End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- id
 license: mit
 base_model: indolem/indobert-base-uncased
 tags:

all_results.json

@@ -1,21 +1,21 @@
 {
-    "accuracy": 0.9070227497527201,
+    "accuracy": 0.9030662710187932,
     "epoch": 20.0,
-    "eval_accuracy": 0.8947368421052632,
-    "eval_f1": 0.8723802851224565,
-    "eval_loss": 0.31320512294769287,
-    "eval_precision": 0.8743110838946724,
-    "eval_recall": 0.8705219130751045,
-    "eval_runtime": 4.938,
+    "eval_accuracy": 0.8872180451127819,
+    "eval_f1": 0.8622036668943447,
+    "eval_loss": 0.32483023405075073,
+    "eval_precision": 0.8671602787456446,
+    "eval_recall": 0.8577014002545917,
+    "eval_runtime": 1.7849,
     "eval_samples": 399,
-    "eval_samples_per_second": 80.802,
-    "eval_steps_per_second": 10.126,
-    "f1": 0.8883990642967194,
-    "precision": 0.8876511954992967,
-    "recall": 0.8891582028860001,
-    "train_loss": 0.232629602463519,
-    "train_runtime": 1905.8691,
+    "eval_samples_per_second": 223.546,
+    "eval_steps_per_second": 28.013,
+    "f1": 0.8838748241912799,
+    "precision": 0.8824176388720238,
+    "recall": 0.8853765637207376,
+    "train_loss": 0.2495564101172275,
+    "train_runtime": 624.4298,
     "train_samples": 3638,
-    "train_samples_per_second": 38.177,
-    "train_steps_per_second": 1.28
+    "train_samples_per_second": 116.522,
+    "train_steps_per_second": 3.908
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.8947368421052632,
-    "eval_f1": 0.8723802851224565,
-    "eval_loss": 0.31320512294769287,
-    "eval_precision": 0.8743110838946724,
-    "eval_recall": 0.8705219130751045,
-    "eval_runtime": 4.938,
+    "eval_accuracy": 0.8872180451127819,
+    "eval_f1": 0.8622036668943447,
+    "eval_loss": 0.32483023405075073,
+    "eval_precision": 0.8671602787456446,
+    "eval_recall": 0.8577014002545917,
+    "eval_runtime": 1.7849,
     "eval_samples": 399,
-    "eval_samples_per_second": 80.802,
-    "eval_steps_per_second": 10.126
+    "eval_samples_per_second": 223.546,
+    "eval_steps_per_second": 28.013
 }

predict_results.json

@@ -1,6 +1,6 @@
 {
-    "accuracy": 0.9070227497527201,
-    "f1": 0.8883990642967194,
-    "precision": 0.8876511954992967,
-    "recall": 0.8891582028860001
+    "accuracy": 0.9030662710187932,
+    "f1": 0.8838748241912799,
+    "precision": 0.8824176388720238,
+    "recall": 0.8853765637207376
 }

predict_results.txt

@@ -29,7 +29,7 @@ index	prediction
 27	1
 28	1
 29	1
-30	0
+30	1
 31	0
 32	1
 33	1
@@ -38,7 +38,7 @@ index	prediction
 36	0
 37	1
 38	1
-39	1
+39	0
 40	1
 41	1
 42	1
@@ -56,7 +56,7 @@ index	prediction
 54	1
 55	1
 56	1
-57	0
+57	1
 58	1
 59	1
 60	1
@@ -86,7 +86,7 @@ index	prediction
 84	0
 85	1
 86	1
-87	1
+87	0
 88	1
 89	1
 90	1
@@ -98,15 +98,15 @@ index	prediction
 96	1
 97	1
 98	1
-99	0
+99	1
 100	1
-101	0
+101	1
 102	1
 103	1
 104	1
-105	0
+105	1
 106	1
-107	0
+107	1
 108	1
 109	1
 110	0
@@ -120,7 +120,7 @@ index	prediction
 118	1
 119	1
 120	1
-121	1
+121	0
 122	1
 123	0
 124	1
@@ -130,7 +130,7 @@ index	prediction
 128	1
 129	1
 130	1
-131	1
+131	0
 132	1
 133	1
 134	1
@@ -152,7 +152,7 @@ index	prediction
 150	1
 151	1
 152	1
-153	1
+153	0
 154	1
 155	1
 156	1
@@ -164,7 +164,7 @@ index	prediction
 162	1
 163	1
 164	0
-165	1
+165	0
 166	1
 167	1
 168	1
@@ -173,10 +173,10 @@ index	prediction
 171	1
 172	1
 173	1
-174	1
+174	0
 175	1
 176	1
-177	0
+177	1
 178	1
 179	1
