| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.998716302952504, | |
| "eval_steps": 500, | |
| "global_step": 3505, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.025673940949935817, | |
| "grad_norm": 5121380.5, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 13.9863, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.051347881899871634, | |
| "grad_norm": 5153433.5, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 13.5554, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.07702182284980745, | |
| "grad_norm": 5378263.0, | |
| "learning_rate": 6e-06, | |
| "loss": 12.3429, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10269576379974327, | |
| "grad_norm": 5853167.0, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 10.9424, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12836970474967907, | |
| "grad_norm": 5810211.5, | |
| "learning_rate": 1e-05, | |
| "loss": 8.8056, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1540436456996149, | |
| "grad_norm": 5421540.5, | |
| "learning_rate": 1.2e-05, | |
| "loss": 6.446, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1797175866495507, | |
| "grad_norm": 3989913.0, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 4.0088, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.20539152759948653, | |
| "grad_norm": 1593379.125, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 2.4106, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23106546854942234, | |
| "grad_norm": 351223.375, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.5032, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.25673940949935814, | |
| "grad_norm": 255315.6875, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0421, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.28241335044929394, | |
| "grad_norm": 150451.453125, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 0.6102, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3080872913992298, | |
| "grad_norm": 44021.93359375, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.3687, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3337612323491656, | |
| "grad_norm": 23825.30078125, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.2745, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3594351732991014, | |
| "grad_norm": 19372.984375, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 0.233, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3851091142490372, | |
| "grad_norm": 19907.76171875, | |
| "learning_rate": 3e-05, | |
| "loss": 0.2075, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.41078305519897307, | |
| "grad_norm": 23577.75390625, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.197, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.43645699614890887, | |
| "grad_norm": 15442.431640625, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 0.1759, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4621309370988447, | |
| "grad_norm": 24589.8671875, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.1626, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4878048780487805, | |
| "grad_norm": 11427.1728515625, | |
| "learning_rate": 3.8e-05, | |
| "loss": 0.1541, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5134788189987163, | |
| "grad_norm": 10105.453125, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1542, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5391527599486521, | |
| "grad_norm": 9209.875, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.1497, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5648267008985879, | |
| "grad_norm": 8737.0283203125, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.1397, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5905006418485238, | |
| "grad_norm": 28856.4609375, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.1466, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6161745827984596, | |
| "grad_norm": 9787.6689453125, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.1358, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.6418485237483954, | |
| "grad_norm": 6893.62353515625, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1265, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6675224646983312, | |
| "grad_norm": 14635.2509765625, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 0.1376, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.693196405648267, | |
| "grad_norm": 7151.79150390625, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 0.1375, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.7188703465982028, | |
| "grad_norm": 6492.265625, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 0.1271, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.7445442875481386, | |
| "grad_norm": 6572.00244140625, | |
