End of training

README.md

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

all_results.json

@@ -1,17 +1,21 @@
 {
+    "accuracy": 0.8615232443125618,
     "epoch": 20.0,
-    "eval_accuracy": 0.9047619047619048,
-    "eval_f1": 0.8878351186601172,
-    "eval_loss": 0.7535876035690308,
-    "eval_precision": 0.879776516905975,
-    "eval_recall": 0.8976177486815784,
-    "eval_runtime": 4.7232,
+    "eval_accuracy": 0.8546365914786967,
+    "eval_f1": 0.8263655462184873,
+    "eval_loss": 0.35401326417922974,
+    "eval_precision": 0.8233396753671443,
+    "eval_recall": 0.8296508456082925,
+    "eval_runtime": 5.0488,
     "eval_samples": 399,
-    "eval_samples_per_second": 84.476,
-    "eval_steps_per_second": 10.586,
-    "train_loss": 0.06173487283655855,
-    "train_runtime": 2892.3535,
+    "eval_samples_per_second": 79.029,
+    "eval_steps_per_second": 9.903,
+    "f1": 0.8344251555846709,
+    "precision": 0.8325509007667684,
+    "recall": 0.8363917467548971,
+    "train_loss": 0.35800845193081215,
+    "train_runtime": 2113.1391,
     "train_samples": 3638,
-    "train_samples_per_second": 25.156,
-    "train_steps_per_second": 0.844
+    "train_samples_per_second": 34.432,
+    "train_steps_per_second": 1.155
 }

eval_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 20.0,
-    "eval_accuracy": 0.9047619047619048,
-    "eval_f1": 0.8878351186601172,
-    "eval_loss": 0.7535876035690308,
-    "eval_precision": 0.879776516905975,
-    "eval_recall": 0.8976177486815784,
-    "eval_runtime": 4.7232,
+    "eval_accuracy": 0.8546365914786967,
+    "eval_f1": 0.8263655462184873,
+    "eval_loss": 0.35401326417922974,
+    "eval_precision": 0.8233396753671443,
+    "eval_recall": 0.8296508456082925,
+    "eval_runtime": 5.0488,
     "eval_samples": 399,
-    "eval_samples_per_second": 84.476,
-    "eval_steps_per_second": 10.586
+    "eval_samples_per_second": 79.029,
+    "eval_steps_per_second": 9.903
 }

predict_results.txt

@@ -1,6 +1,6 @@
 index	prediction
 0	1
-1	1
+1	0
 2	1
 3	1
 4	0
@@ -14,21 +14,21 @@ index	prediction
 12	1
 13	1
 14	1
-15	0
+15	1
 16	1
 17	1
 18	1
 19	1
 20	1
 21	1
-22	1
+22	0
 23	1
-24	0
+24	1
 25	1
 26	1
 27	1
 28	1
-29	1
+29	0
 30	1
 31	1
 32	1
@@ -38,39 +38,39 @@ index	prediction
 36	1
 37	1
 38	1
-39	1
+39	0
 40	1
-41	1
-42	1
-43	1
-44	1
-45	1
+41	0
+42	0
+43	0
+44	0
+45	0
 46	1
 47	1
 48	1
 49	0
 50	1
 51	1
-52	1
+52	0
 53	1
 54	1
 55	1
-56	1
+56	0
 57	0
 58	1
 59	1
-60	1
+60	0
 61	1
 62	1
 63	1
-64	1
+64	0
 65	1
 66	1
 67	1
 68	1
 69	1
 70	1
-71	0
+71	1
 72	1
 73	1
 74	1
@@ -79,31 +79,31 @@ index	prediction
 77	0
 78	1
 79	1
-80	1
-81	0
+80	0
+81	1
 82	1
 83	1
 84	1
 85	0
-86	1
+86	0
 87	1
 88	1
 89	1
 90	1
-91	1
-92	1
-93	1
+91	0
+92	0
+93	0
 94	1
 95	1
 96	1
 97	0
-98	1
+98	0
 99	0
 100	0
 101	0
 102	1
 103	1
-104	1
+104	0
 105	1
 106	1
 107	1
@@ -112,7 +112,7 @@ index	prediction
 110	1
 111	1
 112	1
-113	0
+113	1
 114	1
 115	1
 116	1
@@ -124,7 +124,7 @@ index	prediction
 122	1
 123	1
 124	1
-125	0
+125	1
 126	1
 127	1
 128	1
@@ -134,7 +134,7 @@ index	prediction
 132	1
 133	1
 134	1
-135	1
+135	0
 136	0
 137	1
 138	1
@@ -146,7 +146,7 @@ index	prediction
 144	1
 145	1
 146	1
-147	1
+147	0
 148	1
 149	1
 150	1
@@ -163,24 +163,24 @@ index	prediction
 161	1
 162	1
 163	1
-164	1
+164	0
 165	0
 166	1
 167	1
-168	1
-169	1
+168	0
+169	0
 170	1
 171	1
-172	1
+172	0
 173	0
-174	1
+174	0
 175	1
-176	1
+176	0
 177	0
 178	1
 179	1
 180	1
-181	0
+181	1
 182	1