 180	1
@@ -184,7 +184,7 @@ index	prediction
 182	1
 183	1
 184	1
-185	1
+185	0
 186	1
 187	1
 188	1
@@ -206,7 +206,7 @@ index	prediction
 204	0
 205	1
 206	1
-207	1
+207	0
 208	1
 209	1
 210	1
@@ -217,7 +217,7 @@ index	prediction
 215	0
 216	1
 217	0
-218	1
+218	0
 219	1
 220	1
 221	1
@@ -239,19 +239,19 @@ index	prediction
 237	1
 238	0
 239	1
-240	0
+240	1
 241	1
 242	1
 243	1
 244	1
-245	0
+245	1
 246	1
 247	1
 248	1
 249	1
 250	0
 251	0
-252	0
+252	1
 253	1
 254	1
 255	1
@@ -279,7 +279,7 @@ index	prediction
 277	1
 278	1
 279	1
-280	1
+280	0
 281	1
 282	1
 283	1
@@ -322,7 +322,7 @@ index	prediction
 320	0
 321	0
 322	0
-323	1
+323	0
 324	0
 325	0
 326	0
@@ -344,7 +344,7 @@ index	prediction
 342	0
 343	0
 344	0
-345	0
+345	1
 346	0
 347	0
 348	0
@@ -365,7 +365,7 @@ index	prediction
 363	0
 364	0
 365	0
-366	1
+366	0
 367	0
 368	0
 369	0
@@ -385,10 +385,10 @@ index	prediction
 383	0
 384	0
 385	0
-386	0
-387	0
+386	1
+387	1
 388	0
-389	1
+389	0
 390	0
 391	0
 392	0
@@ -440,7 +440,7 @@ index	prediction
 438	0
 439	0
 440	0
-441	1
+441	0
 442	0
 443	0
 444	0
@@ -472,7 +472,7 @@ index	prediction
 470	0
 471	0
 472	0
-473	0
+473	1
 474	0
 475	0
 476	0
@@ -496,8 +496,8 @@ index	prediction
 494	0
 495	0
 496	0
-497	1
-498	1
+497	0
+498	0
 499	0
 500	0
 501	0
@@ -525,14 +525,14 @@ index	prediction
 523	0
 524	0
 525	0
-526	0
+526	1
 527	0
 528	0
 529	0
 530	0
-531	0
+531	1
 532	0
-533	0
+533	1
 534	0
 535	0
 536	0
@@ -561,7 +561,7 @@ index	prediction
 559	0
 560	0
 561	0
-562	1
+562	0
 563	0
 564	0
 565	0
@@ -580,9 +580,9 @@ index	prediction
 578	0
 579	0
 580	0
-581	1
+581	0
 582	0
-583	0
+583	1
 584	0
 585	0
 586	0
@@ -594,7 +594,7 @@ index	prediction
 592	0
 593	0
 594	0
-595	0
+595	1
 596	0
 597	0
 598	0
@@ -604,15 +604,15 @@ index	prediction
 602	0
 603	1
 604	0
-605	0
+605	1
 606	0
 607	0
 608	0
 609	0
 610	0
-611	0
+611	1
 612	0
-613	1
+613	0
 614	0
 615	0
 616	0
@@ -622,7 +622,7 @@ index	prediction
 620	0
 621	0
 622	0
-623	0
+623	1
 624	0
 625	1
 626	0
@@ -662,9 +662,9 @@ index	prediction
 660	0
 661	0
 662	0
-663	1
+663	0
 664	1
-665	0
+665	1
 666	0
 667	0
 668	0
@@ -685,7 +685,7 @@ index	prediction
 683	0
 684	0
 685	0
-686	1
+686	0
 687	0
 688	0
 689	0
@@ -705,7 +705,7 @@ index	prediction
 703	0
 704	0
 705	0
-706	1
+706	0
 707	0
 708	0
 709	0
@@ -765,12 +765,12 @@ index	prediction
 763	0
 764	0
 765	0
-766	0
+766	1
 767	0
 768	0
 769	0
 770	0
-771	1
+771	0
 772	0
 773	0
 774	0
@@ -840,7 +840,7 @@ index	prediction
 838	0
 839	0
 840	0
-841	1
+841	0
 842	1
 843	0
 844	0
@@ -859,7 +859,7 @@ index	prediction
 857	0
 858	0
 859	0
-860	0
+860	1
 861	0
 862	0
 863	0
@@ -968,19 +968,19 @@ index	prediction
 966	1
 967	0
 968	0
-969	0
+969	1
 970	0
 971	0
 972	0
 973	0
-974	0
+974	1
 975	0
 976	0
 977	0
 978	0
 979	0
 980	1
-981	1
+981	0
 982	0
 983	0
 984	0
@@ -998,7 +998,7 @@ index	prediction
 996	0
 997	0
 998	0
-999	0
+999	1
 1000	0
 1001	0
 1002	0

runs/Jun03_13-55-30_a358b85c7679/events.out.tfevents.1717423570.a358b85c7679.123191.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec63af5327041cb52e73510de6fa75b1fc806de4b66867881961bb3c4dc11ce5
+size 560

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.232629602463519,
-    "train_runtime": 1905.8691,
+    "train_loss": 0.2495564101172275,
+    "train_runtime": 624.4298,
     "train_samples": 3638,
-    "train_samples_per_second": 38.177,
-    "train_steps_per_second": 1.28
+    "train_samples_per_second": 116.522,
