Training in progress, epoch 1

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ca0dc6ea72694ff54a88f458db9aff463c496bd6d83166f6fb57a4e51c5bb172
+oid sha256:849b2219b6f76700a8f32131536f6b6a9d2ed82dad4be3ec62b9fa941c15cf27
 size 409103316

run-18/checkpoint-534/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a2c26a4990298c617435add4fef099a107381df4b5d4dd2ceb1932c28c743541
+oid sha256:b5b3ada415323df4681659f8a3c19121e9379a2b0c7922cade957c3cc9526017
 size 409103316

run-18/checkpoint-534/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c7c82d05f4ce26f22c9acaa7998473a49f5f6cbfcd52e5fa856d9f347b60ad8c
+oid sha256:728b20a4fc94ce708eb15da6dc44b3db5573506ae45b2a48cd0e442e7f0ccbb6
 size 818327802

run-18/checkpoint-534/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dda5eab8086c3361f01b6da1d0d501e5d7ffcc15041db5c9a3384c00b843300f
+oid sha256:8a56050495e1938ae70b239ae8ece3c43ac800b58f0a6e02bd7ed6cd1669797d
 size 1064

run-18/checkpoint-534/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.5817901492118835,
+  "best_metric": 0.6037774682044983,
   "best_model_checkpoint": "BERT-WMM/run-18/checkpoint-534",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,39 +10,39 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7563380281690141,
-      "eval_loss": 0.6019130349159241,
-      "eval_runtime": 2.4196,
-      "eval_samples_per_second": 880.318,
-      "eval_steps_per_second": 55.381,
+      "eval_accuracy": 0.7530516431924883,
+      "eval_loss": 0.6065725684165955,
+      "eval_runtime": 2.6718,
+      "eval_samples_per_second": 797.215,
+      "eval_steps_per_second": 50.153,
       "step": 267
     },
     {
       "epoch": 1.87,
-      "grad_norm": 9.674468994140625,
-      "learning_rate": 5.7987153604787005e-06,
-      "loss": 0.6061,
+      "grad_norm": 9.447751998901367,
+      "learning_rate": 9.825065893169892e-07,
+      "loss": 0.6083,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7619718309859155,
-      "eval_loss": 0.5817901492118835,
-      "eval_runtime": 2.455,
-      "eval_samples_per_second": 867.605,
-      "eval_steps_per_second": 54.582,
+      "eval_accuracy": 0.7615023474178404,
+      "eval_loss": 0.6037774682044983,
+      "eval_runtime": 2.5277,
+      "eval_samples_per_second": 842.674,
+      "eval_steps_per_second": 53.013,
       "step": 534
     }
   ],
   "logging_steps": 500,
-  "max_steps": 801,
+  "max_steps": 534,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "total_flos": 349221394826640.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "per_device_train_batch_size": 32
+    "weight_decay": 0.0016219939810654786
   }
 }

run-18/checkpoint-534/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9cc12ec9008fe3ca8c11132361abe51e0525170497895a06dfb1609f916fd77f
+oid sha256:e1e1153af9790ef8808125f63933179e86e55e5d9fad145931b50ae4a04de8a4
 size 4856

run-19/checkpoint-267/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a63d61dd1dbef133a4a54ab1f2f18adfbb803f4221d5942217c0b43ae7213cff
+oid sha256:849b2219b6f76700a8f32131536f6b6a9d2ed82dad4be3ec62b9fa941c15cf27
 size 409103316

run-19/checkpoint-267/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a6c35c19b3033524502b333f3effce0d2d99fe0651595357c1e9ee00b1bb5471
+oid sha256:28ff46f6452b3ccd1d01e2fc84a1d4cffbeec5188a59986dfd1a7ddffb875f00
 size 818327802

run-19/checkpoint-267/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:123e89548a83b33069b1dc6f7e79d621eac970b4e99c56bc48c19b6733a7f692
+oid sha256:0980580128ccd5e578c966c8ffb616b8165a0951ed4f83a639bfa5862c1defb0
 size 14244

run-19/checkpoint-267/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f3081e0c1671f48c011c79aa776bcc1778551b80f8c0052c7c3a201996bd0303
+oid sha256:ef3be8be9037b2d6a9348611f95c19fb1d17ca7cb1da92486df22b7e0a3c655a
 size 1064

run-19/checkpoint-267/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.64137202501297,
+  "best_metric": 0.6066781878471375,
   "best_model_checkpoint": "BERT-WMM/run-19/checkpoint-267",
   "epoch": 1.0,
   "eval_steps": 500,
@@ -10,26 +10,23 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7361502347417841,
-      "eval_loss": 0.64137202501297,
-      "eval_runtime": 1.9678,
-      "eval_samples_per_second": 1082.453,
-      "eval_steps_per_second": 68.098,
+      "eval_accuracy": 0.7539906103286385,
+      "eval_loss": 0.6066781878471375,
+      "eval_runtime": 2.5156,
+      "eval_samples_per_second": 846.702,
+      "eval_steps_per_second": 53.267,
       "step": 267
     }
   ],
   "logging_steps": 500,
-  "max_steps": 801,
+  "max_steps": 534,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
+  "num_train_epochs": 2,
   "save_steps": 500,
   "total_flos": 0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "learning_rate": 1.4384792339701421e-05,
-    "num_train_epochs": 3,
-    "per_device_train_batch_size": 32,
-    "weight_decay": 0.00035180961962651564
+    "weight_decay": 0.051424292123722175
   }
 }

run-19/checkpoint-267/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bda6d0591379b4005638e47cc567a832be7f1e2ee85c267cf84315e6fce938e6
+oid sha256:6e18fed6d1d6cb4145ea5f498532a391408ab2f035069ebc34d88b7221a52d0a
 size 4856

runs/Apr19_10-19-01_dcdaa3e6ec43/events.out.tfevents.1713524754.dcdaa3e6ec43.789.19

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7a5d29282cb394fe1a6bb799c5bc30cce18862349c8d4e9a9e24079e86697ea9
+size 6121

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e1e1153af9790ef8808125f63933179e86e55e5d9fad145931b50ae4a04de8a4
+oid sha256:6e18fed6d1d6cb4145ea5f498532a391408ab2f035069ebc34d88b7221a52d0a
 size 4856