 183	1
 184	1
@@ -189,7 +189,7 @@ index	prediction
 187	1
 188	1
 189	1
-190	1
+190	0
 191	1
 192	1
 193	1
@@ -217,22 +217,22 @@ index	prediction
 215	1
 216	0
 217	0
-218	1
+218	0
 219	1
 220	0
 221	1
 222	1
 223	1
-224	0
+224	1
 225	1
 226	0
 227	0
-228	1
-229	1
+228	0
+229	0
 230	1
 231	1
-232	1
-233	1
+232	0
+233	0
 234	1
 235	1
 236	1
@@ -248,7 +248,7 @@ index	prediction
 246	0
 247	1
 248	1
-249	0
+249	1
 250	0
 251	1
 252	1
@@ -260,19 +260,19 @@ index	prediction
 258	1
 259	1
 260	1
-261	1
+261	0
 262	1
 263	1
 264	1
-265	1
+265	0
 266	1
 267	1
 268	1
 269	1
 270	1
 271	1
-272	0
-273	1
+272	1
+273	0
 274	1
 275	1
 276	1
@@ -292,11 +292,11 @@ index	prediction
 290	1
 291	1
 292	1
-293	1
+293	0
 294	1
 295	1
 296	1
-297	0
+297	1
 298	0
 299	0
 300	0
@@ -313,9 +313,9 @@ index	prediction
 311	0
 312	0
 313	0
-314	0
+314	1
 315	0
-316	0
+316	1
 317	0
 318	1
 319	0
@@ -327,30 +327,30 @@ index	prediction
 325	0
 326	0
 327	0
-328	1
+328	0
 329	0
-330	1
-331	1
+330	0
+331	0
 332	0
 333	0
 334	0
 335	0
 336	0
 337	0
-338	1
+338	0
 339	0
 340	0
 341	0
 342	0
 343	0
-344	0
+344	1
 345	0
-346	1
+346	0
 347	0
 348	0
 349	0
 350	0
-351	0
+351	1
 352	0
 353	0
 354	0
@@ -375,7 +375,7 @@ index	prediction
 373	0
 374	0
 375	0
-376	0
+376	1
 377	0
 378	0
 379	0
@@ -397,11 +397,11 @@ index	prediction
 395	0
 396	0
 397	0
-398	1
+398	0
 399	0
 400	0
-401	0
-402	1
+401	1
+402	0
 403	0
 404	0
 405	0
@@ -419,12 +419,12 @@ index	prediction
 417	0
 418	0
 419	0
-420	1
-421	0
+420	0
+421	1
 422	0
 423	0
 424	0
-425	0
+425	1
 426	0
 427	0
 428	0
@@ -438,7 +438,7 @@ index	prediction
 436	0
 437	0
 438	0
-439	0
+439	1
 440	0
 441	0
 442	0
@@ -446,7 +446,7 @@ index	prediction
 444	0
 445	0
 446	0
-447	1
+447	0
 448	0
 449	0
 450	0
@@ -454,7 +454,7 @@ index	prediction
 452	0
 453	0
 454	0
-455	0
+455	1
 456	0
 457	0
 458	0
@@ -465,7 +465,7 @@ index	prediction
 463	0
 464	0
 465	0
-466	0
+466	1
 467	0
 468	0
 469	0
@@ -486,12 +486,12 @@ index	prediction
 484	0
 485	0
 486	0
-487	0
+487	1
 488	0
 489	0
 490	0
 491	0
-492	0
+492	1
 493	0
 494	0
 495	0
@@ -502,7 +502,7 @@ index	prediction
 500	0
 501	0
 502	0
-503	0
+503	1
 504	0
 505	0
 506	0
@@ -510,7 +510,7 @@ index	prediction
 508	0
 509	0
 510	0
-511	1
+511	0
 512	0
 513	0
 514	0
@@ -521,9 +521,9 @@ index	prediction
 519	0
 520	0
 521	0
-522	0
+522	1
 523	0
-524	0
+524	1
 525	0
 526	0
 527	0
@@ -533,9 +533,9 @@ index	prediction
 531	0
 532	0
 533	0
-534	0
+534	1
 535	0
-536	1
+536	0
 537	0
 538	0
 539	0
@@ -559,7 +559,7 @@ index	prediction
 557	0
 558	0
 559	0
-560	1
+560	0
 561	0
 562	0
 563	0
@@ -587,20 +587,20 @@ index	prediction
 585	0
 586	0
 587	0
-588	0
+588	1
 589	0
 590	0
 591	0
 592	0
 593	0
 594	0
-595	0
-596	0
+595	1
+596	1
 597	0
 598	0
 599	0
 600	0
-601	0
+601	1
 602	0
 603	0
 604	0
@@ -623,18 +623,18 @@ index	prediction
 621	1
 622	0
 623	0
-624	0
+624	1
 625	0
 626	0
 627	0
-628	0
+628	1
 629	0
 630	0
 631	0
-632	1
+632	0
 633	1
 634	0
-635	0
+635	1
 636	0
 637	0
 638	0
@@ -653,7 +653,7 @@ index	prediction
 651	0
 652	1
 653	0
-654	0
+654	1
 655	0
 656	0
 657	1
@@ -662,7 +662,7 @@ index	prediction
 660	0
 661	0
 662	0
-663	1
+663	0
 664	0
 665	0
 666	0
@@ -702,7 +702,7 @@ index	prediction
 700	0
 701	0
 702	0
-703	1
+703	0
 704	0
 705	0
 706	0
@@ -755,7 +755,7 @@ index	prediction
 753	0
 754	0
 755	0
-756	0
+756	1
 757	0
 758	0
 759	0
@@ -764,14 +764,14 @@ index	prediction
 762	0
 763	0
 764	0
-765	0
+765	1
 766	0
 767	0
 768	0
 769	0
 770	1
 771	0
-772	0
+772	1
