Training in progress, epoch 2, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ae49f66fa40de83d0480645b761e8a753be0e28dae3d7766c6d3f8a1baf76758
+oid sha256:97891a60ef32278715fe2fd622467a12763337c4f29242e0f47e6fa912ad3bfe
 size 800116456

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3edba13ad09666a36a27cebd58dda9e758e7e87ddd5bef0b427617e52e33efac
+oid sha256:5f4c7a6cc33054e1b8f261b1bf71a365bc3b95974dbb2778025c3addcdbbc007
 size 406743860

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21baf4fcb4a38e73544f06f6ef7f7a2d5294a4ec46aaecfe18f9b11ab06b0be2
+oid sha256:863886f073e9c5c689c6ec3b1475c24da88e3fc752e2e629afed4a2fd70f22d2
 size 15984

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:99ac34f10b5b4b30f9a14a4b42e76496b127d0b2a59dc004d710f643389b83f4
+oid sha256:a4274bf3f4a7c08973ea99b74ce6035a79abe0939430e685f8d3ba7036343982
 size 15984

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:38b88187ca347822e8568870ea86a324e54f63d62d2f024d8758ac5294d40a4a
+oid sha256:301fa96e37e1ba2ee75f0fbc01f26e4f12cd1597463ce19b79f6f19cd894f004
 size 15984

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77fbe34d05fd83a2009eafa4d761d1a348b27a341023148bb5b1a192c63511e7
+oid sha256:05b1fde6899045457175e8c52d857720181b80762222d8f14a4261f8c0678492
 size 15984

last-checkpoint/rng_state_4.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cef45ca40916c741f85a2a224b408d1dedf2b7d0472602a0f734da65b3437ad6
+oid sha256:4ee6337bf076a6e596f313be0012d284f2fba533e1d8c79adb3b2cd349de8f1f
 size 15984

last-checkpoint/rng_state_5.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0987883926c20bb8f52b158ac1c13ef4cc918beb36cd85f83e47464909ebac27
+oid sha256:fed1407b50de8e21d020387b421a3eeaf16b75cb4dc805daa58a6a4650647dfe
 size 15984

last-checkpoint/rng_state_6.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6225dc3a627ae28f20a3d380b0ed35a7dd891b3f65db0e8bbd21c856decaf133
+oid sha256:f4ea5f0174438e143d41f6c7ed1b169273deb7031a4afb3e805dfd389ef07e74
 size 15984

last-checkpoint/rng_state_7.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3f877841f36b3220621f0ed636d6be2da52fa59ed16112fd3dca52767c19bb52
+oid sha256:c2f1bd34cca459054a035c07dfbf38f53c01552d6dfad51dfd430634d69f7f8c
 size 15984

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4bca0f9e13713e07601b4dc0bd9d3aa19ee7ad3e516a1c659b4636fc2a3b9f2
+oid sha256:88e7d89c7e9ab9880d864078f36a7dcd958bf79763744c29456aa56ddb874c0f
 size 1064

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.9949937421777222,
+  "epoch": 2.9937421777221527,
   "eval_steps": 500,
-  "global_step": 398,
+  "global_step": 597,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -2793,6 +2793,1399 @@
       "learning_rate": 0.00010205765019574084,
       "loss": 1.3904,
       "step": 398
+    },
+    {
+      "epoch": 2.002503128911139,
+      "grad_norm": 9.335354804992676,
+      "learning_rate": 0.00010152883234697336,
+      "loss": 2.9917,
+      "step": 399
+    },
+    {
+      "epoch": 2.0075093867334166,
+      "grad_norm": 3.17944598197937,
+      "learning_rate": 0.000101,
+      "loss": 1.1571,
+      "step": 400
+    },
+    {
+      "epoch": 2.0125156445556946,
+      "grad_norm": 3.515106678009033,
+      "learning_rate": 0.00010047116765302661,
+      "loss": 1.1999,
+      "step": 401
+    },
+    {
+      "epoch": 2.0175219023779727,
+      "grad_norm": 4.571992874145508,
+      "learning_rate": 9.994234980425921e-05,
+      "loss": 1.2175,
+      "step": 402
+    },
+    {
+      "epoch": 2.0225281602002503,
+      "grad_norm": 4.036806106567383,
+      "learning_rate": 9.941356095150613e-05,
+      "loss": 1.114,
+      "step": 403
+    },
+    {
+      "epoch": 2.0275344180225283,
+      "grad_norm": 3.547945499420166,
+      "learning_rate": 9.888481559178096e-05,
+      "loss": 1.2359,
+      "step": 404
+    },
+    {
+      "epoch": 2.032540675844806,
+      "grad_norm": 4.097041130065918,
+      "learning_rate": 9.835612822090483e-05,
+      "loss": 1.2015,
+      "step": 405
+    },
+    {
+      "epoch": 2.037546933667084,
+      "grad_norm": 4.594753742218018,
+      "learning_rate": 9.782751333310905e-05,
+      "loss": 1.082,
+      "step": 406
+    },
+    {
+      "epoch": 2.0425531914893615,
+      "grad_norm": 3.9699299335479736,
+      "learning_rate": 9.72989854206378e-05,
+      "loss": 1.258,
+      "step": 407
+    },
+    {
+      "epoch": 2.0475594493116396,
+      "grad_norm": 4.059878349304199,
+      "learning_rate": 9.677055897335087e-05,
+      "loss": 1.0336,
+      "step": 408
+    },
+    {
+      "epoch": 2.052565707133917,
+      "grad_norm": 3.936467409133911,
+      "learning_rate": 9.62422484783261e-05,
+      "loss": 1.3053,
+      "step": 409
+    },
+    {
+      "epoch": 2.057571964956195,
+      "grad_norm": 3.5442802906036377,