+    "train_steps_per_second": 3.908
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 4.3584418296813965,
+      "grad_norm": 4.458889961242676,
       "learning_rate": 4.75e-05,
-      "loss": 0.5494,
+      "loss": 0.5533,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.731829573934837,
-      "eval_f1": 0.6369179742314071,
-      "eval_loss": 0.5032888650894165,
-      "eval_precision": 0.6683441169422477,
-      "eval_recall": 0.6277505000909256,
-      "eval_runtime": 5.0519,
-      "eval_samples_per_second": 78.98,
-      "eval_steps_per_second": 9.897,
+      "eval_accuracy": 0.7268170426065163,
+      "eval_f1": 0.6326680574676724,
+      "eval_loss": 0.5133728384971619,
+      "eval_precision": 0.6605799373040753,
+      "eval_recall": 0.6242044008001455,
+      "eval_runtime": 1.774,
+      "eval_samples_per_second": 224.915,
+      "eval_steps_per_second": 28.185,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 4.557066440582275,
+      "grad_norm": 3.8158535957336426,
       "learning_rate": 4.5e-05,
-      "loss": 0.4551,
+      "loss": 0.4779,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7769423558897243,
-      "eval_f1": 0.7522344322344323,
-      "eval_loss": 0.4275681674480438,
-      "eval_precision": 0.7434314656536879,
-      "eval_recall": 0.7796872158574286,
-      "eval_runtime": 5.0559,
-      "eval_samples_per_second": 78.917,
-      "eval_steps_per_second": 9.889,
+      "eval_accuracy": 0.7418546365914787,
+      "eval_f1": 0.7122401394791937,
+      "eval_loss": 0.4949621260166168,
+      "eval_precision": 0.7053803339517626,
+      "eval_recall": 0.734860883797054,
+      "eval_runtime": 1.7746,
+      "eval_samples_per_second": 224.844,
+      "eval_steps_per_second": 28.176,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 3.5165185928344727,
+      "grad_norm": 3.156679630279541,
       "learning_rate": 4.25e-05,
-      "loss": 0.3731,
+      "loss": 0.4097,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.8395989974937343,
-      "eval_f1": 0.7878928832480564,
-      "eval_loss": 0.35077065229415894,
-      "eval_precision": 0.8294492319220682,
-      "eval_recall": 0.7665030005455538,
-      "eval_runtime": 5.0754,
-      "eval_samples_per_second": 78.614,
-      "eval_steps_per_second": 9.851,
+      "eval_accuracy": 0.8245614035087719,
+      "eval_f1": 0.7664715719063545,
+      "eval_loss": 0.3772188425064087,
+      "eval_precision": 0.8092877840475827,
+      "eval_recall": 0.7458628841607565,
+      "eval_runtime": 1.8118,
+      "eval_samples_per_second": 220.228,
+      "eval_steps_per_second": 27.597,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 3.075305461883545,
+      "grad_norm": 3.239713191986084,
       "learning_rate": 4e-05,
-      "loss": 0.3032,
+      "loss": 0.3451,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8671679197994987,
-      "eval_f1": 0.8466330637850383,
-      "eval_loss": 0.3236788511276245,
-      "eval_precision": 0.8353276671885485,
-      "eval_recall": 0.8635206401163849,
-      "eval_runtime": 5.0489,
-      "eval_samples_per_second": 79.027,
-      "eval_steps_per_second": 9.903,
+      "eval_accuracy": 0.8446115288220551,
+      "eval_f1": 0.8169941409717701,
+      "eval_loss": 0.3511227071285248,
+      "eval_precision": 0.8104735988883742,
+      "eval_recall": 0.8250591016548463,
+      "eval_runtime": 1.8097,
+      "eval_samples_per_second": 220.479,