 773	0
 774	0
 775	0
@@ -808,12 +808,12 @@ index	prediction
 806	0
 807	0
 808	0
-809	0
+809	1
 810	0
 811	0
 812	0
 813	0
-814	0
+814	1
 815	0
 816	0
 817	0
@@ -831,7 +831,7 @@ index	prediction
 829	0
 830	0
 831	0
-832	0
+832	1
 833	1
 834	0
 835	0
@@ -840,7 +840,7 @@ index	prediction
 838	0
 839	0
 840	0
-841	1
+841	0
 842	0
 843	0
 844	0
@@ -867,7 +867,7 @@ index	prediction
 865	0
 866	0
 867	0
-868	1
+868	0
 869	0
 870	0
 871	0
@@ -883,7 +883,7 @@ index	prediction
 881	0
 882	0
 883	0
-884	0
+884	1
 885	0
 886	0
 887	0
@@ -893,10 +893,10 @@ index	prediction
 891	1
 892	0
 893	0
-894	0
+894	1
 895	0
 896	0
-897	0
+897	1
 898	0
 899	0
 900	0
@@ -905,13 +905,13 @@ index	prediction
 903	0
 904	0
 905	0
-906	0
+906	1
 907	1
 908	0
 909	0
 910	0
 911	0
-912	0
+912	1
 913	0
 914	0
 915	0
@@ -949,20 +949,20 @@ index	prediction
 947	0
 948	0
 949	0
-950	0
+950	1
 951	0
 952	0
 953	0
 954	0
-955	1
+955	0
 956	0
-957	0
+957	1
 958	0
 959	0
 960	0
 961	0
 962	0
-963	1
+963	0
 964	0
 965	0
 966	0
@@ -991,18 +991,18 @@ index	prediction
 989	0
 990	0
 991	1
-992	0
+992	1
 993	0
 994	0
-995	0
+995	1
 996	0
 997	0
-998	0
+998	1
 999	1
 1000	0
 1001	0
 1002	0
-1003	0
+1003	1
 1004	0
 1005	0
 1006	0

runs/May15_17-27-18_indolem-petl-vm/events.out.tfevents.1715796187.indolem-petl-vm.579186.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c10c37be125002ed98cbc0b129247fa6687b064aa0f13207c7aee099ebc758a
+size 560

test_results.json

@@ -0,0 +1,6 @@
+{
+    "accuracy": 0.8615232443125618,
+    "f1": 0.8344251555846709,
+    "precision": 0.8325509007667684,
+    "recall": 0.8363917467548971
+}

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 20.0,
-    "train_loss": 0.06173487283655855,
-    "train_runtime": 2892.3535,
+    "train_loss": 0.35800845193081215,
+    "train_runtime": 2113.1391,
     "train_samples": 3638,
-    "train_samples_per_second": 25.156,
-    "train_steps_per_second": 0.844
+    "train_samples_per_second": 34.432,
+    "train_steps_per_second": 1.155
 }

trainer_state.json

@@ -10,392 +10,392 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 4.053804397583008,
+      "grad_norm": 3.0131800174713135,
       "learning_rate": 4.75e-05,
-      "loss": 0.4355,
+      "loss": 0.5623,
       "step": 122
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.8696741854636592,
-      "eval_f1": 0.835906358747232,
-      "eval_loss": 0.3243214786052704,
-      "eval_precision": 0.853844109243139,
-      "eval_recall": 0.8227859610838335,
-      "eval_runtime": 4.6306,
-      "eval_samples_per_second": 86.166,
-      "eval_steps_per_second": 10.798,
+      "eval_accuracy": 0.7167919799498746,
+      "eval_f1": 0.5794790005316321,
+      "eval_loss": 0.5053456425666809,
+      "eval_precision": 0.6409822866344606,
+      "eval_recall": 0.5796053827968721,
+      "eval_runtime": 5.6071,
+      "eval_samples_per_second": 71.159,
+      "eval_steps_per_second": 8.917,
       "step": 122
     },
     {
       "epoch": 2.0,
-      "grad_norm": 23.150257110595703,
+      "grad_norm": 5.634490966796875,
       "learning_rate": 4.5e-05,
-      "loss": 0.2295,
+      "loss": 0.518,
       "step": 244
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8701248742380304,
-      "eval_loss": 0.3046626150608063,
-      "eval_precision": 0.8624507874015748,
-      "eval_recall": 0.8794780869248955,
-      "eval_runtime": 4.9587,
-      "eval_samples_per_second": 80.464,
-      "eval_steps_per_second": 10.083,
