Training in progress, epoch 2, checkpoint

checkpoint-162/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:85c309638bb0f97df3e348cd3d4f6306e613c8b0554c335c4be0100bc3aab675
+oid sha256:d39588ab3313e3eee5296f0b2981d8c45c5927e90aa4e8b7a488be00d10bd2ad
 size 592330980

checkpoint-162/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eacb6741bdf4fc0cf7e464e83b2ae6c441ad298b06a67c1fe8f9f691d2d9d761
+oid sha256:4e0f8250b3a7bd66364d28f1dfa6b2eca37140b754887eef5bf0692d5b9e39f9
 size 1014670074

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": 0.28716403245925903,
+  "best_model_checkpoint": "longformer-sep_tok_full_labels/checkpoint-162",
   "epoch": 2.0,
   "eval_steps": 500,
   "global_step": 162,
@@ -11,134 +11,134 @@
     {
       "epoch": 1.0,
       "eval_B-Claim": {
-        "f1-score": 0.37160751565762,
-        "precision": 0.42788461538461536,
-        "recall": 0.3284132841328413,
-        "support": 271.0
+        "f1-score": 0.21800947867298578,
+        "precision": 0.3333333333333333,
+        "recall": 0.1619718309859155,
+        "support": 284.0
       },
       "eval_B-MajorClaim": {
-        "f1-score": 0.3578947368421052,
-        "precision": 0.6666666666666666,
-        "recall": 0.2446043165467626,
-        "support": 139.0
+        "f1-score": 0.0,
+        "precision": 0.0,
+        "recall": 0.0,
+        "support": 141.0
       },
       "eval_B-Premise": {
-        "f1-score": 0.8640915593705293,
-        "precision": 0.7895424836601307,
-        "recall": 0.9541864139020537,
-        "support": 633.0
+        "f1-score": 0.8305489260143198,
+        "precision": 0.71900826446281,
+        "recall": 0.9830508474576272,
+        "support": 708.0
       },
       "eval_I-Claim": {
-        "f1-score": 0.5003402749421533,
-        "precision": 0.5493126120741183,
-        "recall": 0.4593851537115721,
-        "support": 4001.0
+        "f1-score": 0.5590492696211935,
+        "precision": 0.5643589102724319,
+        "recall": 0.5538386068187393,
+        "support": 4077.0
       },
       "eval_I-MajorClaim": {
-        "f1-score": 0.7718093699515347,
-        "precision": 0.6502211636611093,
-        "recall": 0.9493293591654247,
-        "support": 2013.0
+        "f1-score": 0.7399038461538461,
+        "precision": 0.7205056179775281,
+        "recall": 0.7603754940711462,
+        "support": 2024.0
       },
       "eval_I-Premise": {
-        "f1-score": 0.875016720916752,
-        "precision": 0.8846812731043188,
-        "recall": 0.865561044460127,
-        "support": 11336.0
+        "f1-score": 0.8867229093291834,
+        "precision": 0.8883965205974068,
+        "recall": 0.8850555918901243,
+        "support": 12232.0
       },
       "eval_O": {
-        "f1-score": 0.9992483530087988,
-        "precision": 0.9995577178239717,
-        "recall": 0.998939179632249,
-        "support": 11312.0
+        "f1-score": 0.9976548035383666,
+        "precision": 0.9961383616794018,
+        "recall": 0.9991758694577221,
+        "support": 12134.0
       },
-      "eval_accuracy": 0.8614038040733883,
-      "eval_loss": 0.31713685393333435,
+      "eval_accuracy": 0.8699050632911393,
+      "eval_loss": 0.32671716809272766,
       "eval_macro avg": {
-        "f1-score": 0.6771440758127848,
-        "precision": 0.7096952189107044,
-        "recall": 0.685774107364433,
-        "support": 29705.0
-      },
-      "eval_runtime": 4.8338,
-      "eval_samples_per_second": 16.55,
-      "eval_steps_per_second": 2.069,
+        "f1-score": 0.6045556047614136,
+        "precision": 0.6031058583318446,
+        "recall": 0.6204954629544678,
+        "support": 31600.0
+      },
+      "eval_runtime": 6.234,
+      "eval_samples_per_second": 12.833,
+      "eval_steps_per_second": 1.604,
       "eval_weighted avg": {
-        "f1-score": 0.8576207231627551,
-        "precision": 0.8601529227027923,
-        "recall": 0.8614038040733883,