+      "learning_rate": 9.571406841946267e-05,
+      "loss": 1.0357,
+      "step": 410
+    },
+    {
+      "epoch": 2.0625782227784732,
+      "grad_norm": 3.461674213409424,
+      "learning_rate": 9.518603327708372e-05,
+      "loss": 1.0689,
+      "step": 411
+    },
+    {
+      "epoch": 2.067584480600751,
+      "grad_norm": 3.423532247543335,
+      "learning_rate": 9.465815752753935e-05,
+      "loss": 1.0562,
+      "step": 412
+    },
+    {
+      "epoch": 2.072590738423029,
+      "grad_norm": 4.412137508392334,
+      "learning_rate": 9.413045564280998e-05,
+      "loss": 1.2204,
+      "step": 413
+    },
+    {
+      "epoch": 2.0775969962453065,
+      "grad_norm": 3.3397271633148193,
+      "learning_rate": 9.360294209010923e-05,
+      "loss": 0.9651,
+      "step": 414
+    },
+    {
+      "epoch": 2.0826032540675845,
+      "grad_norm": 3.77931547164917,
+      "learning_rate": 9.307563133148767e-05,
+      "loss": 1.2274,
+      "step": 415
+    },
+    {
+      "epoch": 2.0876095118898625,
+      "grad_norm": 3.565459728240967,
+      "learning_rate": 9.254853782343616e-05,
+      "loss": 1.1143,
+      "step": 416
+    },
+    {
+      "epoch": 2.09261576971214,
+      "grad_norm": 3.6844406127929688,
+      "learning_rate": 9.202167601648942e-05,
+      "loss": 1.2619,
+      "step": 417
+    },
+    {
+      "epoch": 2.097622027534418,
+      "grad_norm": 3.3748247623443604,
+      "learning_rate": 9.149506035483005e-05,
+      "loss": 1.0988,
+      "step": 418
+    },
+    {
+      "epoch": 2.1026282853566958,
+      "grad_norm": 3.4693350791931152,
+      "learning_rate": 9.096870527589248e-05,
+      "loss": 1.1261,
+      "step": 419
+    },
+    {
+      "epoch": 2.107634543178974,
+      "grad_norm": 3.5337955951690674,
+      "learning_rate": 9.044262520996702e-05,
+      "loss": 1.0759,
+      "step": 420
+    },
+    {
+      "epoch": 2.1126408010012514,
+      "grad_norm": 3.5257728099823,
+      "learning_rate": 8.991683457980443e-05,
+      "loss": 1.085,
+      "step": 421
+    },
+    {
+      "epoch": 2.1176470588235294,
+      "grad_norm": 3.7548561096191406,
+      "learning_rate": 8.93913478002205e-05,
+      "loss": 1.3311,
+      "step": 422
+    },
+    {
+      "epoch": 2.122653316645807,
+      "grad_norm": 3.253643751144409,
+      "learning_rate": 8.886617927770065e-05,
+      "loss": 1.0647,
+      "step": 423
+    },
+    {
+      "epoch": 2.127659574468085,
+      "grad_norm": 3.59781551361084,
+      "learning_rate": 8.834134341000527e-05,
+      "loss": 1.1489,
+      "step": 424
+    },
+    {
+      "epoch": 2.132665832290363,
+      "grad_norm": 3.869216203689575,
+      "learning_rate": 8.781685458577481e-05,
+      "loss": 1.2613,
+      "step": 425
+    },
+    {
+      "epoch": 2.1376720901126407,
+      "grad_norm": 3.8844690322875977,
+      "learning_rate": 8.729272718413527e-05,
+      "loss": 1.1593,
+      "step": 426
+    },
+    {
+      "epoch": 2.1426783479349187,
+      "grad_norm": 3.599303722381592,
+      "learning_rate": 8.676897557430415e-05,
+      "loss": 1.1382,
+      "step": 427
+    },
+    {
+      "epoch": 2.1476846057571963,
+      "grad_norm": 3.438326120376587,
+      "learning_rate": 8.624561411519644e-05,
+      "loss": 1.0946,
+      "step": 428
+    },
+    {
+      "epoch": 2.1526908635794744,
+      "grad_norm": 3.7482898235321045,
+      "learning_rate": 8.572265715503086e-05,
+      "loss": 1.1727,
+      "step": 429
+    },
+    {
+      "epoch": 2.1576971214017524,
+      "grad_norm": 3.5767128467559814,
+      "learning_rate": 8.520011903093666e-05,
+      "loss": 1.2824,
+      "step": 430
+    },
+    {
+      "epoch": 2.16270337922403,
+      "grad_norm": 3.148529529571533,
+      "learning_rate": 8.467801406856054e-05,
+      "loss": 0.9895,
+      "step": 431
+    },
+    {
+      "epoch": 2.167709637046308,
+      "grad_norm": 3.4491708278656006,
+      "learning_rate": 8.415635658167368e-05,
+      "loss": 1.1678,
+      "step": 432
+    },
+    {
+      "epoch": 2.1727158948685856,
+      "grad_norm": 3.461238384246826,
+      "learning_rate": 8.363516087177962e-05,
+      "loss": 1.1203,
+      "step": 433
+    },
+    {
+      "epoch": 2.1777221526908637,
+      "grad_norm": 3.5150508880615234,
+      "learning_rate": 8.31144412277221e-05,
+      "loss": 1.1886,
+      "step": 434
+    },
+    {
+      "epoch": 2.1827284105131413,
+      "grad_norm": 3.0719783306121826,
+      "learning_rate": 8.25942119252931e-05,
+      "loss": 0.9225,
+      "step": 435
+    },
+    {
+      "epoch": 2.1877346683354193,
+      "grad_norm": 3.7611846923828125,
+      "learning_rate": 8.20744872268418e-05,
+      "loss": 1.2618,
+      "step": 436
+    },
+    {
+      "epoch": 2.192740926157697,
+      "grad_norm": 4.004642963409424,