+      "eval_steps_per_second": 27.629,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 3.069729804992676,
+      "grad_norm": 2.155226707458496,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.2718,
+      "loss": 0.2959,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8469505178365937,
-      "eval_loss": 0.31016021966934204,
-      "eval_precision": 0.844489247311828,
-      "eval_recall": 0.8495635570103655,
-      "eval_runtime": 5.0508,
-      "eval_samples_per_second": 78.998,
-      "eval_steps_per_second": 9.9,
+      "eval_accuracy": 0.8546365914786967,
+      "eval_f1": 0.8255172205802521,
+      "eval_loss": 0.32013869285583496,
+      "eval_precision": 0.8239495798319327,
+      "eval_recall": 0.8271503909801782,
+      "eval_runtime": 1.81,
+      "eval_samples_per_second": 220.445,
+      "eval_steps_per_second": 27.625,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 1.1438698768615723,
+      "grad_norm": 2.97943115234375,
       "learning_rate": 3.5e-05,
-      "loss": 0.2642,
+      "loss": 0.2727,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.87468671679198,
-      "eval_f1": 0.8524146298159436,
-      "eval_loss": 0.30064335465431213,
-      "eval_precision": 0.8451250578971746,
-      "eval_recall": 0.8613384251682124,
-      "eval_runtime": 5.048,
-      "eval_samples_per_second": 79.041,
-      "eval_steps_per_second": 9.905,
+      "eval_accuracy": 0.8646616541353384,
+      "eval_f1": 0.8447157518450185,
+      "eval_loss": 0.3176342844963074,
+      "eval_precision": 0.8325401217487549,
+      "eval_recall": 0.864248045099109,
+      "eval_runtime": 1.8128,
+      "eval_samples_per_second": 220.097,
+      "eval_steps_per_second": 27.581,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.622890830039978,
+      "grad_norm": 1.0954539775848389,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.2394,
+      "loss": 0.2595,
       "step": 854
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.8721804511278195,
-      "eval_f1": 0.8404212771630449,
-      "eval_loss": 0.3013169765472412,
-      "eval_precision": 0.854416558018253,
-      "eval_recall": 0.8295599199854519,
-      "eval_runtime": 5.0554,
-      "eval_samples_per_second": 78.925,
-      "eval_steps_per_second": 9.89,
+      "eval_accuracy": 0.87468671679198,
+      "eval_f1": 0.8524146298159436,
+      "eval_loss": 0.2958522439002991,
+      "eval_precision": 0.8451250578971746,
+      "eval_recall": 0.8613384251682124,
+      "eval_runtime": 1.8171,
+      "eval_samples_per_second": 219.582,
+      "eval_steps_per_second": 27.516,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 2.1770668029785156,
+      "grad_norm": 6.516312122344971,
       "learning_rate": 3e-05,
-      "loss": 0.2234,
+      "loss": 0.2409,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8796992481203008,
-      "eval_f1": 0.8533986527862829,
-      "eval_loss": 0.290397047996521,
-      "eval_precision": 0.8572003218020917,
-      "eval_recall": 0.8498817966903074,
-      "eval_runtime": 5.0587,
-      "eval_samples_per_second": 78.874,
-      "eval_steps_per_second": 9.884,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8649122807017544,
+      "eval_loss": 0.28329744935035706,
+      "eval_precision": 0.8710116366366366,
+      "eval_recall": 0.8594744498999818,
+      "eval_runtime": 1.8171,
+      "eval_samples_per_second": 219.577,