| "learning_rate": 5.8e-05, | |
| "loss": 0.1334, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7702182284980744, | |
| "grad_norm": 9629.9423828125, | |
| "learning_rate": 6e-05, | |
| "loss": 0.1384, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7958921694480102, | |
| "grad_norm": 6990.53466796875, | |
| "learning_rate": 6.2e-05, | |
| "loss": 0.1349, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.8215661103979461, | |
| "grad_norm": 8195.203125, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.1231, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.8472400513478819, | |
| "grad_norm": 5189.6279296875, | |
| "learning_rate": 6.6e-05, | |
| "loss": 0.1274, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8729139922978177, | |
| "grad_norm": 8494.017578125, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 0.1324, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8985879332477535, | |
| "grad_norm": 6397.20166015625, | |
| "learning_rate": 7e-05, | |
| "loss": 0.1223, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.9242618741976893, | |
| "grad_norm": 5670.24072265625, | |
| "learning_rate": 7.2e-05, | |
| "loss": 0.129, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.9499358151476252, | |
| "grad_norm": 6880.46533203125, | |
| "learning_rate": 7.4e-05, | |
| "loss": 0.1261, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.975609756097561, | |
| "grad_norm": 5716.75537109375, | |
| "learning_rate": 7.6e-05, | |
| "loss": 0.1263, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0012836970474968, | |
| "grad_norm": 7472.732421875, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 0.1229, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.0269576379974326, | |
| "grad_norm": 6848.66162109375, | |
| "learning_rate": 8e-05, | |
| "loss": 0.1236, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 6818.19921875, | |
| "learning_rate": 8.2e-05, | |
| "loss": 0.1198, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0783055198973042, | |
| "grad_norm": 6299.18115234375, | |
| "learning_rate": 8.4e-05, | |
| "loss": 0.127, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.10397946084724, | |
| "grad_norm": 5963.009765625, | |
| "learning_rate": 8.6e-05, | |
| "loss": 0.1228, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.1296534017971758, | |
| "grad_norm": 5657.501953125, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 0.1248, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.1553273427471118, | |
| "grad_norm": 15452.26953125, | |
| "learning_rate": 9e-05, | |
| "loss": 0.1237, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.1810012836970474, | |
| "grad_norm": 6293.9599609375, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 0.1218, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.2066752246469834, | |
| "grad_norm": 6370.39794921875, | |
| "learning_rate": 9.4e-05, | |
| "loss": 0.1191, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.2323491655969192, | |
| "grad_norm": 6101.39697265625, | |
| "learning_rate": 9.6e-05, | |
| "loss": 0.1161, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.258023106546855, | |
| "grad_norm": 6093.3916015625, | |
| "learning_rate": 9.8e-05, | |
| "loss": 0.1247, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.2836970474967908, | |
| "grad_norm": 4891.51171875, | |
| "learning_rate": 0.0001, | |
| "loss": 0.117, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.2836970474967908, | |
| "eval_gen_len": 38.948, | |
| "eval_loss": 0.10336166620254517, | |
| "eval_rouge1": 0.5729, | |
| "eval_rouge2": 0.2787, | |
| "eval_rougeL": 0.5355, | |
| "eval_runtime": 65.3435, | |
| "eval_samples_per_second": 7.652, | |
| "eval_steps_per_second": 0.49, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.3093709884467266, | |
| "grad_norm": 4472.9599609375, | |
| "learning_rate": 9.970501474926254e-05, | |
| "loss": 0.1215, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.3350449293966624, | |
| "grad_norm": 5620.90576171875, | |
| "learning_rate": 9.941002949852508e-05, | |
| "loss": 0.1212, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.3607188703465982, | |
| "grad_norm": 8320.6953125, | |
| "learning_rate": 9.911504424778762e-05, | |
| "loss": 0.1145, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.386392811296534, | |
| "grad_norm": 6414.8359375, | |
| "learning_rate": 9.882005899705014e-05, | |
| "loss": 0.1248, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.4120667522464698, | |
| "grad_norm": 6748.93017578125, | |
| "learning_rate": 9.85250737463127e-05, | |
| "loss": 0.1167, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.4377406931964056, | |
| "grad_norm": 6198.98779296875, | |
| "learning_rate": 9.823008849557522e-05, | |
| "loss": 0.126, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.4634146341463414, | |