+      "eval_accuracy": 0.7293233082706767,
+      "eval_f1": 0.599784530797236,
+      "eval_loss": 0.4860531687736511,
+      "eval_precision": 0.6673625792811839,
+      "eval_recall": 0.5959719949081652,
+      "eval_runtime": 5.7755,
+      "eval_samples_per_second": 69.085,
+      "eval_steps_per_second": 8.657,
       "step": 244
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.44390636682510376,
+      "grad_norm": 1.959808111190796,
       "learning_rate": 4.25e-05,
-      "loss": 0.1337,
+      "loss": 0.4835,
       "step": 366
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.879667048676036,
-      "eval_loss": 0.3747338354587555,
-      "eval_precision": 0.8778361344537815,
-      "eval_recall": 0.8815693762502272,
-      "eval_runtime": 4.996,
-      "eval_samples_per_second": 79.865,
-      "eval_steps_per_second": 10.008,
+      "eval_accuracy": 0.7694235588972431,
+      "eval_f1": 0.7145034843205575,
+      "eval_loss": 0.45518842339515686,
+      "eval_precision": 0.7210824478299833,
+      "eval_recall": 0.7093562465902892,
+      "eval_runtime": 5.2584,
+      "eval_samples_per_second": 75.878,
+      "eval_steps_per_second": 9.509,
       "step": 366
     },
     {
       "epoch": 4.0,
-      "grad_norm": 15.949886322021484,
+      "grad_norm": 4.635169506072998,
       "learning_rate": 4e-05,
-      "loss": 0.1038,
+      "loss": 0.4497,
       "step": 488
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.8822055137844611,
-      "eval_f1": 0.8651222336500356,
-      "eval_loss": 0.41882890462875366,
-      "eval_precision": 0.8518339768339769,
-      "eval_recall": 0.8866612111292962,
-      "eval_runtime": 5.018,
-      "eval_samples_per_second": 79.514,
-      "eval_steps_per_second": 9.964,
+      "eval_accuracy": 0.7944862155388471,
+      "eval_f1": 0.7520912893253319,
+      "eval_loss": 0.4223441481590271,
+      "eval_precision": 0.7520912893253319,
+      "eval_recall": 0.7520912893253319,
+      "eval_runtime": 5.0487,
+      "eval_samples_per_second": 79.03,
+      "eval_steps_per_second": 9.903,
       "step": 488
     },
     {
       "epoch": 5.0,
-      "grad_norm": 3.7783217430114746,
+      "grad_norm": 8.219679832458496,
       "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.072,
+      "loss": 0.4266,
       "step": 610
     },
     {
       "epoch": 5.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_f1": 0.8622036668943447,
-      "eval_loss": 0.6270534992218018,
-      "eval_precision": 0.8671602787456446,
-      "eval_recall": 0.8577014002545917,
-      "eval_runtime": 4.9576,
-      "eval_samples_per_second": 80.483,
-      "eval_steps_per_second": 10.086,
+      "eval_accuracy": 0.8170426065162907,
+      "eval_f1": 0.7740779522978476,
+      "eval_loss": 0.399569034576416,
+      "eval_precision": 0.7814051164566629,
+      "eval_recall": 0.7680487361338425,
+      "eval_runtime": 5.0767,
+      "eval_samples_per_second": 78.595,
+      "eval_steps_per_second": 9.849,
       "step": 610
     },
     {
       "epoch": 6.0,
-      "grad_norm": 44.57243347167969,
+      "grad_norm": 4.150725841522217,
       "learning_rate": 3.5e-05,
-      "loss": 0.0462,
+      "loss": 0.3907,
       "step": 732
     },
     {
       "epoch": 6.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8695225637671682,
-      "eval_loss": 0.6129250526428223,
-      "eval_precision": 0.8631532846715328,
-      "eval_recall": 0.8769776322967813,