+      "eval_steps_per_second": 27.516,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 8.559947967529297,
+      "grad_norm": 6.709987163543701,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.2098,
+      "loss": 0.2298,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8701248742380304,
-      "eval_loss": 0.2983732223510742,
-      "eval_precision": 0.8624507874015748,
-      "eval_recall": 0.8794780869248955,
-      "eval_runtime": 5.0499,
-      "eval_samples_per_second": 79.011,
-      "eval_steps_per_second": 9.901,
+      "eval_accuracy": 0.8771929824561403,
+      "eval_f1": 0.850729517396184,
+      "eval_loss": 0.2893889546394348,
+      "eval_precision": 0.8535087719298247,
+      "eval_recall": 0.8481087470449173,
+      "eval_runtime": 1.8128,
+      "eval_samples_per_second": 220.099,
+      "eval_steps_per_second": 27.581,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 2.985607385635376,
+      "grad_norm": 4.345912933349609,
       "learning_rate": 2.5e-05,
-      "loss": 0.2029,
+      "loss": 0.2221,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8495004213314072,
-      "eval_loss": 0.31886106729507446,
-      "eval_precision": 0.8761860561056105,
-      "eval_recall": 0.8316512093107837,
-      "eval_runtime": 5.0651,
-      "eval_samples_per_second": 78.775,
-      "eval_steps_per_second": 9.872,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8614765038536611,
+      "eval_loss": 0.2884393632411957,
+      "eval_precision": 0.8686536646744258,
+      "eval_recall": 0.8552009456264775,
+      "eval_runtime": 1.8154,
+      "eval_samples_per_second": 219.786,
+      "eval_steps_per_second": 27.542,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 0.5187826156616211,
+      "grad_norm": 2.3872387409210205,
       "learning_rate": 2.25e-05,
-      "loss": 0.1917,
+      "loss": 0.1986,
       "step": 1342
     },
     {
       "epoch": 11.0,
       "eval_accuracy": 0.8847117794486216,
       "eval_f1": 0.8587719298245614,
-      "eval_loss": 0.28483322262763977,
+      "eval_loss": 0.2855367362499237,
       "eval_precision": 0.864771021021021,
       "eval_recall": 0.8534278959810875,
-      "eval_runtime": 5.0495,
-      "eval_samples_per_second": 79.018,
-      "eval_steps_per_second": 9.902,
+      "eval_runtime": 1.8165,
+      "eval_samples_per_second": 219.656,
+      "eval_steps_per_second": 27.526,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 12.659485816955566,
+      "grad_norm": 5.007177829742432,
       "learning_rate": 2e-05,
-      "loss": 0.1797,
+      "loss": 0.1964,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8771929824561403,
-      "eval_f1": 0.850729517396184,
-      "eval_loss": 0.30030354857444763,
-      "eval_precision": 0.8535087719298247,
-      "eval_recall": 0.8481087470449173,
-      "eval_runtime": 5.0516,
-      "eval_samples_per_second": 78.985,
-      "eval_steps_per_second": 9.898,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.8521068445832446,
+      "eval_loss": 0.29210031032562256,
+      "eval_precision": 0.8693800752624282,
+      "eval_recall": 0.8391525731951264,
+      "eval_runtime": 1.8152,
+      "eval_samples_per_second": 219.815,
+      "eval_steps_per_second": 27.546,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 2.4105708599090576,
+      "grad_norm": 1.4582099914550781,
       "learning_rate": 1.75e-05,
-      "loss": 0.1658,