| "grad_norm": 6716.78515625, | |
| "learning_rate": 9.793510324483777e-05, | |
| "loss": 0.1173, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.4890885750962772, | |
| "grad_norm": 5211.44921875, | |
| "learning_rate": 9.76401179941003e-05, | |
| "loss": 0.1117, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.514762516046213, | |
| "grad_norm": 5593.43798828125, | |
| "learning_rate": 9.734513274336283e-05, | |
| "loss": 0.1113, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.540436456996149, | |
| "grad_norm": 6742.3544921875, | |
| "learning_rate": 9.705014749262537e-05, | |
| "loss": 0.1184, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.5661103979460846, | |
| "grad_norm": 5100.9228515625, | |
| "learning_rate": 9.675516224188791e-05, | |
| "loss": 0.1122, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.5917843388960207, | |
| "grad_norm": 6717.302734375, | |
| "learning_rate": 9.646017699115044e-05, | |
| "loss": 0.1211, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.6174582798459562, | |
| "grad_norm": 8748.5849609375, | |
| "learning_rate": 9.616519174041299e-05, | |
| "loss": 0.1198, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.6431322207958923, | |
| "grad_norm": 6425.04248046875, | |
| "learning_rate": 9.587020648967551e-05, | |
| "loss": 0.1146, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.6688061617458279, | |
| "grad_norm": 5612.4580078125, | |
| "learning_rate": 9.557522123893806e-05, | |
| "loss": 0.1113, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.6944801026957639, | |
| "grad_norm": 6821.66259765625, | |
| "learning_rate": 9.528023598820059e-05, | |
| "loss": 0.1144, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.7201540436456995, | |
| "grad_norm": 7305.1376953125, | |
| "learning_rate": 9.498525073746313e-05, | |
| "loss": 0.1139, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.7458279845956355, | |
| "grad_norm": 6363.8701171875, | |
| "learning_rate": 9.469026548672566e-05, | |
| "loss": 0.1121, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.7715019255455713, | |
| "grad_norm": 5113.75537109375, | |
| "learning_rate": 9.43952802359882e-05, | |
| "loss": 0.115, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.797175866495507, | |
| "grad_norm": 6163.470703125, | |
| "learning_rate": 9.410029498525074e-05, | |
| "loss": 0.1244, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.822849807445443, | |
| "grad_norm": 6645.03466796875, | |
| "learning_rate": 9.380530973451328e-05, | |
| "loss": 0.1168, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.8485237483953787, | |
| "grad_norm": 9238.5634765625, | |
| "learning_rate": 9.351032448377582e-05, | |
| "loss": 0.1198, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.8741976893453145, | |
| "grad_norm": 5459.36962890625, | |
| "learning_rate": 9.321533923303836e-05, | |
| "loss": 0.1175, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.8998716302952503, | |
| "grad_norm": 5480.2998046875, | |
| "learning_rate": 9.29203539823009e-05, | |
| "loss": 0.1115, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.925545571245186, | |
| "grad_norm": 5369.232421875, | |
| "learning_rate": 9.262536873156342e-05, | |
| "loss": 0.1126, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.951219512195122, | |
| "grad_norm": 5776.08154296875, | |
| "learning_rate": 9.233038348082597e-05, | |
| "loss": 0.1184, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.976893453145058, | |
| "grad_norm": 4656.33203125, | |
| "learning_rate": 9.20353982300885e-05, | |
| "loss": 0.1145, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.0025673940949935, | |
| "grad_norm": 6602.3056640625, | |
| "learning_rate": 9.174041297935103e-05, | |
| "loss": 0.126, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.0282413350449295, | |
| "grad_norm": 5742.654296875, | |
| "learning_rate": 9.144542772861357e-05, | |
| "loss": 0.1036, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.053915275994865, | |
| "grad_norm": 5546.1728515625, | |
| "learning_rate": 9.115044247787611e-05, | |
| "loss": 0.1077, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.079589216944801, | |
| "grad_norm": 6285.85693359375, | |
| "learning_rate": 9.085545722713865e-05, | |
| "loss": 0.1156, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 5655.2763671875, | |
| "learning_rate": 9.056047197640119e-05, | |
| "loss": 0.1136, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.1309370988446728, | |
| "grad_norm": 5984.7080078125, | |
| "learning_rate": 9.026548672566371e-05, | |
| "loss": 0.1076, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.1566110397946083, | |
| "grad_norm": 5037.02490234375, | |
| "learning_rate": 8.997050147492626e-05, | |
