End of training

all_results.json

@@ -1,14 +1,14 @@
 {
-    "epoch": 3.16,
-    "eval_accuracy": 0.7967386311437759,
-    "eval_f1": 0.807738814993954,
-    "eval_loss": 0.7215016484260559,
-    "eval_precision": 0.7822014051522248,
-    "eval_recall": 0.835,
-    "eval_runtime": 2.2224,
+    "epoch": 13.16,
+    "eval_accuracy": 0.843706936150666,
+    "eval_f1": 0.8903638151425762,
+    "eval_loss": 0.6187416911125183,
+    "eval_precision": 0.875725338491296,
+    "eval_recall": 0.9055,
+    "eval_runtime": 2.2631,
     "eval_samples": 54,
-    "eval_samples_per_second": 24.298,
-    "eval_steps_per_second": 1.8,
+    "eval_samples_per_second": 23.861,
+    "eval_steps_per_second": 1.767,
     "predict_accuracy": 0.8218373936014088,
     "predict_f1": 0.9038133181559477,
     "predict_loss": 0.6888472437858582,
@@ -17,9 +17,9 @@
     "predict_runtime": 2.0596,
     "predict_samples_per_second": 8.74,
     "predict_steps_per_second": 0.971,
-    "train_loss": 1.0857223510742187,
-    "train_runtime": 75.9454,
+    "train_loss": 0.43045907592773436,
+    "train_runtime": 242.2202,
     "train_samples": 150,
-    "train_samples_per_second": 6.32,
-    "train_steps_per_second": 1.58
+    "train_samples_per_second": 8.257,
+    "train_steps_per_second": 2.064
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
-    "epoch": 3.16,
-    "eval_accuracy": 0.7967386311437759,
-    "eval_f1": 0.807738814993954,
-    "eval_loss": 0.7215016484260559,
-    "eval_precision": 0.7822014051522248,
-    "eval_recall": 0.835,
-    "eval_runtime": 2.2224,
+    "epoch": 13.16,
+    "eval_accuracy": 0.843706936150666,
+    "eval_f1": 0.8903638151425762,
+    "eval_loss": 0.6187416911125183,
+    "eval_precision": 0.875725338491296,
+    "eval_recall": 0.9055,
+    "eval_runtime": 2.2631,
     "eval_samples": 54,
-    "eval_samples_per_second": 24.298,
-    "eval_steps_per_second": 1.8
+    "eval_samples_per_second": 23.861,
+    "eval_steps_per_second": 1.767
 }

runs/Aug24_17-25-51_bernini/events.out.tfevents.1724513165.bernini.30085.0

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4b4d729d2e90d3aaa91fb346b6f513c7a698c52521481e4630923414523477cb
-size 8948
+oid sha256:32d6032c284b59e0446a09b95f497071f6060212597fb56c0b5a23aa09dee71b
+size 9302

runs/Aug24_17-25-51_bernini/events.out.tfevents.1724513416.bernini.30085.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c6f1d5f254aaca03278ea8df80b1c17b487e7f9b3082019b0e9e59e1714b1cf
+size 512

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.16,
-    "train_loss": 1.0857223510742187,
-    "train_runtime": 75.9454,
+    "epoch": 13.16,
+    "train_loss": 0.43045907592773436,
+    "train_runtime": 242.2202,
     "train_samples": 150,
-    "train_samples_per_second": 6.32,
-    "train_steps_per_second": 1.58
+    "train_samples_per_second": 8.257,
+    "train_steps_per_second": 2.064
 }