-      "eval_runtime": 4.9521,
-      "eval_samples_per_second": 80.572,
-      "eval_steps_per_second": 10.097,
+      "eval_accuracy": 0.8195488721804511,
+      "eval_f1": 0.784453781512605,
+      "eval_loss": 0.3830115497112274,
+      "eval_precision": 0.7818241274748796,
+      "eval_recall": 0.787324968176032,
+      "eval_runtime": 5.0718,
+      "eval_samples_per_second": 78.67,
+      "eval_steps_per_second": 9.858,
       "step": 732
     },
     {
       "epoch": 7.0,
-      "grad_norm": 0.024074144661426544,
+      "grad_norm": 3.297985076904297,
       "learning_rate": 3.2500000000000004e-05,
-      "loss": 0.0459,
+      "loss": 0.3742,
       "step": 854
     },
     {
       "epoch": 7.0,
-      "eval_accuracy": 0.8897243107769424,
-      "eval_f1": 0.8649122807017544,
-      "eval_loss": 0.5890637636184692,
-      "eval_precision": 0.8710116366366366,
-      "eval_recall": 0.8594744498999818,
-      "eval_runtime": 4.9483,
-      "eval_samples_per_second": 80.633,
-      "eval_steps_per_second": 10.104,
+      "eval_accuracy": 0.8345864661654135,
+      "eval_f1": 0.798423147581139,
+      "eval_loss": 0.3684135675430298,
+      "eval_precision": 0.8016430472182685,
+      "eval_recall": 0.7954628114202582,
+      "eval_runtime": 5.0743,
+      "eval_samples_per_second": 78.632,
+      "eval_steps_per_second": 9.854,
       "step": 854
     },
     {
       "epoch": 8.0,
-      "grad_norm": 0.022918157279491425,
+      "grad_norm": 8.395323753356934,
       "learning_rate": 3e-05,
-      "loss": 0.0391,
+      "loss": 0.3616,
       "step": 976
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8872180451127819,
-      "eval_f1": 0.8680720368560659,
-      "eval_loss": 0.5972921252250671,
-      "eval_precision": 0.8587217615098657,
-      "eval_recall": 0.8802054919076197,
-      "eval_runtime": 4.9878,
-      "eval_samples_per_second": 79.995,
-      "eval_steps_per_second": 10.024,
+      "eval_accuracy": 0.8270676691729323,
+      "eval_f1": 0.7967966933608887,
+      "eval_loss": 0.3719731867313385,
+      "eval_precision": 0.7902444649446494,
+      "eval_recall": 0.8051463902527732,
+      "eval_runtime": 5.0484,
+      "eval_samples_per_second": 79.035,
+      "eval_steps_per_second": 9.904,
       "step": 976
     },
     {
       "epoch": 9.0,
-      "grad_norm": 106.23094177246094,
+      "grad_norm": 3.748974561691284,
       "learning_rate": 2.7500000000000004e-05,
-      "loss": 0.0307,
+      "loss": 0.3294,
       "step": 1098
     },
     {
       "epoch": 9.0,
-      "eval_accuracy": 0.87468671679198,
-      "eval_f1": 0.8584865509022812,
-      "eval_loss": 0.7086873054504395,
-      "eval_precision": 0.8441043083900227,
-      "eval_recall": 0.8863429714493545,
-      "eval_runtime": 4.9636,
-      "eval_samples_per_second": 80.386,
-      "eval_steps_per_second": 10.073,
+      "eval_accuracy": 0.8370927318295739,
+      "eval_f1": 0.8076965854743632,
+      "eval_loss": 0.36888691782951355,
+      "eval_precision": 0.8018925518925519,
+      "eval_recall": 0.8147390434624477,
+      "eval_runtime": 5.0543,
+      "eval_samples_per_second": 78.943,
+      "eval_steps_per_second": 9.893,
       "step": 1098
     },
     {
       "epoch": 10.0,
-      "grad_norm": 0.024997469037771225,
+      "grad_norm": 9.309541702270508,
       "learning_rate": 2.5e-05,
-      "loss": 0.0199,
+      "loss": 0.3207,
       "step": 1220
     },
     {