| "loss": 0.1181, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.1822849807445444, | |
| "grad_norm": 5577.619140625, | |
| "learning_rate": 8.96755162241888e-05, | |
| "loss": 0.1147, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.20795892169448, | |
| "grad_norm": 4631.0732421875, | |
| "learning_rate": 8.938053097345133e-05, | |
| "loss": 0.1106, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.233632862644416, | |
| "grad_norm": 4876.57177734375, | |
| "learning_rate": 8.908554572271388e-05, | |
| "loss": 0.1099, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.2593068035943515, | |
| "grad_norm": 4877.7177734375, | |
| "learning_rate": 8.87905604719764e-05, | |
| "loss": 0.1134, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.2849807445442876, | |
| "grad_norm": 5603.54150390625, | |
| "learning_rate": 8.849557522123895e-05, | |
| "loss": 0.1091, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.3106546854942236, | |
| "grad_norm": 4900.8056640625, | |
| "learning_rate": 8.820058997050148e-05, | |
| "loss": 0.111, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.336328626444159, | |
| "grad_norm": 5891.17919921875, | |
| "learning_rate": 8.790560471976402e-05, | |
| "loss": 0.113, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.3620025673940948, | |
| "grad_norm": 6097.0400390625, | |
| "learning_rate": 8.761061946902655e-05, | |
| "loss": 0.1131, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.387676508344031, | |
| "grad_norm": 5553.201171875, | |
| "learning_rate": 8.731563421828909e-05, | |
| "loss": 0.105, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.413350449293967, | |
| "grad_norm": 4939.27392578125, | |
| "learning_rate": 8.702064896755162e-05, | |
| "loss": 0.115, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.4390243902439024, | |
| "grad_norm": 5012.8740234375, | |
| "learning_rate": 8.672566371681417e-05, | |
| "loss": 0.1095, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.4646983311938384, | |
| "grad_norm": 6639.900390625, | |
| "learning_rate": 8.64306784660767e-05, | |
| "loss": 0.1096, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.490372272143774, | |
| "grad_norm": 6572.72705078125, | |
| "learning_rate": 8.613569321533924e-05, | |
| "loss": 0.1091, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.51604621309371, | |
| "grad_norm": 4813.23291015625, | |
| "learning_rate": 8.584070796460177e-05, | |
| "loss": 0.1114, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.5417201540436456, | |
| "grad_norm": 6162.396484375, | |
| "learning_rate": 8.554572271386431e-05, | |
| "loss": 0.1101, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.5673940949935816, | |
| "grad_norm": 6070.53857421875, | |
| "learning_rate": 8.525073746312685e-05, | |
| "loss": 0.1103, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.5673940949935816, | |
| "eval_gen_len": 38.946, | |
| "eval_loss": 0.0969744473695755, | |
| "eval_rouge1": 0.5903, | |
| "eval_rouge2": 0.3078, | |
| "eval_rougeL": 0.556, | |
| "eval_runtime": 65.5669, | |
| "eval_samples_per_second": 7.626, | |
| "eval_steps_per_second": 0.488, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.593068035943517, | |
| "grad_norm": 6648.20751953125, | |
| "learning_rate": 8.495575221238938e-05, | |
| "loss": 0.109, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.6187419768934532, | |
| "grad_norm": 4565.7236328125, | |
| "learning_rate": 8.466076696165192e-05, | |
| "loss": 0.1062, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.644415917843389, | |
| "grad_norm": 5601.0908203125, | |
| "learning_rate": 8.436578171091446e-05, | |
| "loss": 0.1099, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.670089858793325, | |
| "grad_norm": 5187.435546875, | |
| "learning_rate": 8.4070796460177e-05, | |
| "loss": 0.1127, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.6957637997432604, | |
| "grad_norm": 5046.6767578125, | |
| "learning_rate": 8.377581120943954e-05, | |
| "loss": 0.122, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.7214377406931964, | |
| "grad_norm": 4885.3818359375, | |
| "learning_rate": 8.348082595870208e-05, | |
| "loss": 0.105, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.7471116816431325, | |
| "grad_norm": 5473.66015625, | |
| "learning_rate": 8.31858407079646e-05, | |
| "loss": 0.1081, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.772785622593068, | |
| "grad_norm": 6577.04931640625, | |
| "learning_rate": 8.289085545722715e-05, | |
| "loss": 0.1161, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.7984595635430036, | |
| "grad_norm": 5628.3349609375, | |
| "learning_rate": 8.259587020648968e-05, | |
| "loss": 0.1057, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.8241335044929397, | |
| "grad_norm": 5690.630859375, | |
