Training in progress, epoch 2, checkpoint

checkpoint-162/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43c99e6a866cd6da18b8d9e9189c03d600a154fd5b5f19a93d91b0364eaa4952
+oid sha256:21666e7da60d6216240ddc1704956854ac9c3a07aefceddba6d6cf93516993ec
 size 592324828

checkpoint-162/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50861c73a583a913a44587c4fa457f2a6245f0bd1f3fa030a197744b62add8d4
+oid sha256:e6e68689a69e7cb07bfd1818731e22aa883af728766a36e31c39ff0e9078abe7
 size 1014657786

checkpoint-162/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0bd8264931534616018335b535732049961e4647585455e22b74dcb3e6034c63
+oid sha256:46048b97dd03acf4cb6c82fc924d60322fb8fb6d0b9ecb6d39c987081744ecab
 size 1064

checkpoint-162/trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
+  "best_metric": 0.5075464248657227,
+  "best_model_checkpoint": "longformer-simple/checkpoint-162",
   "epoch": 2.0,
   "eval_steps": 500,
   "global_step": 162,
@@ -11,98 +11,98 @@
     {
       "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
+        "f1-score": 0.6889408441875129,
+        "precision": 0.708455608235389,
+        "recall": 0.6772234834145282,
+        "support": 29334.0
       },
-      "eval_runtime": 4.7759,
-      "eval_samples_per_second": 16.751,
-      "eval_steps_per_second": 2.094,
+      "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
+        "f1-score": 0.7407811680534537,
+        "precision": 0.7243501600962077,
+        "recall": 0.7717868927348766,
+        "support": 29334.0
       },
-      "eval_runtime": 4.7439,
-      "eval_samples_per_second": 16.864,
-      "eval_steps_per_second": 2.108,
+      "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
     }
   ],
   "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": 287577465453600.0,
   "train_batch_size": 4,

checkpoint-162/training_args.bin

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