trainer_state.json

@@ -1,169 +1,151 @@
 {
-  "best_metric": null,
-  "best_model_checkpoint": null,
-  "epoch": 3.1578947368421053,
-  "global_step": 120,
+  "best_metric": 0.8903638151425762,
+  "best_model_checkpoint": "/media/HDD/mpizzol/Tirocinio/layoutlmv3-base-finetuned-funsd/checkpoint-500",
+  "epoch": 13.157894736842104,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.26,
-      "eval_accuracy": 0.3226917776757005,
-      "eval_f1": 0.07924874026568941,
-      "eval_loss": 1.737101435661316,
-      "eval_precision": 0.07311918850380389,
-      "eval_recall": 0.0865,
-      "eval_runtime": 2.2526,
-      "eval_samples_per_second": 23.972,
-      "eval_steps_per_second": 1.776,
-      "step": 10
-    },
-    {
-      "epoch": 0.53,
-      "eval_accuracy": 0.4538355535140101,
-      "eval_f1": 0.14812929498600153,
-      "eval_loss": 1.4963377714157104,
-      "eval_precision": 0.15085536547433903,
-      "eval_recall": 0.1455,
-      "eval_runtime": 2.2423,
-      "eval_samples_per_second": 24.082,
-      "eval_steps_per_second": 1.784,
-      "step": 20
+      "epoch": 1.32,
+      "eval_accuracy": 0.7606798346348186,
+      "eval_f1": 0.7277096851718338,
+      "eval_loss": 0.9062727689743042,
+      "eval_precision": 0.7006015733456733,
+      "eval_recall": 0.757,
+      "eval_runtime": 2.4285,
+      "eval_samples_per_second": 22.236,
+      "eval_steps_per_second": 1.647,
+      "step": 50
     },
     {
-      "epoch": 0.79,
-      "eval_accuracy": 0.5964630225080386,
-      "eval_f1": 0.4187978142076503,
-      "eval_loss": 1.3012946844100952,
-      "eval_precision": 0.3720388349514563,
-      "eval_recall": 0.479,
-      "eval_runtime": 2.2373,
-      "eval_samples_per_second": 24.136,
-      "eval_steps_per_second": 1.788,
-      "step": 30
+      "epoch": 2.63,
+      "eval_accuracy": 0.7967386311437759,
+      "eval_f1": 0.8242074927953891,
+      "eval_loss": 0.6386759281158447,
+      "eval_precision": 0.7929759704251387,
+      "eval_recall": 0.858,
+      "eval_runtime": 2.2823,
+      "eval_samples_per_second": 23.66,
+      "eval_steps_per_second": 1.753,
+      "step": 100
     },
     {
-      "epoch": 1.05,
-      "eval_accuracy": 0.6367707854846119,
-      "eval_f1": 0.5357978474496958,
-      "eval_loss": 1.155036449432373,
-      "eval_precision": 0.5035180299032542,
-      "eval_recall": 0.5725,
-      "eval_runtime": 2.3937,
-      "eval_samples_per_second": 22.56,
-      "eval_steps_per_second": 1.671,
-      "step": 40
+      "epoch": 3.95,
+      "eval_accuracy": 0.825447864033073,
+      "eval_f1": 0.8485576923076923,
+      "eval_loss": 0.5691221356391907,
+      "eval_precision": 0.8171296296296297,
+      "eval_recall": 0.8825,
+      "eval_runtime": 2.2496,
+      "eval_samples_per_second": 24.004,
+      "eval_steps_per_second": 1.778,
+      "step": 150
     },
     {
-      "epoch": 1.32,
-      "eval_accuracy": 0.7382866329811667,
-      "eval_f1": 0.6415929203539823,
-      "eval_loss": 1.0083192586898804,
-      "eval_precision": 0.6310444874274661,
-      "eval_recall": 0.6525,
-      "eval_runtime": 2.2203,
-      "eval_samples_per_second": 24.321,
-      "eval_steps_per_second": 1.802,
-      "step": 50
+      "epoch": 5.26,
+      "eval_accuracy": 0.822347266881029,
+      "eval_f1": 0.8555474629764507,
+      "eval_loss": 0.5723332166671753,
+      "eval_precision": 0.831524303916942,
+      "eval_recall": 0.881,
+      "eval_runtime": 2.2738,
+      "eval_samples_per_second": 23.749,
+      "eval_steps_per_second": 1.759,
+      "step": 200
     },
     {
-      "epoch": 1.58,
-      "eval_accuracy": 0.7382866329811667,
-      "eval_f1": 0.6930741190765491,
-      "eval_loss": 0.9313498139381409,
-      "eval_precision": 0.6742316784869976,
-      "eval_recall": 0.713,
-      "eval_runtime": 2.2084,
-      "eval_samples_per_second": 24.452,
-      "eval_steps_per_second": 1.811,
-      "step": 60
+      "epoch": 6.58,
+      "eval_accuracy": 0.8292374827744603,
+      "eval_f1": 0.8729388942774006,
+      "eval_loss": 0.5896942019462585,
+      "eval_precision": 0.847457627118644,
+      "eval_recall": 0.9,
+      "eval_runtime": 2.2927,
+      "eval_samples_per_second": 23.553,
+      "eval_steps_per_second": 1.745,
+      "step": 250
     },
     {
-      "epoch": 1.84,
-      "eval_accuracy": 0.7544786403307303,
-      "eval_f1": 0.7475519465010746,
-      "eval_loss": 0.8519546389579773,
-      "eval_precision": 0.7155921353452218,
-      "eval_recall": 0.7825,
