Training in progress, epoch 4, checkpoint

checkpoint-324/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:27f47dcdb29a8d1c80c0a340b99e024b80cd6acb825831f65ed0576f5361034c
+oid sha256:6da327faf37cb0c48058ebbe215843503ee19d048023a20529214c8e0142e4a5
 size 592324828

checkpoint-324/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fdc3cc96f13a409b98a2a4916802ef86199a770d1d97f421e314fdc04610b2a1
+oid sha256:328ef2f4fc9fa61f0d8454ac23f36313867c304c38b7ae519dbcbca43c1f1d2a
 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

@@ -11,188 +11,188 @@
     {
       "epoch": 1.0,
       "eval_Claim": {
-        "f1-score": 0.5254768748366867,
-        "precision": 0.5768789443488239,
-        "recall": 0.4824856046065259,
-        "support": 4168.0
+        "f1-score": 0.580553004795429,
+        "precision": 0.5136306192453511,
+        "recall": 0.6675269826372595,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.7984361424847959,
-        "precision": 0.7495921696574225,
-        "recall": 0.854089219330855,
-        "support": 2152.0
+        "f1-score": 0.6929252383341696,
+        "precision": 0.7583745194947831,
+        "recall": 0.6378752886836028,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.9990277532260916,
-        "precision": 0.9988511841640155,
-        "recall": 0.9992043847241867,
-        "support": 11312.0
+        "f1-score": 0.9968272281511393,
+        "precision": 0.9967861557478368,
+        "recall": 0.996868303939344,
+        "support": 12134.0
       },
       "eval_Premise": {
-        "f1-score": 0.8854183656825966,
-        "precision": 0.8719781543651113,
-        "recall": 0.8992793837488611,
-        "support": 12073.0
+        "f1-score": 0.8769487750556793,
+        "precision": 0.9107806691449815,
+        "recall": 0.8455403021704119,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.8755765022723447,
-      "eval_loss": 0.28303390741348267,
+      "eval_accuracy": 0.8654113924050633,
+      "eval_loss": 0.32089120149612427,
       "eval_macro avg": {
-        "f1-score": 0.8020897840575427,
-        "precision": 0.7993251131338432,
-        "recall": 0.8087646481026072,
-        "support": 29705.0
-      },
-      "eval_runtime": 4.8106,
-      "eval_samples_per_second": 16.63,
-      "eval_steps_per_second": 2.079,
+        "f1-score": 0.7868135615841043,
+        "precision": 0.7948929909082382,
+        "recall": 0.7869527193576545,
+        "support": 31600.0
+      },
+      "eval_runtime": 5.9789,
+      "eval_samples_per_second": 13.38,
+      "eval_steps_per_second": 1.673,
       "eval_weighted avg": {
-        "f1-score": 0.8718761173649596,
-        "precision": 0.8700202202343714,
-        "recall": 0.8755765022723447,
-        "support": 29705.0
+        "f1-score": 0.8703967313850799,
+        "precision": 0.8797989523023909,
+        "recall": 0.8654113924050633,
+        "support": 31600.0
       },
       "step": 81
     },
     {
       "epoch": 2.0,
       "eval_Claim": {
-        "f1-score": 0.5578782336033447,
-        "precision": 0.6124497991967871,
-        "recall": 0.5122360844529751,
-        "support": 4168.0
+        "f1-score": 0.6029902384776968,
+        "precision": 0.6369094231271208,
+        "recall": 0.5725011731581418,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.784421283598464,
-        "precision": 0.9571619812583668,
-        "recall": 0.6644981412639405,
-        "support": 2152.0
+        "f1-score": 0.8179940325912326,
+        "precision": 0.8129562043795621,
+        "recall": 0.8230946882217091,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.9998674150351351,
-        "precision": 0.9997348652231551,
-        "recall": 1.0,
-        "support": 11312.0
+        "f1-score": 0.9984359565360553,
+        "precision": 0.9972866304884065,
+        "recall": 0.9995879347288611,
+        "support": 12134.0
       },
       "eval_Premise": {
-        "f1-score": 0.8977749872463995,
-        "precision": 0.8530201342281879,
-        "recall": 0.9474861260664292,
-        "support": 12073.0
+        "f1-score": 0.9017917895214335,
+        "precision": 0.8891539321654864,
+        "recall": 0.9147940793005599,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.8859114627167144,
-      "eval_loss": 0.27363118529319763,
+      "eval_accuracy": 0.8949050632911393,
+      "eval_loss": 0.25671303272247314,
       "eval_macro avg": {
-        "f1-score": 0.8099854798708358,
-        "precision": 0.8555916949766242,
-        "recall": 0.7810550879458362,
-        "support": 29705.0
-      },
-      "eval_runtime": 4.7727,
-      "eval_samples_per_second": 16.762,
-      "eval_steps_per_second": 2.095,
+        "f1-score": 0.8303030042816045,
+        "precision": 0.8340765475401439,
+        "recall": 0.8274944688523179,
+        "support": 31600.0
+      },
+      "eval_runtime": 6.0616,
+      "eval_samples_per_second": 13.198,
+      "eval_steps_per_second": 1.65,
       "eval_weighted avg": {
-        "f1-score": 0.8807489883812781,
-        "precision": 0.8826802296805741,
-        "recall": 0.8859114627167144,
-        "support": 29705.0