+      "learning_rate": 8.155528138088337e-05,
+      "loss": 1.2625,
+      "step": 437
+    },
+    {
+      "epoch": 2.197747183979975,
+      "grad_norm": 3.3355510234832764,
+      "learning_rate": 8.103660862170826e-05,
+      "loss": 1.0592,
+      "step": 438
+    },
+    {
+      "epoch": 2.202753441802253,
+      "grad_norm": 3.7413265705108643,
+      "learning_rate": 8.051848316899227e-05,
+      "loss": 1.208,
+      "step": 439
+    },
+    {
+      "epoch": 2.2077596996245306,
+      "grad_norm": 4.234354496002197,
+      "learning_rate": 8.000091922740633e-05,
+      "loss": 1.4002,
+      "step": 440
+    },
+    {
+      "epoch": 2.2127659574468086,
+      "grad_norm": 4.029099464416504,
+      "learning_rate": 7.948393098622737e-05,
+      "loss": 1.1973,
+      "step": 441
+    },
+    {
+      "epoch": 2.217772215269086,
+      "grad_norm": 3.4637832641601562,
+      "learning_rate": 7.896753261894923e-05,
+      "loss": 1.1302,
+      "step": 442
+    },
+    {
+      "epoch": 2.2227784730913642,
+      "grad_norm": 3.666888952255249,
+      "learning_rate": 7.845173828289392e-05,
+      "loss": 1.1006,
+      "step": 443
+    },
+    {
+      "epoch": 2.2277847309136423,
+      "grad_norm": 3.490281581878662,
+      "learning_rate": 7.793656211882377e-05,
+      "loss": 1.095,
+      "step": 444
+    },
+    {
+      "epoch": 2.23279098873592,
+      "grad_norm": 3.6505391597747803,
+      "learning_rate": 7.74220182505536e-05,
+      "loss": 1.2131,
+      "step": 445
+    },
+    {
+      "epoch": 2.237797246558198,
+      "grad_norm": 3.622451066970825,
+      "learning_rate": 7.690812078456336e-05,
+      "loss": 1.238,
+      "step": 446
+    },
+    {
+      "epoch": 2.2428035043804755,
+      "grad_norm": 3.1041157245635986,
+      "learning_rate": 7.639488380961173e-05,
+      "loss": 0.9489,
+      "step": 447
+    },
+    {
+      "epoch": 2.2478097622027535,
+      "grad_norm": 3.3886702060699463,
+      "learning_rate": 7.588232139634968e-05,
+      "loss": 1.1466,
+      "step": 448
+    },
+    {
+      "epoch": 2.252816020025031,
+      "grad_norm": 3.712101697921753,
+      "learning_rate": 7.537044759693463e-05,
+      "loss": 1.0877,
+      "step": 449
+    },
+    {
+      "epoch": 2.257822277847309,
+      "grad_norm": 3.551365613937378,
+      "learning_rate": 7.48592764446454e-05,
+      "loss": 1.1355,
+      "step": 450
+    },
+    {
+      "epoch": 2.2628285356695867,
+      "grad_norm": 3.4715287685394287,
+      "learning_rate": 7.434882195349736e-05,
+      "loss": 1.1936,
+      "step": 451
+    },
+    {
+      "epoch": 2.267834793491865,
+      "grad_norm": 4.415275573730469,
+      "learning_rate": 7.383909811785817e-05,
+      "loss": 1.3533,
+      "step": 452
+    },
+    {
+      "epoch": 2.272841051314143,
+      "grad_norm": 3.295712471008301,
+      "learning_rate": 7.333011891206432e-05,
+      "loss": 1.2051,
+      "step": 453
+    },
+    {
+      "epoch": 2.2778473091364204,
+      "grad_norm": 4.3286213874816895,
+      "learning_rate": 7.282189829003785e-05,
+      "loss": 1.246,
+      "step": 454
+    },
+    {
+      "epoch": 2.2828535669586985,
+      "grad_norm": 3.628694772720337,
+      "learning_rate": 7.231445018490381e-05,
+      "loss": 1.2319,
+      "step": 455
+    },
+    {
+      "epoch": 2.287859824780976,
+      "grad_norm": 3.730273723602295,
+      "learning_rate": 7.180778850860835e-05,
+      "loss": 1.0754,
+      "step": 456
+    },
+    {
+      "epoch": 2.292866082603254,
+      "grad_norm": 3.6538302898406982,
+      "learning_rate": 7.130192715153731e-05,
+      "loss": 1.2115,
+      "step": 457
+    },
+    {
+      "epoch": 2.297872340425532,
+      "grad_norm": 3.4317686557769775,
+      "learning_rate": 7.079687998213526e-05,
+      "loss": 1.2313,
+      "step": 458
+    },
+    {
+      "epoch": 2.3028785982478097,
+      "grad_norm": 3.6949033737182617,
+      "learning_rate": 7.029266084652548e-05,
+      "loss": 1.1821,
+      "step": 459
+    },
+    {
+      "epoch": 2.3078848560700878,
+      "grad_norm": 3.3573741912841797,
+      "learning_rate": 6.978928356813031e-05,
+      "loss": 1.1804,
+      "step": 460
+    },
+    {
+      "epoch": 2.3128911138923653,
+      "grad_norm": 3.4236843585968018,
+      "learning_rate": 6.92867619472921e-05,
+      "loss": 1.1519,
+      "step": 461
+    },
+    {
+      "epoch": 2.3178973717146434,
+      "grad_norm": 3.1184580326080322,
+      "learning_rate": 6.878510976089493e-05,
+      "loss": 1.0577,
+      "step": 462
+    },
+    {
+      "epoch": 2.322903629536921,
+      "grad_norm": 3.2034711837768555,
+      "learning_rate": 6.828434076198693e-05,
+      "loss": 1.1435,
+      "step": 463
+    },
+    {
+      "epoch": 2.327909887359199,
+      "grad_norm": 3.5450615882873535,