| "learning_rate": 8.230088495575221e-05, | |
| "loss": 0.1116, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.8498074454428757, | |
| "grad_norm": 4804.5458984375, | |
| "learning_rate": 8.200589970501475e-05, | |
| "loss": 0.103, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.8754813863928113, | |
| "grad_norm": 4892.4580078125, | |
| "learning_rate": 8.171091445427729e-05, | |
| "loss": 0.1116, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.901155327342747, | |
| "grad_norm": 5333.4130859375, | |
| "learning_rate": 8.141592920353983e-05, | |
| "loss": 0.106, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.926829268292683, | |
| "grad_norm": 5062.69580078125, | |
| "learning_rate": 8.112094395280237e-05, | |
| "loss": 0.1086, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.952503209242619, | |
| "grad_norm": 4429.697265625, | |
| "learning_rate": 8.082595870206489e-05, | |
| "loss": 0.1079, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.9781771501925545, | |
| "grad_norm": 5827.998046875, | |
| "learning_rate": 8.053097345132744e-05, | |
| "loss": 0.1151, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 3.0038510911424905, | |
| "grad_norm": 5520.6826171875, | |
| "learning_rate": 8.023598820058997e-05, | |
| "loss": 0.1063, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 3.029525032092426, | |
| "grad_norm": 5321.1328125, | |
| "learning_rate": 7.99410029498525e-05, | |
| "loss": 0.1028, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 3.055198973042362, | |
| "grad_norm": 6147.1103515625, | |
| "learning_rate": 7.964601769911504e-05, | |
| "loss": 0.1027, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 3.0808729139922977, | |
| "grad_norm": 5837.15673828125, | |
| "learning_rate": 7.935103244837758e-05, | |
| "loss": 0.1129, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.1065468549422337, | |
| "grad_norm": 5890.27490234375, | |
| "learning_rate": 7.905604719764012e-05, | |
| "loss": 0.105, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 3.1322207958921693, | |
| "grad_norm": 5193.16259765625, | |
| "learning_rate": 7.876106194690266e-05, | |
| "loss": 0.1025, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 5144.00048828125, | |
| "learning_rate": 7.84660766961652e-05, | |
| "loss": 0.1053, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.183568677792041, | |
| "grad_norm": 4484.66552734375, | |
| "learning_rate": 7.817109144542774e-05, | |
| "loss": 0.1111, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 3.209242618741977, | |
| "grad_norm": 5307.7431640625, | |
| "learning_rate": 7.787610619469027e-05, | |
| "loss": 0.1088, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.2349165596919125, | |
| "grad_norm": 5560.54736328125, | |
| "learning_rate": 7.75811209439528e-05, | |
| "loss": 0.1056, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.2605905006418485, | |
| "grad_norm": 5743.2255859375, | |
| "learning_rate": 7.728613569321535e-05, | |
| "loss": 0.1057, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.2862644415917845, | |
| "grad_norm": 5270.234375, | |
| "learning_rate": 7.699115044247787e-05, | |
| "loss": 0.1072, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.31193838254172, | |
| "grad_norm": 5415.4580078125, | |
| "learning_rate": 7.669616519174043e-05, | |
| "loss": 0.1044, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.337612323491656, | |
| "grad_norm": 5077.26123046875, | |
| "learning_rate": 7.640117994100295e-05, | |
| "loss": 0.1063, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.3632862644415917, | |
| "grad_norm": 5815.587890625, | |
| "learning_rate": 7.610619469026549e-05, | |
| "loss": 0.104, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.3889602053915278, | |
| "grad_norm": 5168.4140625, | |
| "learning_rate": 7.581120943952803e-05, | |
| "loss": 0.1057, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.4146341463414633, | |
| "grad_norm": 6175.1962890625, | |
| "learning_rate": 7.551622418879057e-05, | |
| "loss": 0.1085, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.4403080872913994, | |
| "grad_norm": 4866.1806640625, | |
| "learning_rate": 7.522123893805309e-05, | |
| "loss": 0.1088, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.465982028241335, | |
| "grad_norm": 4987.28662109375, | |
| "learning_rate": 7.492625368731564e-05, | |
| "loss": 0.1144, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.491655969191271, | |
| "grad_norm": 6046.6435546875, | |
| "learning_rate": 7.463126843657817e-05, | |
| "loss": 0.1066, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.5173299101412066, | |
| "grad_norm": 6097.72802734375, | |
| "learning_rate": 7.433628318584072e-05, | |
| "loss": 0.1109, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.5430038510911426, | |