       "epoch": 10.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8717112228173498,
-      "eval_loss": 0.7264124155044556,
-      "eval_precision": 0.8869295958279009,
-      "eval_recall": 0.8597926895799237,
-      "eval_runtime": 4.965,
-      "eval_samples_per_second": 80.363,
-      "eval_steps_per_second": 10.07,
+      "eval_accuracy": 0.8395989974937343,
+      "eval_f1": 0.8110907261644079,
+      "eval_loss": 0.36315786838531494,
+      "eval_precision": 0.8046983557202408,
+      "eval_recall": 0.819012547735952,
+      "eval_runtime": 5.0709,
+      "eval_samples_per_second": 78.684,
+      "eval_steps_per_second": 9.86,
       "step": 1220
     },
     {
       "epoch": 11.0,
-      "grad_norm": 0.004392046481370926,
+      "grad_norm": 1.2568168640136719,
       "learning_rate": 2.25e-05,
-      "loss": 0.0105,
+      "loss": 0.3214,
       "step": 1342
     },
     {
       "epoch": 11.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8757339815412664,
-      "eval_loss": 0.6738360524177551,
-      "eval_precision": 0.8766906299500427,
-      "eval_recall": 0.8747954173486088,
-      "eval_runtime": 4.9486,
-      "eval_samples_per_second": 80.628,
-      "eval_steps_per_second": 10.104,
+      "eval_accuracy": 0.8370927318295739,
+      "eval_f1": 0.8085765951950401,
+      "eval_loss": 0.3577338457107544,
+      "eval_precision": 0.8017470018450185,
+      "eval_recall": 0.817239498090562,
+      "eval_runtime": 5.1071,
+      "eval_samples_per_second": 78.126,
+      "eval_steps_per_second": 9.79,
       "step": 1342
     },
     {
       "epoch": 12.0,
-      "grad_norm": 0.004026818089187145,
+      "grad_norm": 1.915198802947998,
       "learning_rate": 2e-05,
-      "loss": 0.0131,
+      "loss": 0.3167,
       "step": 1464
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.899749373433584,
-      "eval_f1": 0.882467302933899,
-      "eval_loss": 0.7488105297088623,
-      "eval_precision": 0.8732988802756245,
-      "eval_recall": 0.8940716493907983,
-      "eval_runtime": 4.9589,
-      "eval_samples_per_second": 80.462,
-      "eval_steps_per_second": 10.083,
+      "eval_accuracy": 0.8395989974937343,
+      "eval_f1": 0.8119476846942383,
+      "eval_loss": 0.36069995164871216,
+      "eval_precision": 0.8045650301464256,
+      "eval_recall": 0.8215130023640662,
+      "eval_runtime": 5.0598,
+      "eval_samples_per_second": 78.857,
+      "eval_steps_per_second": 9.882,
       "step": 1464
     },
     {
       "epoch": 13.0,
-      "grad_norm": 0.004543425515294075,
+      "grad_norm": 1.9545631408691406,
       "learning_rate": 1.75e-05,
-      "loss": 0.0102,
+      "loss": 0.289,
       "step": 1586
     },
     {
       "epoch": 13.0,
-      "eval_accuracy": 0.8972431077694235,
-      "eval_f1": 0.8792560061999484,
-      "eval_loss": 0.7154756784439087,
-      "eval_precision": 0.8707622232472325,
-      "eval_recall": 0.889798145117294,
-      "eval_runtime": 4.9704,
-      "eval_samples_per_second": 80.275,
-      "eval_steps_per_second": 10.06,
+      "eval_accuracy": 0.8345864661654135,
+      "eval_f1": 0.8060710498409331,
+      "eval_loss": 0.3684280812740326,
+      "eval_precision": 0.7988372093023256,
+      "eval_recall": 0.8154664484451719,
+      "eval_runtime": 5.1019,
+      "eval_samples_per_second": 78.206,
+      "eval_steps_per_second": 9.8,
       "step": 1586
     },
     {
       "epoch": 14.0,
-      "grad_norm": 0.0037931231781840324,
+      "grad_norm": 5.748187065124512,