+      "learning_rate": 6.77844686794031e-05,
+      "loss": 1.2075,
+      "step": 464
+    },
+    {
+      "epoch": 2.3329161451814766,
+      "grad_norm": 3.3356549739837646,
+      "learning_rate": 6.728550721738915e-05,
+      "loss": 1.1745,
+      "step": 465
+    },
+    {
+      "epoch": 2.3379224030037546,
+      "grad_norm": 3.2924506664276123,
+      "learning_rate": 6.678747005522557e-05,
+      "loss": 1.0487,
+      "step": 466
+    },
+    {
+      "epoch": 2.3429286608260327,
+      "grad_norm": 3.386472225189209,
+      "learning_rate": 6.629037084685278e-05,
+      "loss": 1.1496,
+      "step": 467
+    },
+    {
+      "epoch": 2.3479349186483103,
+      "grad_norm": 3.7050280570983887,
+      "learning_rate": 6.579422322049668e-05,
+      "loss": 1.3198,
+      "step": 468
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 3.130352258682251,
+      "learning_rate": 6.529904077829505e-05,
+      "loss": 1.0512,
+      "step": 469
+    },
+    {
+      "epoch": 2.357947434292866,
+      "grad_norm": 3.446882963180542,
+      "learning_rate": 6.480483709592468e-05,
+      "loss": 1.2949,
+      "step": 470
+    },
+    {
+      "epoch": 2.362953692115144,
+      "grad_norm": 3.6783792972564697,
+      "learning_rate": 6.43116257222292e-05,
+      "loss": 1.1859,
+      "step": 471
+    },
+    {
+      "epoch": 2.367959949937422,
+      "grad_norm": 3.3411688804626465,
+      "learning_rate": 6.381942017884753e-05,
+      "loss": 1.1589,
+      "step": 472
+    },
+    {
+      "epoch": 2.3729662077596996,
+      "grad_norm": 3.6124532222747803,
+      "learning_rate": 6.33282339598433e-05,
+      "loss": 1.0375,
+      "step": 473
+    },
+    {
+      "epoch": 2.3779724655819776,
+      "grad_norm": 3.412095785140991,
+      "learning_rate": 6.283808053133484e-05,
+      "loss": 1.0928,
+      "step": 474
+    },
+    {
+      "epoch": 2.382978723404255,
+      "grad_norm": 4.005087852478027,
+      "learning_rate": 6.234897333112594e-05,
+      "loss": 1.3221,
+      "step": 475
+    },
+    {
+      "epoch": 2.3879849812265332,
+      "grad_norm": 2.9916348457336426,
+      "learning_rate": 6.186092576833761e-05,
+      "loss": 1.028,
+      "step": 476
+    },
+    {
+      "epoch": 2.392991239048811,
+      "grad_norm": 3.539137601852417,
+      "learning_rate": 6.137395122304033e-05,
+      "loss": 1.3132,
+      "step": 477
+    },
+    {
+      "epoch": 2.397997496871089,
+      "grad_norm": 3.288222312927246,
+      "learning_rate": 6.088806304588717e-05,
+      "loss": 1.1697,
+      "step": 478
+    },
+    {
+      "epoch": 2.4030037546933665,
+      "grad_norm": 2.9620723724365234,
+      "learning_rate": 6.0403274557748035e-05,
+      "loss": 0.9548,
+      "step": 479
+    },
+    {
+      "epoch": 2.4080100125156445,
+      "grad_norm": 3.1602985858917236,
+      "learning_rate": 5.9919599049344194e-05,
+      "loss": 1.0504,
+      "step": 480
+    },
+    {
+      "epoch": 2.4130162703379225,
+      "grad_norm": 3.3587441444396973,
+      "learning_rate": 5.943704978088402e-05,
+      "loss": 1.072,
+      "step": 481
+    },
+    {
+      "epoch": 2.4180225281602,
+      "grad_norm": 3.729177951812744,
+      "learning_rate": 5.89556399816995e-05,
+      "loss": 1.2125,
+      "step": 482
+    },
+    {
+      "epoch": 2.423028785982478,
+      "grad_norm": 3.408430576324463,
+      "learning_rate": 5.847538284988341e-05,
+      "loss": 1.1585,
+      "step": 483
+    },
+    {
+      "epoch": 2.4280350438047558,
+      "grad_norm": 3.401205062866211,
+      "learning_rate": 5.7996291551927666e-05,
+      "loss": 1.1512,
+      "step": 484
+    },
+    {
+      "epoch": 2.433041301627034,
+      "grad_norm": 3.5880188941955566,
+      "learning_rate": 5.751837922236217e-05,
+      "loss": 1.2643,
+      "step": 485
+    },
+    {
+      "epoch": 2.438047559449312,
+      "grad_norm": 3.426093578338623,
+      "learning_rate": 5.704165896339494e-05,
+      "loss": 1.2169,
+      "step": 486
+    },
+    {
+      "epoch": 2.4430538172715894,
+      "grad_norm": 3.17655611038208,
+      "learning_rate": 5.656614384455257e-05,
+      "loss": 1.0476,
+      "step": 487
+    },
+    {
+      "epoch": 2.4480600750938675,
+      "grad_norm": 3.2203094959259033,
+      "learning_rate": 5.609184690232235e-05,
+      "loss": 1.0628,
+      "step": 488
+    },
+    {
+      "epoch": 2.453066332916145,
+      "grad_norm": 3.196530342102051,
+      "learning_rate": 5.5618781139794465e-05,
+      "loss": 1.0765,
+      "step": 489
+    },
+    {
+      "epoch": 2.458072590738423,
+      "grad_norm": 3.410573959350586,
+      "learning_rate": 5.514695952630578e-05,
+      "loss": 0.9914,
+      "step": 490
+    },
+    {
+      "epoch": 2.4630788485607007,
+      "grad_norm": 3.6145739555358887,
