Training in progress, epoch 4, checkpoint

checkpoint-324/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:67bba726d5b93a5953d8c322309502f56f3d294303980872106da73402842fcd
+oid sha256:735c491c209b68773e3343ca55476d1523ee7c593cb93b690b2cff220c6a1920
 size 592324828

checkpoint-324/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad7419a2296c7b12859acd3f69a8a13a0da0939bc756ffeb1fd315ac1573d369
+oid sha256:83d61128a0f1db593229398fe9627c91ccc99dc947eb5eb400d67d433e3f40f9
 size 1014657786

checkpoint-324/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f95cf6428a521f225a19b2f71296e812e7a5d112c98910331720e5ff6f1eff62
+oid sha256:bd6dfe6953b9cbab86b93ee2d3f16f1f3460960873f680d5a6b7263a2c839a40
 size 1064

checkpoint-324/trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
+  "best_metric": 0.4725272059440613,
+  "best_model_checkpoint": "longformer-simple/checkpoint-324",
   "epoch": 4.0,
   "eval_steps": 500,
   "global_step": 324,
@@ -11,188 +11,188 @@
     {
       "epoch": 1.0,
       "eval_Claim": {
-        "f1-score": 0.41092129347162903,
-        "precision": 0.5640703517587939,
-        "recall": 0.32317658349328215,
-        "support": 4168.0
+        "f1-score": 0.46242900830056793,
+        "precision": 0.43257049448304047,
+        "recall": 0.496715157203191,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.6523102655237298,
-        "precision": 0.5929304446978335,
-        "recall": 0.724907063197026,
-        "support": 2152.0
+        "f1-score": 0.5626477541371159,
+        "precision": 0.6522533495736906,
+        "recall": 0.49468822170900695,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.8834432924913603,
-        "precision": 0.8545380875202593,
-        "recall": 0.9143724257533059,
-        "support": 9226.0
+        "f1-score": 0.8691687871077185,
+        "precision": 0.911293908403735,
+        "recall": 0.8307661126874747,
+        "support": 9868.0
       },
       "eval_Premise": {
-        "f1-score": 0.8664166767469054,
-        "precision": 0.8441231929604023,
-        "recall": 0.8899196554294707,
-        "support": 12073.0
+        "f1-score": 0.8615178272046495,
+        "precision": 0.8377046804810897,
+        "recall": 0.88672444205844,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.7997031029363844,
-      "eval_loss": 0.5104541182518005,
+      "eval_accuracy": 0.7823004022635849,
+      "eval_loss": 0.5488611459732056,
       "eval_macro avg": {
-        "f1-score": 0.7032728820584061,
-        "precision": 0.7139155192343223,
-        "recall": 0.7130939319682712,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.7759,
-      "eval_samples_per_second": 16.751,
-      "eval_steps_per_second": 2.094,
+        "f1-score": 0.6889408441875129,
+        "precision": 0.708455608235389,
+        "recall": 0.6772234834145282,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.8698,
+      "eval_samples_per_second": 13.629,
+      "eval_steps_per_second": 1.704,
       "eval_weighted avg": {
-        "f1-score": 0.7866826459135919,
-        "precision": 0.7857670171690952,
-        "recall": 0.7997031029363844,
-        "support": 27619.0
+        "f1-score": 0.7840489998358311,
+        "precision": 0.7899101236188295,
+        "recall": 0.7823004022635849,
+        "support": 29334.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_Claim": {
-        "f1-score": 0.5200312174817897,
-        "precision": 0.5678977272727272,
-        "recall": 0.47960652591170827,
-        "support": 4168.0
+        "f1-score": 0.5457327068079991,
+        "precision": 0.5107383923092657,
+        "recall": 0.5858751759737213,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7526302911671151,
-        "precision": 0.7948320413436692,
-        "recall": 0.7146840148698885,
-        "support": 2152.0
+        "f1-score": 0.6722085773663329,
+        "precision": 0.5687340153452686,
+        "recall": 0.8217090069284064,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.9116096466629278,
-        "precision": 0.944560669456067,
-        "recall": 0.8808801213960546,
-        "support": 9226.0
+        "f1-score": 0.8773589977955805,
+        "precision": 0.9344863131370977,
+        "recall": 0.8268139440616133,
+        "support": 9868.0
       },
       "eval_Premise": {
-        "f1-score": 0.8738735224125151,
-        "precision": 0.8259587020648967,
-        "recall": 0.9276898865236478,
-        "support": 12073.0
+        "f1-score": 0.8678243902439025,
+        "precision": 0.8834419195931988,
+        "recall": 0.852749443975765,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.8278359100619139,
-      "eval_loss": 0.44692954421043396,
+      "eval_accuracy": 0.8029590236585532,
+      "eval_loss": 0.5075464248657227,
       "eval_macro avg": {
-        "f1-score": 0.7645361694310869,
-        "precision": 0.7833122850343401,
-        "recall": 0.7507151371753248,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.7439,
-      "eval_samples_per_second": 16.864,
-      "eval_steps_per_second": 2.108,
+        "f1-score": 0.7407811680534537,
+        "precision": 0.7243501600962077,
+        "recall": 0.7717868927348766,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.9473,
+      "eval_samples_per_second": 13.451,
+      "eval_steps_per_second": 1.681,
       "eval_weighted avg": {
-        "f1-score": 0.8236335905447046,