+        "f1-score": 0.8928603328205832,
+        "precision": 0.8914339316361279,
+        "recall": 0.8949050632911393,
+        "support": 31600.0
       },
       "step": 162
     },
     {
       "epoch": 3.0,
       "eval_Claim": {
-        "f1-score": 0.64760736196319,
-        "precision": 0.6627322953289804,
-        "recall": 0.6331573896353166,
-        "support": 4168.0
+        "f1-score": 0.6010706092276319,
+        "precision": 0.6579241071428571,
+        "recall": 0.5532613796339747,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.9016018306636155,
-        "precision": 0.8881875563570785,
-        "recall": 0.9154275092936803,
-        "support": 2152.0
+        "f1-score": 0.8101840181497353,
+        "precision": 0.8917869034406215,
+        "recall": 0.7422632794457275,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 0.9996904982977407,
-        "precision": 1.0,
-        "recall": 0.9993811881188119,
-        "support": 11312.0
+        "f1-score": 0.997983456109305,
+        "precision": 0.9967118783394986,
+        "recall": 0.9992582825119499,
+        "support": 12134.0
       },
       "eval_Premise": {
-        "f1-score": 0.8982820417748115,
-        "precision": 0.8936065573770492,
-        "recall": 0.9030067091857865,
-        "support": 12073.0
+        "f1-score": 0.9085203780271707,
+        "precision": 0.8758630507509432,
+        "recall": 0.9437073395199018,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.9027436458508669,
-      "eval_loss": 0.22585685551166534,
+      "eval_accuracy": 0.8985759493670886,
+      "eval_loss": 0.2674676775932312,
       "eval_macro avg": {
-        "f1-score": 0.8617954331748394,
-        "precision": 0.861131602265777,
-        "recall": 0.8627431990583988,
-        "support": 29705.0
-      },
-      "eval_runtime": 4.8003,
-      "eval_samples_per_second": 16.666,
-      "eval_steps_per_second": 2.083,
+        "f1-score": 0.8294396153784607,
+        "precision": 0.8555714849184801,
+        "recall": 0.8096225702778885,
+        "support": 31600.0
+      },
+      "eval_runtime": 5.978,
+      "eval_samples_per_second": 13.382,
+      "eval_steps_per_second": 1.673,
       "eval_weighted avg": {
-        "f1-score": 0.9019670975035187,
-        "precision": 0.9013351218793044,
-        "recall": 0.9027436458508669,
-        "support": 29705.0
+        "f1-score": 0.894668981055346,
+        "precision": 0.8939642861109123,
+        "recall": 0.8985759493670886,
+        "support": 31600.0
       },
       "step": 243
     },
     {
       "epoch": 4.0,
       "eval_Claim": {
-        "f1-score": 0.6477602966229081,
-        "precision": 0.5561865427637239,
-        "recall": 0.7754318618042226,
-        "support": 4168.0
+        "f1-score": 0.66,
+        "precision": 0.6399294843543412,
+        "recall": 0.681370248709526,
+        "support": 4262.0
       },
       "eval_MajorClaim": {
-        "f1-score": 0.8984409799554566,
-        "precision": 0.8627031650983746,
-        "recall": 0.9372676579925651,
-        "support": 2152.0
+        "f1-score": 0.8682903759754079,
+        "precision": 0.8895348837209303,
+        "recall": 0.8480369515011548,
+        "support": 2165.0
       },
       "eval_O": {
-        "f1-score": 1.0,
-        "precision": 1.0,
-        "recall": 1.0,
-        "support": 11312.0
+        "f1-score": 0.9983125488743466,
+        "precision": 0.9971224204554797,
+        "recall": 0.9995055216746332,
+        "support": 12134.0
       },
       "eval_Premise": {
-        "f1-score": 0.861495720750997,
-        "precision": 0.9384029675907849,
-        "recall": 0.7962395427814131,
-        "support": 12073.0
+        "f1-score": 0.9011362757980985,
+        "precision": 0.9082976236852357,
+        "recall": 0.8940869698596519,
+        "support": 13039.0
       },
-      "eval_accuracy": 0.8811311227066151,
-      "eval_loss": 0.32341843843460083,
+      "eval_accuracy": 0.9027215189873418,
+      "eval_loss": 0.2527652680873871,
       "eval_macro avg": {
-        "f1-score": 0.8519242493323405,
-        "precision": 0.8393231688632209,
-        "recall": 0.8772347656445502,
-        "support": 29705.0
-      },
-      "eval_runtime": 4.7708,
-      "eval_samples_per_second": 16.769,
-      "eval_steps_per_second": 2.096,
+        "f1-score": 0.8569348001619632,
+        "precision": 0.8587211030539967,
+        "recall": 0.8557499229362415,
+        "support": 31600.0
+      },
+      "eval_runtime": 5.9795,
+      "eval_samples_per_second": 13.379,
+      "eval_steps_per_second": 1.672,
       "eval_weighted avg": {
-        "f1-score": 0.8869263673393438,
-        "precision": 0.9027457246003854,
-        "recall": 0.8811311227066151,
-        "support": 29705.0
+        "f1-score": 0.9036775010177053,
+        "precision": 0.9049240079307782,
+        "recall": 0.9027215189873418,
+        "support": 31600.0
       },
       "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 sha256:6c4152f1ae85467509d6d1bf2d3b859abf08e786a1c118f9c84f899edc2df5d1
+oid sha256:4fa93f134f1d4f6f4c7db9ba5be632f50fd4a52b8d6cce9d00588b8a1094a71e
 size 4856