+      "learning_rate": 5.467639499708423e-05,
+      "loss": 1.111,
+      "step": 491
+    },
+    {
+      "epoch": 2.4680851063829787,
+      "grad_norm": 3.244371175765991,
+      "learning_rate": 5.420710045289399e-05,
+      "loss": 1.0385,
+      "step": 492
+    },
+    {
+      "epoch": 2.4730913642052563,
+      "grad_norm": 3.114959239959717,
+      "learning_rate": 5.373908875968211e-05,
+      "loss": 1.1609,
+      "step": 493
+    },
+    {
+      "epoch": 2.4780976220275344,
+      "grad_norm": 3.3504631519317627,
+      "learning_rate": 5.3272372748225556e-05,
+      "loss": 1.2219,
+      "step": 494
+    },
+    {
+      "epoch": 2.4831038798498124,
+      "grad_norm": 3.6841323375701904,
+      "learning_rate": 5.2806965213779544e-05,
+      "loss": 1.2159,
+      "step": 495
+    },
+    {
+      "epoch": 2.48811013767209,
+      "grad_norm": 3.017163038253784,
+      "learning_rate": 5.234287891572674e-05,
+      "loss": 0.9543,
+      "step": 496
+    },
+    {
+      "epoch": 2.493116395494368,
+      "grad_norm": 3.568748950958252,
+      "learning_rate": 5.1880126577227464e-05,
+      "loss": 1.3312,
+      "step": 497
+    },
+    {
+      "epoch": 2.4981226533166456,
+      "grad_norm": 2.6787469387054443,
+      "learning_rate": 5.141872088487078e-05,
+      "loss": 0.9538,
+      "step": 498
+    },
+    {
+      "epoch": 2.5031289111389237,
+      "grad_norm": 3.8604955673217773,
+      "learning_rate": 5.095867448832683e-05,
+      "loss": 1.3486,
+      "step": 499
+    },
+    {
+      "epoch": 2.5081351689612017,
+      "grad_norm": 3.5043580532073975,
+      "learning_rate": 5.050000000000002e-05,
+      "loss": 1.1804,
+      "step": 500
+    },
+    {
+      "epoch": 2.5131414267834793,
+      "grad_norm": 3.233349323272705,
+      "learning_rate": 5.004270999468307e-05,
+      "loss": 1.1243,
+      "step": 501
+    },
+    {
+      "epoch": 2.5181476846057573,
+      "grad_norm": 3.219106912612915,
+      "learning_rate": 4.95868170092125e-05,
+      "loss": 1.3048,
+      "step": 502
+    },
+    {
+      "epoch": 2.523153942428035,
+      "grad_norm": 3.1391568183898926,
+      "learning_rate": 4.913233354212485e-05,
+      "loss": 1.1749,
+      "step": 503
+    },
+    {
+      "epoch": 2.528160200250313,
+      "grad_norm": 2.795653820037842,
+      "learning_rate": 4.867927205331386e-05,
+      "loss": 0.9872,
+      "step": 504
+    },
+    {
+      "epoch": 2.533166458072591,
+      "grad_norm": 3.4815118312835693,
+      "learning_rate": 4.822764496368917e-05,
+      "loss": 1.2715,
+      "step": 505
+    },
+    {
+      "epoch": 2.5381727158948686,
+      "grad_norm": 3.4137794971466064,
+      "learning_rate": 4.7777464654835564e-05,
+      "loss": 1.1253,
+      "step": 506
+    },
+    {
+      "epoch": 2.543178973717146,
+      "grad_norm": 3.228795289993286,
+      "learning_rate": 4.732874346867362e-05,
+      "loss": 1.111,
+      "step": 507
+    },
+    {
+      "epoch": 2.5481852315394242,
+      "grad_norm": 2.9751367568969727,
+      "learning_rate": 4.6881493707121315e-05,
+      "loss": 0.9749,
+      "step": 508
+    },
+    {
+      "epoch": 2.5531914893617023,
+      "grad_norm": 4.011078834533691,
+      "learning_rate": 4.643572763175684e-05,
+      "loss": 1.1068,
+      "step": 509
+    },
+    {
+      "epoch": 2.55819774718398,
+      "grad_norm": 3.416471481323242,
+      "learning_rate": 4.5991457463482264e-05,
+      "loss": 1.1882,
+      "step": 510
+    },
+    {
+      "epoch": 2.563204005006258,
+      "grad_norm": 2.948613166809082,
+      "learning_rate": 4.554869538218868e-05,
+      "loss": 0.9455,
+      "step": 511
+    },
+    {
+      "epoch": 2.5682102628285355,
+      "grad_norm": 3.4453470706939697,
+      "learning_rate": 4.5107453526422255e-05,
+      "loss": 1.3593,
+      "step": 512
+    },
+    {
+      "epoch": 2.5732165206508135,
+      "grad_norm": 3.316791296005249,
+      "learning_rate": 4.46677439930513e-05,
+      "loss": 1.1298,
+      "step": 513
+    },
+    {
+      "epoch": 2.5782227784730916,
+      "grad_norm": 2.7384519577026367,
+      "learning_rate": 4.422957883693483e-05,
+      "loss": 0.9876,
+      "step": 514
+    },
+    {
+      "epoch": 2.583229036295369,
+      "grad_norm": 3.192760705947876,
+      "learning_rate": 4.3792970070591906e-05,
+      "loss": 1.0523,
+      "step": 515
+    },
+    {
+      "epoch": 2.588235294117647,
+      "grad_norm": 3.4807369709014893,
+      "learning_rate": 4.3357929663872406e-05,
+      "loss": 1.0323,
+      "step": 516
+    },
+    {
+      "epoch": 2.593241551939925,
+      "grad_norm": 3.2211036682128906,
+      "learning_rate": 4.29244695436289e-05,
+      "loss": 0.9843,
+      "step": 517
+    },
+    {
+      "epoch": 2.598247809762203,
+      "grad_norm": 3.5839602947235107,