-      "eval_runtime": 2.216,
-      "eval_samples_per_second": 24.369,
-      "eval_steps_per_second": 1.805,
-      "step": 70
+      "epoch": 7.89,
+      "eval_accuracy": 0.8283187873220027,
+      "eval_f1": 0.874515503875969,
+      "eval_loss": 0.6122345924377441,
+      "eval_precision": 0.8482142857142857,
+      "eval_recall": 0.9025,
+      "eval_runtime": 2.2852,
+      "eval_samples_per_second": 23.63,
+      "eval_steps_per_second": 1.75,
+      "step": 300
     },
     {
-      "epoch": 2.11,
-      "eval_accuracy": 0.7816949931097841,
-      "eval_f1": 0.777831094049904,
-      "eval_loss": 0.7899913191795349,
-      "eval_precision": 0.7476937269372693,
-      "eval_recall": 0.8105,
-      "eval_runtime": 2.218,
-      "eval_samples_per_second": 24.346,
-      "eval_steps_per_second": 1.803,
-      "step": 80
+      "epoch": 9.21,
+      "eval_accuracy": 0.8392282958199357,
+      "eval_f1": 0.8740884783665532,
+      "eval_loss": 0.6045289635658264,
+      "eval_precision": 0.8505203405865658,
+      "eval_recall": 0.899,
+      "eval_runtime": 2.4063,
+      "eval_samples_per_second": 22.441,
+      "eval_steps_per_second": 1.662,
+      "step": 350
     },
     {
-      "epoch": 2.37,
-      "eval_accuracy": 0.7456361966008268,
-      "eval_f1": 0.7814885496183206,
-      "eval_loss": 0.8020110726356506,
-      "eval_precision": 0.7472627737226277,
-      "eval_recall": 0.819,
-      "eval_runtime": 2.3405,
-      "eval_samples_per_second": 23.072,
-      "eval_steps_per_second": 1.709,
-      "step": 90
+      "epoch": 10.53,
+      "eval_accuracy": 0.8446256316031235,
+      "eval_f1": 0.8851733464470125,
+      "eval_loss": 0.5662067532539368,
+      "eval_precision": 0.8708272859216255,
+      "eval_recall": 0.9,
+      "eval_runtime": 2.2543,
+      "eval_samples_per_second": 23.954,
+      "eval_steps_per_second": 1.774,
+      "step": 400
     },
     {
-      "epoch": 2.63,
-      "eval_accuracy": 0.7959347726228755,
-      "eval_f1": 0.8073572120038721,
-      "eval_loss": 0.7402865886688232,
-      "eval_precision": 0.7823639774859287,
-      "eval_recall": 0.834,
-      "eval_runtime": 2.2034,
-      "eval_samples_per_second": 24.507,
-      "eval_steps_per_second": 1.815,
-      "step": 100
+      "epoch": 11.84,
+      "eval_accuracy": 0.843706936150666,
+      "eval_f1": 0.888943488943489,
+      "eval_loss": 0.5973214507102966,
+      "eval_precision": 0.8739130434782608,
+      "eval_recall": 0.9045,
+      "eval_runtime": 2.2748,
+      "eval_samples_per_second": 23.738,
+      "eval_steps_per_second": 1.758,
+      "step": 450
     },
     {
-      "epoch": 2.89,
-      "eval_accuracy": 0.7952457510335323,
-      "eval_f1": 0.8107584201599225,
-      "eval_loss": 0.7252823114395142,
-      "eval_precision": 0.7865538316878232,
-      "eval_recall": 0.8365,
-      "eval_runtime": 2.2471,
-      "eval_samples_per_second": 24.031,
-      "eval_steps_per_second": 1.78,
-      "step": 110
+      "epoch": 13.16,
+      "learning_rate": 0.0,
+      "loss": 0.4305,
+      "step": 500
     },
     {
-      "epoch": 3.16,
-      "eval_accuracy": 0.7967386311437759,
-      "eval_f1": 0.807738814993954,
-      "eval_loss": 0.7215016484260559,
-      "eval_precision": 0.7822014051522248,
-      "eval_recall": 0.835,
-      "eval_runtime": 2.2177,
-      "eval_samples_per_second": 24.35,
-      "eval_steps_per_second": 1.804,
-      "step": 120
+      "epoch": 13.16,
+      "eval_accuracy": 0.843706936150666,
+      "eval_f1": 0.8903638151425762,
+      "eval_loss": 0.6187416911125183,
+      "eval_precision": 0.875725338491296,
+      "eval_recall": 0.9055,
+      "eval_runtime": 2.216,
+      "eval_samples_per_second": 24.369,
+      "eval_steps_per_second": 1.805,
+      "step": 500
     },
     {
-      "epoch": 3.16,
-      "step": 120,
-      "total_flos": 124948932691968.0,
-      "train_loss": 1.0857223510742187,
-      "train_runtime": 75.9454,
-      "train_samples_per_second": 6.32,
-      "train_steps_per_second": 1.58
+      "epoch": 13.16,
+      "step": 500,
+      "total_flos": 520356947539968.0,
+      "train_loss": 0.43045907592773436,
+      "train_runtime": 242.2202,
+      "train_samples_per_second": 8.257,
+      "train_steps_per_second": 2.064
     }
   ],
-  "max_steps": 120,
-  "num_train_epochs": 4,
-  "total_flos": 124948932691968.0,
+  "max_steps": 500,
+  "num_train_epochs": 14,
+  "total_flos": 520356947539968.0,
   "trial_name": null,
   "trial_params": null
 }