       "learning_rate": 1.5e-05,
-      "loss": 0.0061,
+      "loss": 0.2997,
       "step": 1708
     },
     {
       "epoch": 14.0,
-      "eval_accuracy": 0.9072681704260651,
-      "eval_f1": 0.8894993300948346,
-      "eval_loss": 0.7196279168128967,
-      "eval_precision": 0.8850535598035154,
-      "eval_recall": 0.8943898890707401,
-      "eval_runtime": 4.9583,
-      "eval_samples_per_second": 80.47,
-      "eval_steps_per_second": 10.084,
+      "eval_accuracy": 0.849624060150376,
+      "eval_f1": 0.8176861216035092,
+      "eval_loss": 0.3479882776737213,
+      "eval_precision": 0.8193355786895284,
+      "eval_recall": 0.8161029278050556,
+      "eval_runtime": 5.0557,
+      "eval_samples_per_second": 78.92,
+      "eval_steps_per_second": 9.89,
       "step": 1708
     },
     {
       "epoch": 15.0,
-      "grad_norm": 0.0027608012314885855,
+      "grad_norm": 4.010083198547363,
       "learning_rate": 1.25e-05,
-      "loss": 0.0138,
+      "loss": 0.2986,
       "step": 1830
     },
     {
       "epoch": 15.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.884617951284618,
-      "eval_loss": 0.7618029713630676,
-      "eval_precision": 0.8772893772893773,
-      "eval_recall": 0.8933442444080741,
-      "eval_runtime": 5.0532,
-      "eval_samples_per_second": 78.96,
-      "eval_steps_per_second": 9.895,
+      "eval_accuracy": 0.849624060150376,
+      "eval_f1": 0.821236559139785,
+      "eval_loss": 0.35758015513420105,
+      "eval_precision": 0.8169406150583245,
+      "eval_recall": 0.8261047463175123,
+      "eval_runtime": 5.0955,
+      "eval_samples_per_second": 78.304,
+      "eval_steps_per_second": 9.813,
       "step": 1830
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.0022813216783106327,
+      "grad_norm": 0.9220337271690369,
       "learning_rate": 1e-05,
-      "loss": 0.0075,
+      "loss": 0.2914,
       "step": 1952
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8873149414352814,
-      "eval_loss": 0.7252941727638245,
-      "eval_precision": 0.8806277372262774,
-      "eval_recall": 0.8951172940534643,
-      "eval_runtime": 5.598,
-      "eval_samples_per_second": 71.275,
-      "eval_steps_per_second": 8.932,
+      "eval_accuracy": 0.849624060150376,
+      "eval_f1": 0.8195005730140539,
+      "eval_loss": 0.34965991973876953,
+      "eval_precision": 0.8179621848739496,
+      "eval_recall": 0.8211038370612839,
+      "eval_runtime": 5.0617,
+      "eval_samples_per_second": 78.827,
+      "eval_steps_per_second": 9.878,
       "step": 1952
     },
     {
       "epoch": 17.0,
-      "grad_norm": 15.638340950012207,
+      "grad_norm": 1.7026562690734863,
       "learning_rate": 7.5e-06,
-      "loss": 0.0063,
+      "loss": 0.278,
       "step": 2074
     },
     {
       "epoch": 17.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8840781602687784,
-      "eval_loss": 0.7560042142868042,
-      "eval_precision": 0.87816715542522,
-      "eval_recall": 0.89084378977996,
-      "eval_runtime": 4.9548,
-      "eval_samples_per_second": 80.528,
-      "eval_steps_per_second": 10.091,
+      "eval_accuracy": 0.8521303258145363,
+      "eval_f1": 0.8229427559286084,
+      "eval_loss": 0.3539772927761078,
+      "eval_precision": 0.8206541218637993,
+      "eval_recall": 0.8253773413347881,
+      "eval_runtime": 5.1199,
+      "eval_samples_per_second": 77.931,
+      "eval_steps_per_second": 9.766,
       "step": 2074