-        "precision": 0.8242076985653165,
-        "recall": 0.8278359100619139,
-        "support": 27619.0
+        "f1-score": 0.8097969994222006,
+        "precision": 0.8232353684753937,
+        "recall": 0.8029590236585532,
+        "support": 29334.0
       },
       "step": 162
     },
     {
       "epoch": 3.0,
       "eval_Claim": {
-        "f1-score": 0.6218357913273168,
-        "precision": 0.5744204961366409,
-        "recall": 0.6777831094049904,
-        "support": 4168.0
+        "f1-score": 0.4978109708987896,
+        "precision": 0.5516552511415526,
+        "recall": 0.45354293758798686,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7799868909766223,
-        "precision": 0.7360824742268042,
-        "recall": 0.8294609665427509,
-        "support": 2152.0
+        "f1-score": 0.7357876283496118,
+        "precision": 0.8036105032822757,
+        "recall": 0.6785219399538106,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.9093637121422617,
-        "precision": 0.9330596419204014,
-        "recall": 0.8868415347929763,
-        "support": 9226.0
+        "f1-score": 0.8891011355194736,
+        "precision": 0.9331699710403207,
+        "recall": 0.8490068909606809,
+        "support": 9868.0
       },
       "eval_Premise": {
-        "f1-score": 0.876420695504665,
-        "precision": 0.89797514556357,
-        "recall": 0.855876749772219,
-        "support": 12073.0
+        "f1-score": 0.8770979581655561,
+        "precision": 0.8191560170394037,
+        "recall": 0.9438607255157604,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.8372859263550454,
-      "eval_loss": 0.40934404730796814,
+      "eval_accuracy": 0.821129065248517,
+      "eval_loss": 0.5219377279281616,
       "eval_macro avg": {
-        "f1-score": 0.7969017724877165,
-        "precision": 0.7853844394618541,
-        "recall": 0.8124905901282342,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.7941,
-      "eval_samples_per_second": 16.687,
-      "eval_steps_per_second": 2.086,
+        "f1-score": 0.7499494232333578,
+        "precision": 0.7768979356258882,
+        "recall": 0.7312331235045597,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.8364,
+      "eval_samples_per_second": 13.707,
+      "eval_steps_per_second": 1.713,
       "eval_weighted avg": {
-        "f1-score": 0.8414917278933443,
-        "precision": 0.8482528803063183,
-        "recall": 0.8372859263550454,
-        "support": 27619.0
+        "f1-score": 0.815598992812927,
+        "precision": 0.8174973750724106,
+        "recall": 0.821129065248517,
+        "support": 29334.0
       },
       "step": 243
     },
     {
       "epoch": 4.0,
       "eval_Claim": {
-        "f1-score": 0.608644284729743,
-        "precision": 0.516112873601603,
-        "recall": 0.7416026871401151,
-        "support": 4168.0
+        "f1-score": 0.5811481218993623,
+        "precision": 0.5851569933396765,
+        "recall": 0.5771938057250118,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.788337924701561,
-        "precision": 0.7790381125226861,
-        "recall": 0.7978624535315985,
-        "support": 2152.0
+        "f1-score": 0.761714412613813,
+        "precision": 0.7335329341317365,
+        "recall": 0.792147806004619,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.9068679925629612,
-        "precision": 0.9442684500762643,
-        "recall": 0.8723173639713853,
-        "support": 9226.0
+        "f1-score": 0.8977645611156685,
+        "precision": 0.9085720215857203,
+        "recall": 0.8872111876773409,
+        "support": 9868.0
       },
       "eval_Premise": {
-        "f1-score": 0.8528029247910863,
-        "precision": 0.8985600293497202,
-        "recall": 0.8114801623457302,
-        "support": 12073.0
+        "f1-score": 0.8891009734682191,
+        "precision": 0.8851474612344178,
+        "recall": 0.8930899608865711,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.8201962417176581,
-      "eval_loss": 0.5160673260688782,
+      "eval_accuracy": 0.8377650507943001,
+      "eval_loss": 0.4725272059440613,
       "eval_macro avg": {
-        "f1-score": 0.789163281696338,
-        "precision": 0.7844948663875684,
-        "recall": 0.8058156667472073,
-        "support": 27619.0
-      },
-      "eval_runtime": 4.7669,
-      "eval_samples_per_second": 16.782,
-      "eval_steps_per_second": 2.098,
+        "f1-score": 0.7824320172742658,
+        "precision": 0.7781023525728877,
+        "recall": 0.7874106900733857,
+        "support": 29334.0
+      },
+      "eval_runtime": 5.8763,
+      "eval_samples_per_second": 13.614,
+      "eval_steps_per_second": 1.702,
       "eval_weighted avg": {
-        "f1-score": 0.8289940404467937,
-        "precision": 0.8468005514342695,
-        "recall": 0.8201962417176581,
-        "support": 27619.0
+        "f1-score": 0.8378705011585708,
+        "precision": 0.8382513248807655,
+        "recall": 0.8377650507943001,
+        "support": 29334.0
       },
       "step": 324
     }
   ],
   "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": 575154930907200.0,
   "train_batch_size": 4,

checkpoint-324/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e651c478af6098730eda01c59bbe4834fbad7ed476e7ec6e4265141ee6647e9
+oid sha256:91855299d72d2ceb66b00b0c173ef2e96bd5d782fa49a7debc2d6f3b7cbd2cd4
 size 4856