "eval_loss": 0.19801080226898193,
+      "eval_accuracy": 0.9272516533715143,
+      "eval_loss": 0.23897655308246613,
       "eval_macro avg": {
-        "f1-score": 0.9146260934305644,
-        "precision": 0.9093920485262728,
-        "recall": 0.9203951718787179,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.6132,
-      "eval_samples_per_second": 17.342,
-      "eval_steps_per_second": 2.168,
+        "f1-score": 0.9006777451828384,
+        "precision": 0.8916601706396955,
+        "recall": 0.910852760259998,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.4904,
+      "eval_samples_per_second": 14.571,
+      "eval_steps_per_second": 1.821,
       "eval_weighted avg": {
-        "f1-score": 0.9363812792925563,
-        "precision": 0.9364514800186444,
-        "recall": 0.9364929939534379,
-        "support": 27619.0
+        "f1-score": 0.9271915992526735,
+        "precision": 0.9273787107073643,
+        "recall": 0.9272516533715143,
+        "support": 29334.0
       },
       "step": 324
     },
     {
       "epoch": 5.0,
       "eval_B": {
-        "f1-score": 0.883877159309021,
-        "precision": 0.8847262247838616,
-        "recall": 0.8830297219558965,
-        "support": 1043.0
+        "f1-score": 0.8624407072013798,
+        "precision": 0.8431703204047217,
+        "recall": 0.8826125330979699,
+        "support": 1133.0
       },
       "eval_I": {
-        "f1-score": 0.9591500482927107,
-        "precision": 0.9456643513331839,
-        "recall": 0.9730259365994236,
-        "support": 17350.0
+        "f1-score": 0.948190701170407,
+        "precision": 0.9335059992600032,
+        "recall": 0.9633447880870561,
+        "support": 18333.0
       },
       "eval_O": {
-        "f1-score": 0.9191176470588236,
-        "precision": 0.9454503781801513,
-        "recall": 0.8942120095382614,
-        "support": 9226.0
+        "f1-score": 0.8955333298423835,
+        "precision": 0.9265359193845487,
+        "recall": 0.8665383056343737,
+        "support": 9868.0
       },
-      "eval_accuracy": 0.9432999022412107,
-      "eval_loss": 0.20074255764484406,
+      "eval_accuracy": 0.9276607349832958,
+      "eval_loss": 0.25390708446502686,
       "eval_macro avg": {
-        "f1-score": 0.9207149515535183,
-        "precision": 0.9252803180990656,
-        "recall": 0.9167558893645271,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.6092,
-      "eval_samples_per_second": 17.356,
-      "eval_steps_per_second": 2.17,
+        "f1-score": 0.9020549127380568,
+        "precision": 0.9010707463497579,
+        "recall": 0.9041652089397999,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.4919,
+      "eval_samples_per_second": 14.567,
+      "eval_steps_per_second": 1.821,
       "eval_weighted avg": {
-        "f1-score": 0.9429348139614955,
-        "precision": 0.9432916158141273,
-        "recall": 0.9432999022412107,
-        "support": 27619.0
+        "f1-score": 0.9271646670996412,
+        "precision": 0.9276721180179627,
+        "recall": 0.9276607349832958,
+        "support": 29334.0
       },
       "step": 405
     }
   ],
   "logging_steps": 500,
-  "max_steps": 4050,
+  "max_steps": 1620,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 50,
+  "num_train_epochs": 20,
   "save_steps": 500,
   "total_flos": 718933263678000.0,
   "train_batch_size": 4,

checkpoint-405/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:821ca433db9ddb761291ab61fc06d651ad248cd9d560cdb25f7f905e5aaff14e
+oid sha256:1094850123635c13193a80398a7bb1e351dd871416d3ce265431b259bf603290
 size 4856