| "grad_norm": 6237.98388671875, | |
| "learning_rate": 7.404129793510324e-05, | |
| "loss": 0.1061, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.568677792041078, | |
| "grad_norm": 5106.3720703125, | |
| "learning_rate": 7.374631268436578e-05, | |
| "loss": 0.1052, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.594351732991014, | |
| "grad_norm": 5083.40234375, | |
| "learning_rate": 7.345132743362832e-05, | |
| "loss": 0.1012, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.62002567394095, | |
| "grad_norm": 5191.7353515625, | |
| "learning_rate": 7.315634218289086e-05, | |
| "loss": 0.1081, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.645699614890886, | |
| "grad_norm": 6808.04638671875, | |
| "learning_rate": 7.28613569321534e-05, | |
| "loss": 0.1046, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.6713735558408214, | |
| "grad_norm": 5069.12353515625, | |
| "learning_rate": 7.256637168141593e-05, | |
| "loss": 0.1018, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.6970474967907574, | |
| "grad_norm": 5624.8330078125, | |
| "learning_rate": 7.227138643067847e-05, | |
| "loss": 0.1087, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.7227214377406934, | |
| "grad_norm": 8308.7177734375, | |
| "learning_rate": 7.197640117994101e-05, | |
| "loss": 0.1045, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.748395378690629, | |
| "grad_norm": 5380.990234375, | |
| "learning_rate": 7.168141592920355e-05, | |
| "loss": 0.1083, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.7740693196405646, | |
| "grad_norm": 4756.90576171875, | |
| "learning_rate": 7.138643067846607e-05, | |
| "loss": 0.1058, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.7997432605905006, | |
| "grad_norm": 5780.90625, | |
| "learning_rate": 7.109144542772862e-05, | |
| "loss": 0.1029, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.8254172015404366, | |
| "grad_norm": 5286.49560546875, | |
| "learning_rate": 7.079646017699115e-05, | |
| "loss": 0.1029, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.851091142490372, | |
| "grad_norm": 5398.28369140625, | |
| "learning_rate": 7.050147492625369e-05, | |
| "loss": 0.1086, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.851091142490372, | |
| "eval_gen_len": 38.946, | |
| "eval_loss": 0.09477131813764572, | |
| "eval_rouge1": 0.5988, | |
| "eval_rouge2": 0.3178, | |
| "eval_rougeL": 0.5661, | |
| "eval_runtime": 65.5594, | |
| "eval_samples_per_second": 7.627, | |
| "eval_steps_per_second": 0.488, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.8767650834403082, | |
| "grad_norm": 4950.18212890625, | |
| "learning_rate": 7.020648967551623e-05, | |
| "loss": 0.1025, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.902439024390244, | |
| "grad_norm": 4885.29248046875, | |
| "learning_rate": 6.991150442477876e-05, | |
| "loss": 0.1067, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.92811296534018, | |
| "grad_norm": 6418.5791015625, | |
| "learning_rate": 6.96165191740413e-05, | |
| "loss": 0.1009, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.9537869062901154, | |
| "grad_norm": 6914.34375, | |
| "learning_rate": 6.932153392330384e-05, | |
| "loss": 0.1085, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.9794608472400514, | |
| "grad_norm": 5611.89306640625, | |
| "learning_rate": 6.902654867256638e-05, | |
| "loss": 0.1125, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 4.005134788189987, | |
| "grad_norm": 4575.3046875, | |
| "learning_rate": 6.873156342182892e-05, | |
| "loss": 0.1037, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 4.030808729139923, | |
| "grad_norm": 5809.431640625, | |
| "learning_rate": 6.843657817109145e-05, | |
| "loss": 0.0999, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 4.056482670089859, | |
| "grad_norm": 6907.01025390625, | |
| "learning_rate": 6.814159292035398e-05, | |
| "loss": 0.1007, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 4.082156611039794, | |
| "grad_norm": 6448.38330078125, | |
| "learning_rate": 6.784660766961653e-05, | |
| "loss": 0.11, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 4.10783055198973, | |
| "grad_norm": 15915.1982421875, | |
| "learning_rate": 6.755162241887906e-05, | |
| "loss": 0.1106, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.133504492939666, | |
| "grad_norm": 5690.85107421875, | |
| "learning_rate": 6.725663716814161e-05, | |
| "loss": 0.1037, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 4.159178433889602, | |
| "grad_norm": 4913.6220703125, | |
| "learning_rate": 6.696165191740413e-05, | |
| "loss": 0.1052, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 4.184852374839538, | |