+      "loss": 0.1783,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8595070422535211,
-      "eval_loss": 0.3010116219520569,
-      "eval_precision": 0.8633733523114054,
-      "eval_recall": 0.8559283506092017,
-      "eval_runtime": 5.0413,
-      "eval_samples_per_second": 79.146,
-      "eval_steps_per_second": 9.918,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8649122807017544,
+      "eval_loss": 0.3103856146335602,
+      "eval_precision": 0.8710116366366366,
+      "eval_recall": 0.8594744498999818,
+      "eval_runtime": 1.8138,
+      "eval_samples_per_second": 219.982,
+      "eval_steps_per_second": 27.567,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 4.914756774902344,
+      "grad_norm": 6.738508701324463,
       "learning_rate": 1.5e-05,
-      "loss": 0.1551,
+      "loss": 0.1788,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.8847117794486216,
-      "eval_f1": 0.8622899159663866,
-      "eval_loss": 0.3076874911785126,
-      "eval_precision": 0.8589244307033712,
-      "eval_recall": 0.8659301691216585,
-      "eval_runtime": 5.0475,
-      "eval_samples_per_second": 79.05,
-      "eval_steps_per_second": 9.906,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8689068100358424,
+      "eval_loss": 0.3015482425689697,
+      "eval_precision": 0.863953693884765,
+      "eval_recall": 0.8744771776686671,
+      "eval_runtime": 1.8152,
+      "eval_samples_per_second": 219.806,
+      "eval_steps_per_second": 27.545,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 4.140334606170654,
+      "grad_norm": 2.6257522106170654,
       "learning_rate": 1.25e-05,
-      "loss": 0.1517,
+      "loss": 0.172,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8703663593044124,
-      "eval_loss": 0.30141717195510864,
-      "eval_precision": 0.8789149003479912,
-      "eval_recall": 0.8630205491907619,
-      "eval_runtime": 5.0937,
-      "eval_samples_per_second": 78.332,
-      "eval_steps_per_second": 9.816,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8595070422535211,
+      "eval_loss": 0.3011764883995056,
+      "eval_precision": 0.8633733523114054,
+      "eval_recall": 0.8559283506092017,
+      "eval_runtime": 1.8131,
+      "eval_samples_per_second": 220.062,
+      "eval_steps_per_second": 27.577,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.8993310332298279,
+      "grad_norm": 4.563363552093506,
       "learning_rate": 1e-05,
-      "loss": 0.1532,
+      "loss": 0.1563,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8736504011098378,
-      "eval_loss": 0.3066878616809845,
-      "eval_precision": 0.8718487394957983,
-      "eval_recall": 0.8755228223313329,
-      "eval_runtime": 5.0442,
-      "eval_samples_per_second": 79.1,
-      "eval_steps_per_second": 9.912,
+      "eval_accuracy": 0.8897243107769424,
+      "eval_f1": 0.8695225637671682,
+      "eval_loss": 0.3159307837486267,
+      "eval_precision": 0.8631532846715328,
+      "eval_recall": 0.8769776322967813,
+      "eval_runtime": 1.8172,
+      "eval_samples_per_second": 219.574,
+      "eval_steps_per_second": 27.516,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.4546063542366028,
+      "grad_norm": 1.206107258796692,
       "learning_rate": 7.5e-06,
-      "loss": 0.136,
+      "loss": 0.1512,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8669758137843244,
-      "eval_loss": 0.3174092173576355,
-      "eval_precision": 0.8669758137843244,