+      "learning_rate": 4.249260159338946e-05,
+      "loss": 1.1994,
+      "step": 518
+    },
+    {
+      "epoch": 2.603254067584481,
+      "grad_norm": 3.289240837097168,
+      "learning_rate": 4.2062337653032146e-05,
+      "loss": 1.084,
+      "step": 519
+    },
+    {
+      "epoch": 2.6082603254067585,
+      "grad_norm": 3.4030802249908447,
+      "learning_rate": 4.1633689518460225e-05,
+      "loss": 1.2044,
+      "step": 520
+    },
+    {
+      "epoch": 2.613266583229036,
+      "grad_norm": 3.1492092609405518,
+      "learning_rate": 4.1206668941278826e-05,
+      "loss": 1.0259,
+      "step": 521
+    },
+    {
+      "epoch": 2.618272841051314,
+      "grad_norm": 3.439842939376831,
+      "learning_rate": 4.078128762847279e-05,
+      "loss": 1.2384,
+      "step": 522
+    },
+    {
+      "epoch": 2.623279098873592,
+      "grad_norm": 3.105792284011841,
+      "learning_rate": 4.035755724208573e-05,
+      "loss": 1.0454,
+      "step": 523
+    },
+    {
+      "epoch": 2.6282853566958697,
+      "grad_norm": 3.2097530364990234,
+      "learning_rate": 3.9935489398900145e-05,
+      "loss": 1.1008,
+      "step": 524
+    },
+    {
+      "epoch": 2.6332916145181477,
+      "grad_norm": 2.9623773097991943,
+      "learning_rate": 3.951509567011922e-05,
+      "loss": 0.9429,
+      "step": 525
+    },
+    {
+      "epoch": 2.6382978723404253,
+      "grad_norm": 3.3015732765197754,
+      "learning_rate": 3.90963875810494e-05,
+      "loss": 1.1089,
+      "step": 526
+    },
+    {
+      "epoch": 2.6433041301627034,
+      "grad_norm": 3.1678755283355713,
+      "learning_rate": 3.86793766107844e-05,
+      "loss": 1.0696,
+      "step": 527
+    },
+    {
+      "epoch": 2.6483103879849814,
+      "grad_norm": 2.9453158378601074,
+      "learning_rate": 3.826407419189066e-05,
+      "loss": 0.9968,
+      "step": 528
+    },
+    {
+      "epoch": 2.653316645807259,
+      "grad_norm": 2.9824790954589844,
+      "learning_rate": 3.785049171009381e-05,
+      "loss": 1.103,
+      "step": 529
+    },
+    {
+      "epoch": 2.658322903629537,
+      "grad_norm": 3.202838659286499,
+      "learning_rate": 3.743864050396644e-05,
+      "loss": 1.1283,
+      "step": 530
+    },
+    {
+      "epoch": 2.6633291614518146,
+      "grad_norm": 3.488142728805542,
+      "learning_rate": 3.7028531864617444e-05,
+      "loss": 1.1765,
+      "step": 531
+    },
+    {
+      "epoch": 2.6683354192740927,
+      "grad_norm": 3.3499858379364014,
+      "learning_rate": 3.662017703538234e-05,
+      "loss": 1.2133,
+      "step": 532
+    },
+    {
+      "epoch": 2.6733416770963707,
+      "grad_norm": 3.035024881362915,
+      "learning_rate": 3.621358721151505e-05,
+      "loss": 1.1001,
+      "step": 533
+    },
+    {
+      "epoch": 2.6783479349186483,
+      "grad_norm": 3.237748861312866,
+      "learning_rate": 3.5808773539880973e-05,
+      "loss": 1.1989,
+      "step": 534
+    },
+    {
+      "epoch": 2.683354192740926,
+      "grad_norm": 3.3954966068267822,
+      "learning_rate": 3.540574711865146e-05,
+      "loss": 1.202,
+      "step": 535
+    },
+    {
+      "epoch": 2.688360450563204,
+      "grad_norm": 3.1598644256591797,
+      "learning_rate": 3.500451899699935e-05,
+      "loss": 1.1158,
+      "step": 536
+    },
+    {
+      "epoch": 2.693366708385482,
+      "grad_norm": 3.440000295639038,
+      "learning_rate": 3.460510017479631e-05,
+      "loss": 1.2114,
+      "step": 537
+    },
+    {
+      "epoch": 2.6983729662077596,
+      "grad_norm": 2.966639518737793,
+      "learning_rate": 3.420750160231118e-05,
+      "loss": 0.9118,
+      "step": 538
+    },
+    {
+      "epoch": 2.7033792240300376,
+      "grad_norm": 3.380932092666626,
+      "learning_rate": 3.381173417990957e-05,
+      "loss": 1.0835,
+      "step": 539
+    },
+    {
+      "epoch": 2.708385481852315,
+      "grad_norm": 3.229613780975342,
+      "learning_rate": 3.3417808757755355e-05,
+      "loss": 1.138,
+      "step": 540
+    },
+    {
+      "epoch": 2.7133917396745932,
+      "grad_norm": 3.15791916847229,
+      "learning_rate": 3.302573613551292e-05,
+      "loss": 1.2285,
+      "step": 541
+    },
+    {
+      "epoch": 2.7183979974968713,
+      "grad_norm": 2.8461785316467285,
+      "learning_rate": 3.263552706205128e-05,
+      "loss": 0.9287,
+      "step": 542
+    },
+    {
+      "epoch": 2.723404255319149,
+      "grad_norm": 2.8767101764678955,
+      "learning_rate": 3.22471922351493e-05,
+      "loss": 1.0412,
+      "step": 543
+    },
+    {
+      "epoch": 2.728410513141427,
+      "grad_norm": 2.858546018600464,
+      "learning_rate": 3.186074230120244e-05,
+      "loss": 1.0807,
+      "step": 544
+    },
+    {
+      "epoch": 2.7334167709637045,
+      "grad_norm": 2.865001678466797,