| "grad_norm": 5320.2470703125, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.1018, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 6042.61376953125, | |
| "learning_rate": 6.637168141592921e-05, | |
| "loss": 0.1013, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.2362002567394095, | |
| "grad_norm": 5034.08203125, | |
| "learning_rate": 6.607669616519175e-05, | |
| "loss": 0.1085, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 4.2618741976893455, | |
| "grad_norm": 6053.23876953125, | |
| "learning_rate": 6.578171091445427e-05, | |
| "loss": 0.0986, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.2875481386392815, | |
| "grad_norm": 5543.45556640625, | |
| "learning_rate": 6.548672566371682e-05, | |
| "loss": 0.0986, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.313222079589217, | |
| "grad_norm": 6083.2236328125, | |
| "learning_rate": 6.519174041297935e-05, | |
| "loss": 0.1028, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.338896020539153, | |
| "grad_norm": 5847.65087890625, | |
| "learning_rate": 6.48967551622419e-05, | |
| "loss": 0.1001, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.364569961489089, | |
| "grad_norm": 5046.88623046875, | |
| "learning_rate": 6.460176991150442e-05, | |
| "loss": 0.102, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.390243902439025, | |
| "grad_norm": 7761.01611328125, | |
| "learning_rate": 6.430678466076696e-05, | |
| "loss": 0.1017, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.41591784338896, | |
| "grad_norm": 5590.93505859375, | |
| "learning_rate": 6.40117994100295e-05, | |
| "loss": 0.1058, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.441591784338896, | |
| "grad_norm": 4478.46484375, | |
| "learning_rate": 6.371681415929204e-05, | |
| "loss": 0.0995, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.467265725288832, | |
| "grad_norm": 6958.63720703125, | |
| "learning_rate": 6.342182890855458e-05, | |
| "loss": 0.1028, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.492939666238768, | |
| "grad_norm": 5210.4853515625, | |
| "learning_rate": 6.312684365781711e-05, | |
| "loss": 0.1078, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.518613607188703, | |
| "grad_norm": 4667.54345703125, | |
| "learning_rate": 6.283185840707965e-05, | |
| "loss": 0.1025, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.544287548138639, | |
| "grad_norm": 5578.6943359375, | |
| "learning_rate": 6.253687315634219e-05, | |
| "loss": 0.1029, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.569961489088575, | |
| "grad_norm": 6289.7841796875, | |
| "learning_rate": 6.224188790560473e-05, | |
| "loss": 0.1062, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.595635430038511, | |
| "grad_norm": 5193.00244140625, | |
| "learning_rate": 6.194690265486725e-05, | |
| "loss": 0.1104, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.621309370988447, | |
| "grad_norm": 5092.68408203125, | |
| "learning_rate": 6.16519174041298e-05, | |
| "loss": 0.0996, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.646983311938382, | |
| "grad_norm": 5535.3857421875, | |
| "learning_rate": 6.135693215339233e-05, | |
| "loss": 0.1066, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.672657252888318, | |
| "grad_norm": 6088.28515625, | |
| "learning_rate": 6.106194690265487e-05, | |
| "loss": 0.1031, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.698331193838254, | |
| "grad_norm": 5986.71240234375, | |
| "learning_rate": 6.0766961651917406e-05, | |
| "loss": 0.1043, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.7240051347881895, | |
| "grad_norm": 5196.69140625, | |
| "learning_rate": 6.0471976401179945e-05, | |
| "loss": 0.1035, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.7496790757381255, | |
| "grad_norm": 5394.7138671875, | |
| "learning_rate": 6.017699115044248e-05, | |
| "loss": 0.1017, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.775353016688062, | |
| "grad_norm": 5689.53173828125, | |
| "learning_rate": 5.988200589970502e-05, | |
| "loss": 0.107, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.801026957637998, | |
| "grad_norm": 5098.4541015625, | |
| "learning_rate": 5.958702064896755e-05, | |
| "loss": 0.1032, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.826700898587934, | |
| "grad_norm": 4243.0087890625, | |
| "learning_rate": 5.92920353982301e-05, | |
| "loss": 0.1017, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.852374839537869, | |
| "grad_norm": 5340.123046875, | |
| "learning_rate": 5.899705014749263e-05, | |
| "loss": 0.0986, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.878048780487805, | |
| "grad_norm": 5436.1259765625, | |
| "learning_rate": 5.870206489675516e-05, | |
| "loss": 0.1, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.903722721437741, | |
| "grad_norm": 5866.4375, | |