-        "support": 29705.0
+        "f1-score": 0.8664142595402331,
+        "precision": 0.8644597558999653,
+        "recall": 0.8699050632911393,
+        "support": 31600.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_B-Claim": {
-        "f1-score": 0.4708624708624709,
-        "precision": 0.6392405063291139,
-        "recall": 0.3726937269372694,
-        "support": 271.0
+        "f1-score": 0.5352697095435685,
+        "precision": 0.6515151515151515,
+        "recall": 0.45422535211267606,
+        "support": 284.0
       },
       "eval_B-MajorClaim": {
-        "f1-score": 0.796875,
-        "precision": 0.8717948717948718,
-        "recall": 0.7338129496402878,
-        "support": 139.0
+        "f1-score": 0.7766990291262136,
+        "precision": 0.7142857142857143,
+        "recall": 0.851063829787234,
+        "support": 141.0
       },
       "eval_B-Premise": {
-        "f1-score": 0.8736616702355461,
-        "precision": 0.796875,
-        "recall": 0.966824644549763,
-        "support": 633.0
+        "f1-score": 0.8802721088435375,
+        "precision": 0.8490813648293963,
+        "recall": 0.9138418079096046,
+        "support": 708.0
       },
       "eval_I-Claim": {
-        "f1-score": 0.5100589925881107,
-        "precision": 0.6459770114942529,
-        "recall": 0.4213946513371657,
-        "support": 4001.0
+        "f1-score": 0.5747557221255521,
+        "precision": 0.6325869180907484,
+        "recall": 0.5266127054206524,
+        "support": 4077.0
       },
       "eval_I-MajorClaim": {
-        "f1-score": 0.8401387776888176,
-        "precision": 0.9077277970011534,
-        "recall": 0.7819175360158966,
-        "support": 2013.0
+        "f1-score": 0.7833260963960051,
+        "precision": 0.6986831913245546,
+        "recall": 0.8913043478260869,
+        "support": 2024.0
       },
       "eval_I-Premise": {
-        "f1-score": 0.8912891699864469,
-        "precision": 0.8338584492430646,
-        "recall": 0.9572159491884262,
-        "support": 11336.0
+        "f1-score": 0.8995515695067264,
+        "precision": 0.8971377459749553,
+        "recall": 0.9019784172661871,
+        "support": 12232.0
       },
       "eval_O": {
-        "f1-score": 0.9996904982977407,
-        "precision": 1.0,
-        "recall": 0.9993811881188119,
-        "support": 11312.0
+        "f1-score": 0.9971231300345225,
+        "precision": 0.9945072962780783,
+        "recall": 0.9997527608373167,
+        "support": 12134.0
       },
-      "eval_accuracy": 0.8830499915839084,
-      "eval_loss": 0.2966194748878479,
+      "eval_accuracy": 0.8864240506329114,
+      "eval_loss": 0.28716403245925903,
       "eval_macro avg": {
-        "f1-score": 0.7689395113798762,
-        "precision": 0.8136390908374939,
-        "recall": 0.7476058065410885,
-        "support": 29705.0
-      },
-      "eval_runtime": 4.8625,
-      "eval_samples_per_second": 16.452,
-      "eval_steps_per_second": 2.057,
+        "f1-score": 0.7781424807965893,
+        "precision": 0.7768281974712284,
+        "recall": 0.791254174451394,
+        "support": 31600.0
+      },
+      "eval_runtime": 6.2816,
+      "eval_samples_per_second": 12.736,
+      "eval_steps_per_second": 1.592,
       "eval_weighted avg": {
-        "f1-score": 0.8731020208182413,
-        "precision": 0.874440834821272,
-        "recall": 0.8830499915839084,
-        "support": 29705.0
+        "f1-score": 0.8834146129726252,
+        "precision": 0.8835831101628018,
+        "recall": 0.8864240506329114,
+        "support": 31600.0
       },
       "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": 287581625436000.0,
   "train_batch_size": 4,

checkpoint-162/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e3aa7d1dcb1ecb96052506e9849f1d3fa19e688784f8ea37f784dfaca9546785
+oid sha256:4a6eed06a4e3317966121cd80a8275e0b82aa59ebaa0185c3b4676ff92bfed33
 size 4920