+      "learning_rate": 3.147618785493083e-05,
+      "loss": 1.0954,
+      "step": 545
+    },
+    {
+      "epoch": 2.7384230287859825,
+      "grad_norm": 3.1909632682800293,
+      "learning_rate": 3.109353943908893e-05,
+      "loss": 1.1101,
+      "step": 546
+    },
+    {
+      "epoch": 2.7434292866082606,
+      "grad_norm": 3.1679904460906982,
+      "learning_rate": 3.071280754417626e-05,
+      "loss": 1.031,
+      "step": 547
+    },
+    {
+      "epoch": 2.748435544430538,
+      "grad_norm": 3.0772807598114014,
+      "learning_rate": 3.033400260815008e-05,
+      "loss": 1.1844,
+      "step": 548
+    },
+    {
+      "epoch": 2.7534418022528158,
+      "grad_norm": 3.3308348655700684,
+      "learning_rate": 2.9957135016139122e-05,
+      "loss": 1.3589,
+      "step": 549
+    },
+    {
+      "epoch": 2.758448060075094,
+      "grad_norm": 3.0506107807159424,
+      "learning_rate": 2.9582215100158706e-05,
+      "loss": 1.1545,
+      "step": 550
+    },
+    {
+      "epoch": 2.763454317897372,
+      "grad_norm": 3.096111536026001,
+      "learning_rate": 2.920925313882776e-05,
+      "loss": 1.1531,
+      "step": 551
+    },
+    {
+      "epoch": 2.7684605757196494,
+      "grad_norm": 2.8558101654052734,
+      "learning_rate": 2.8838259357086884e-05,
+      "loss": 0.975,
+      "step": 552
+    },
+    {
+      "epoch": 2.7734668335419275,
+      "grad_norm": 3.4529948234558105,
+      "learning_rate": 2.846924392591794e-05,
+      "loss": 1.2278,
+      "step": 553
+    },
+    {
+      "epoch": 2.778473091364205,
+      "grad_norm": 2.830240249633789,
+      "learning_rate": 2.8102216962065423e-05,
+      "loss": 0.9457,
+      "step": 554
+    },
+    {
+      "epoch": 2.783479349186483,
+      "grad_norm": 2.8564674854278564,
+      "learning_rate": 2.7737188527758972e-05,
+      "loss": 0.9811,
+      "step": 555
+    },
+    {
+      "epoch": 2.788485607008761,
+      "grad_norm": 3.3456573486328125,
+      "learning_rate": 2.7374168630437456e-05,
+      "loss": 1.1045,
+      "step": 556
+    },
+    {
+      "epoch": 2.7934918648310387,
+      "grad_norm": 2.9954707622528076,
+      "learning_rate": 2.7013167222474756e-05,
+      "loss": 1.1134,
+      "step": 557
+    },
+    {
+      "epoch": 2.7984981226533168,
+      "grad_norm": 2.952683448791504,
+      "learning_rate": 2.6654194200906833e-05,
+      "loss": 1.0336,
+      "step": 558
+    },
+    {
+      "epoch": 2.8035043804755944,
+      "grad_norm": 3.3193068504333496,
+      "learning_rate": 2.629725940716041e-05,
+      "loss": 1.1333,
+      "step": 559
+    },
+    {
+      "epoch": 2.8085106382978724,
+      "grad_norm": 3.06868577003479,
+      "learning_rate": 2.5942372626783172e-05,
+      "loss": 1.0866,
+      "step": 560
+    },
+    {
+      "epoch": 2.8135168961201504,
+      "grad_norm": 2.9460933208465576,
+      "learning_rate": 2.5589543589175485e-05,
+      "loss": 1.0895,
+      "step": 561
+    },
+    {
+      "epoch": 2.818523153942428,
+      "grad_norm": 3.093489170074463,
+      "learning_rate": 2.523878196732358e-05,
+      "loss": 1.1445,
+      "step": 562
+    },
+    {
+      "epoch": 2.8235294117647056,
+      "grad_norm": 3.155773162841797,
+      "learning_rate": 2.489009737753459e-05,
+      "loss": 1.1296,
+      "step": 563
+    },
+    {
+      "epoch": 2.8285356695869837,
+      "grad_norm": 2.886901378631592,
+      "learning_rate": 2.4543499379172615e-05,
+      "loss": 1.0545,
+      "step": 564
+    },
+    {
+      "epoch": 2.8335419274092617,
+      "grad_norm": 3.0896921157836914,
+      "learning_rate": 2.4198997474396877e-05,
+      "loss": 1.0446,
+      "step": 565
+    },
+    {
+      "epoch": 2.8385481852315393,
+      "grad_norm": 2.9552993774414062,
+      "learning_rate": 2.3856601107901166e-05,
+      "loss": 0.8962,
+      "step": 566
+    },
+    {
+      "epoch": 2.8435544430538173,
+      "grad_norm": 2.9749550819396973,
+      "learning_rate": 2.351631966665476e-05,
+      "loss": 1.0768,
+      "step": 567
+    },
+    {
+      "epoch": 2.848560700876095,
+      "grad_norm": 3.1430954933166504,
+      "learning_rate": 2.31781624796453e-05,
+      "loss": 1.1391,
+      "step": 568
+    },
+    {
+      "epoch": 2.853566958698373,
+      "grad_norm": 3.2870116233825684,
+      "learning_rate": 2.2842138817622883e-05,
+      "loss": 0.985,
+      "step": 569
+    },
+    {
+      "epoch": 2.858573216520651,
+      "grad_norm": 2.746265172958374,
+      "learning_rate": 2.250825789284594e-05,
+      "loss": 0.9044,
+      "step": 570
+    },
+    {
+      "epoch": 2.8635794743429286,
+      "grad_norm": 3.157536506652832,
+      "learning_rate": 2.217652885882869e-05,
+      "loss": 1.0997,
+      "step": 571
+    },
+    {
+      "epoch": 2.8685857321652066,