     },
     {
       "epoch": 18.0,
-      "grad_norm": 0.002121408935636282,
+      "grad_norm": 5.839470863342285,
       "learning_rate": 5e-06,
-      "loss": 0.0066,
+      "loss": 0.2887,
       "step": 2196
     },
     {
       "epoch": 18.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.8856624319419237,
-      "eval_loss": 0.748332142829895,
-      "eval_precision": 0.8758364312267658,
-      "eval_recall": 0.8983451536643026,
-      "eval_runtime": 4.9482,
-      "eval_samples_per_second": 80.635,
-      "eval_steps_per_second": 10.105,
+      "eval_accuracy": 0.8521303258145363,
+      "eval_f1": 0.8229427559286084,
+      "eval_loss": 0.35161107778549194,
+      "eval_precision": 0.8206541218637993,
+      "eval_recall": 0.8253773413347881,
+      "eval_runtime": 5.1154,
+      "eval_samples_per_second": 77.999,
+      "eval_steps_per_second": 9.774,
       "step": 2196
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.004570267163217068,
+      "grad_norm": 7.782900810241699,
       "learning_rate": 2.5e-06,
-      "loss": 0.0023,
+      "loss": 0.2829,
       "step": 2318
     },
     {
       "epoch": 19.0,
-      "eval_accuracy": 0.9022556390977443,
-      "eval_f1": 0.884617951284618,
-      "eval_loss": 0.7535205483436584,
-      "eval_precision": 0.8772893772893773,
-      "eval_recall": 0.8933442444080741,
-      "eval_runtime": 4.9538,
-      "eval_samples_per_second": 80.543,
-      "eval_steps_per_second": 10.093,
+      "eval_accuracy": 0.8521303258145363,
+      "eval_f1": 0.8229427559286084,
+      "eval_loss": 0.35369938611984253,
+      "eval_precision": 0.8206541218637993,
+      "eval_recall": 0.8253773413347881,
+      "eval_runtime": 5.0565,
+      "eval_samples_per_second": 78.908,
+      "eval_steps_per_second": 9.888,
       "step": 2318
     },
     {
       "epoch": 20.0,
-      "grad_norm": 0.0018295175395905972,
+      "grad_norm": 5.059621334075928,
       "learning_rate": 0.0,
-      "loss": 0.0021,
+      "loss": 0.2771,
       "step": 2440
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.9047619047619048,
-      "eval_f1": 0.8878351186601172,
-      "eval_loss": 0.7535876035690308,
-      "eval_precision": 0.879776516905975,
-      "eval_recall": 0.8976177486815784,
-      "eval_runtime": 4.9557,
-      "eval_samples_per_second": 80.514,
-      "eval_steps_per_second": 10.089,
+      "eval_accuracy": 0.8546365914786967,
+      "eval_f1": 0.8263655462184873,
+      "eval_loss": 0.35401326417922974,
+      "eval_precision": 0.8233396753671443,
+      "eval_recall": 0.8296508456082925,
+      "eval_runtime": 5.0854,
+      "eval_samples_per_second": 78.459,
+      "eval_steps_per_second": 9.832,
       "step": 2440
     },
     {
       "epoch": 20.0,
       "step": 2440,
-      "total_flos": 7584162436176000.0,
-      "train_loss": 0.06173487283655855,
-      "train_runtime": 2892.3535,
-      "train_samples_per_second": 25.156,
-      "train_steps_per_second": 0.844
+      "total_flos": 7597037114448000.0,
+      "train_loss": 0.35800845193081215,
+      "train_runtime": 2113.1391,
+      "train_samples_per_second": 34.432,
+      "train_steps_per_second": 1.155
     }
   ],
   "logging_steps": 500,
@@ -403,7 +403,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 20,
   "save_steps": 500,
-  "total_flos": 7584162436176000.0,
+  "total_flos": 7597037114448000.0,
   "train_batch_size": 30,
   "trial_name": null,
   "trial_params": null