-      "eval_recall": 0.8669758137843244,
-      "eval_runtime": 5.0564,
-      "eval_samples_per_second": 78.91,
-      "eval_steps_per_second": 9.889,
+      "eval_accuracy": 0.8847117794486216,
+      "eval_f1": 0.8572517421602788,
+      "eval_loss": 0.32489535212516785,
+      "eval_precision": 0.8679426449878376,
+      "eval_recall": 0.8484269867248591,
+      "eval_runtime": 1.8131,
+      "eval_samples_per_second": 220.063,
+      "eval_steps_per_second": 27.577,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 4.372452259063721,
+      "grad_norm": 4.8656744956970215,
       "learning_rate": 5e-06,
-      "loss": 0.1438,
+      "loss": 0.151,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8663031558425733,
-      "eval_loss": 0.31285080313682556,
-      "eval_precision": 0.8682026944274341,
-      "eval_recall": 0.8644753591562102,
-      "eval_runtime": 5.0509,
-      "eval_samples_per_second": 78.996,
-      "eval_steps_per_second": 9.899,
+      "eval_accuracy": 0.8822055137844611,
+      "eval_f1": 0.855319904024935,
+      "eval_loss": 0.3245084285736084,
+      "eval_precision": 0.862378106322743,
+      "eval_recall": 0.8491543917075832,
+      "eval_runtime": 1.8159,
+      "eval_samples_per_second": 219.726,
+      "eval_steps_per_second": 27.535,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 6.023264408111572,
+      "grad_norm": 3.600020170211792,
       "learning_rate": 2.5e-06,
-      "loss": 0.1507,
+      "loss": 0.1461,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.8922305764411027,
-      "eval_f1": 0.8683279483657071,
-      "eval_loss": 0.316543847322464,
-      "eval_precision": 0.873366724738676,
-      "eval_recall": 0.863747954173486,
-      "eval_runtime": 5.058,
-      "eval_samples_per_second": 78.885,
-      "eval_steps_per_second": 9.885,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8614765038536611,
+      "eval_loss": 0.328171044588089,
+      "eval_precision": 0.8686536646744258,
+      "eval_recall": 0.8552009456264775,
+      "eval_runtime": 1.8138,
+      "eval_samples_per_second": 219.976,
+      "eval_steps_per_second": 27.566,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 4.744778156280518,
+      "grad_norm": 3.394913673400879,
       "learning_rate": 0.0,
-      "loss": 0.1326,
+      "loss": 0.1555,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.8947368421052632,
-      "eval_f1": 0.8723802851224565,
-      "eval_loss": 0.31320512294769287,
-      "eval_precision": 0.8743110838946724,
-      "eval_recall": 0.8705219130751045,
-      "eval_runtime": 5.0604,
-      "eval_samples_per_second": 78.847,
-      "eval_steps_per_second": 9.881,
+      "eval_accuracy": 0.8872180451127819,
+      "eval_f1": 0.8622036668943447,
+      "eval_loss": 0.32483023405075073,
+      "eval_precision": 0.8671602787456446,
+      "eval_recall": 0.8577014002545917,
+      "eval_runtime": 1.8125,
+      "eval_samples_per_second": 220.138,
+      "eval_steps_per_second": 27.586,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
       "total_flos": 8444128359504000.0,
-      "train_loss": 0.232629602463519,
-      "train_runtime": 1905.8691,
-      "train_samples_per_second": 38.177,
-      "train_steps_per_second": 1.28
+      "train_loss": 0.2495564101172275,
+      "train_runtime": 624.4298,
+      "train_samples_per_second": 116.522,
+      "train_steps_per_second": 3.908
     }
   ],
   "logging_steps": 500,