+      "grad_norm": 3.519000291824341,
+      "learning_rate": 2.1846960810090188e-05,
+      "loss": 1.2928,
+      "step": 572
+    },
+    {
+      "epoch": 2.873591989987484,
+      "grad_norm": 3.1560218334198,
+      "learning_rate": 2.151956278190494e-05,
+      "loss": 1.136,
+      "step": 573
+    },
+    {
+      "epoch": 2.8785982478097623,
+      "grad_norm": 3.072340965270996,
+      "learning_rate": 2.119434375005527e-05,
+      "loss": 1.087,
+      "step": 574
+    },
+    {
+      "epoch": 2.8836045056320403,
+      "grad_norm": 2.9986274242401123,
+      "learning_rate": 2.087131263058526e-05,
+      "loss": 0.9521,
+      "step": 575
+    },
+    {
+      "epoch": 2.888610763454318,
+      "grad_norm": 3.377012014389038,
+      "learning_rate": 2.055047827955618e-05,
+      "loss": 1.205,
+      "step": 576
+    },
+    {
+      "epoch": 2.8936170212765955,
+      "grad_norm": 3.326692819595337,
+      "learning_rate": 2.0231849492803852e-05,
+      "loss": 1.1575,
+      "step": 577
+    },
+    {
+      "epoch": 2.8986232790988735,
+      "grad_norm": 2.875267744064331,
+      "learning_rate": 1.991543500569745e-05,
+      "loss": 0.9717,
+      "step": 578
+    },
+    {
+      "epoch": 2.9036295369211516,
+      "grad_norm": 3.402390956878662,
+      "learning_rate": 1.960124349289992e-05,
+      "loss": 1.3638,
+      "step": 579
+    },
+    {
+      "epoch": 2.908635794743429,
+      "grad_norm": 3.127502202987671,
+      "learning_rate": 1.928928356813032e-05,
+      "loss": 1.1322,
+      "step": 580
+    },
+    {
+      "epoch": 2.913642052565707,
+      "grad_norm": 3.1292340755462646,
+      "learning_rate": 1.8979563783927565e-05,
+      "loss": 1.1917,
+      "step": 581
+    },
+    {
+      "epoch": 2.918648310387985,
+      "grad_norm": 2.690685510635376,
+      "learning_rate": 1.8672092631416013e-05,
+      "loss": 0.845,
+      "step": 582
+    },
+    {
+      "epoch": 2.923654568210263,
+      "grad_norm": 3.146888256072998,
+      "learning_rate": 1.8366878540072614e-05,
+      "loss": 1.1461,
+      "step": 583
+    },
+    {
+      "epoch": 2.928660826032541,
+      "grad_norm": 3.0849928855895996,
+      "learning_rate": 1.8063929877495892e-05,
+      "loss": 1.0927,
+      "step": 584
+    },
+    {
+      "epoch": 2.9336670838548184,
+      "grad_norm": 2.757366180419922,
+      "learning_rate": 1.7763254949176414e-05,
+      "loss": 0.903,
+      "step": 585
+    },
+    {
+      "epoch": 2.9386733416770965,
+      "grad_norm": 3.00736665725708,
+      "learning_rate": 1.7464861998269243e-05,
+      "loss": 1.0627,
+      "step": 586
+    },
+    {
+      "epoch": 2.943679599499374,
+      "grad_norm": 2.767845630645752,
+      "learning_rate": 1.7168759205367893e-05,
+      "loss": 1.0366,
+      "step": 587
+    },
+    {
+      "epoch": 2.948685857321652,
+      "grad_norm": 3.163112163543701,
+      "learning_rate": 1.6874954688279956e-05,
+      "loss": 1.1653,
+      "step": 588
+    },
+    {
+      "epoch": 2.95369211514393,
+      "grad_norm": 3.0217740535736084,
+      "learning_rate": 1.6583456501804725e-05,
+      "loss": 1.1163,
+      "step": 589
+    },
+    {
+      "epoch": 2.9586983729662077,
+      "grad_norm": 2.995216131210327,
+      "learning_rate": 1.6294272637512183e-05,
+      "loss": 0.9859,
+      "step": 590
+    },
+    {
+      "epoch": 2.9637046307884853,
+      "grad_norm": 3.1709847450256348,
+      "learning_rate": 1.600741102352409e-05,
+      "loss": 1.0269,
+      "step": 591
+    },
+    {
+      "epoch": 2.9687108886107634,
+      "grad_norm": 3.2043511867523193,
+      "learning_rate": 1.57228795242965e-05,
+      "loss": 1.1052,
+      "step": 592
+    },
+    {
+      "epoch": 2.9737171464330414,
+      "grad_norm": 2.9635226726531982,
+      "learning_rate": 1.544068594040417e-05,
+      "loss": 0.8923,
+      "step": 593
+    },
+    {
+      "epoch": 2.978723404255319,
+      "grad_norm": 3.1605000495910645,
+      "learning_rate": 1.516083800832676e-05,
+      "loss": 1.1042,
+      "step": 594
+    },
+    {
+      "epoch": 2.983729662077597,
+      "grad_norm": 3.1283183097839355,
+      "learning_rate": 1.488334340023669e-05,
+      "loss": 1.1628,
+      "step": 595
+    },
+    {
+      "epoch": 2.9887359198998746,
+      "grad_norm": 3.149639129638672,
+      "learning_rate": 1.4608209723788835e-05,
+      "loss": 1.2558,
+      "step": 596
+    },
+    {
+      "epoch": 2.9937421777221527,
+      "grad_norm": 2.6690926551818848,
+      "learning_rate": 1.4335444521911899e-05,
+      "loss": 0.8779,
+      "step": 597
     }
   ],
   "logging_steps": 1,
@@ -2812,7 +4205,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.2530562039493427e+18,
+  "total_flos": 1.8782958184195686e+18,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null