| "learning_rate": 5.8407079646017705e-05, | |
| "loss": 0.104, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.929396662387677, | |
| "grad_norm": 5687.0595703125, | |
| "learning_rate": 5.8112094395280236e-05, | |
| "loss": 0.1003, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.955070603337612, | |
| "grad_norm": 5049.65869140625, | |
| "learning_rate": 5.781710914454278e-05, | |
| "loss": 0.1051, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.980744544287548, | |
| "grad_norm": 4348.83251953125, | |
| "learning_rate": 5.752212389380531e-05, | |
| "loss": 0.1067, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 5.006418485237484, | |
| "grad_norm": 5278.10498046875, | |
| "learning_rate": 5.7227138643067844e-05, | |
| "loss": 0.1028, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 5.03209242618742, | |
| "grad_norm": 5227.31689453125, | |
| "learning_rate": 5.693215339233039e-05, | |
| "loss": 0.1015, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 5.057766367137355, | |
| "grad_norm": 5626.7041015625, | |
| "learning_rate": 5.663716814159292e-05, | |
| "loss": 0.0953, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 5.083440308087291, | |
| "grad_norm": 4941.787109375, | |
| "learning_rate": 5.634218289085545e-05, | |
| "loss": 0.1, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 5.109114249037227, | |
| "grad_norm": 5543.74365234375, | |
| "learning_rate": 5.6047197640118e-05, | |
| "loss": 0.0975, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 5.134788189987163, | |
| "grad_norm": 6526.22509765625, | |
| "learning_rate": 5.575221238938053e-05, | |
| "loss": 0.1046, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.134788189987163, | |
| "eval_gen_len": 38.946, | |
| "eval_loss": 0.09348437190055847, | |
| "eval_rouge1": 0.6042, | |
| "eval_rouge2": 0.3248, | |
| "eval_rougeL": 0.5712, | |
| "eval_runtime": 65.0847, | |
| "eval_samples_per_second": 7.682, | |
| "eval_steps_per_second": 0.492, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.160462130937099, | |
| "grad_norm": 5070.046875, | |
| "learning_rate": 5.545722713864307e-05, | |
| "loss": 0.0981, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 5.186136071887034, | |
| "grad_norm": 5264.22509765625, | |
| "learning_rate": 5.5162241887905605e-05, | |
| "loss": 0.1023, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 5.21181001283697, | |
| "grad_norm": 10262.3994140625, | |
| "learning_rate": 5.486725663716814e-05, | |
| "loss": 0.1007, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 5.2374839537869065, | |
| "grad_norm": 4638.310546875, | |
| "learning_rate": 5.457227138643069e-05, | |
| "loss": 0.1019, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "grad_norm": 5691.34033203125, | |
| "learning_rate": 5.427728613569322e-05, | |
| "loss": 0.1048, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.288831835686778, | |
| "grad_norm": 5892.60986328125, | |
| "learning_rate": 5.398230088495575e-05, | |
| "loss": 0.1002, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 5.314505776636714, | |
| "grad_norm": 5043.25, | |
| "learning_rate": 5.3687315634218295e-05, | |
| "loss": 0.1026, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 5.34017971758665, | |
| "grad_norm": 5076.90283203125, | |
| "learning_rate": 5.339233038348083e-05, | |
| "loss": 0.103, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.365853658536586, | |
| "grad_norm": 5730.2998046875, | |
| "learning_rate": 5.309734513274337e-05, | |
| "loss": 0.0995, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 5.391527599486521, | |
| "grad_norm": 5071.3759765625, | |
| "learning_rate": 5.28023598820059e-05, | |
| "loss": 0.1006, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.417201540436457, | |
| "grad_norm": 4912.38134765625, | |
| "learning_rate": 5.2507374631268435e-05, | |
| "loss": 0.0965, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 5.442875481386393, | |
| "grad_norm": 5349.1376953125, | |
| "learning_rate": 5.221238938053098e-05, | |
| "loss": 0.1016, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 5.468549422336329, | |
| "grad_norm": 6012.4912109375, | |
| "learning_rate": 5.191740412979351e-05, | |
| "loss": 0.0985, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 5.494223363286264, | |
| "grad_norm": 6078.17333984375, | |
| "learning_rate": 5.162241887905604e-05, | |
| "loss": 0.1001, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 5.5198973042362, | |
| "grad_norm": 6352.015625, | |
| "learning_rate": 5.132743362831859e-05, | |
| "loss": 0.0995, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.545571245186136, | |
| "grad_norm": 10780.03125, | |
| "learning_rate": 5.103244837758112e-05, | |
| "loss": 0.1037, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 5.571245186136072, | |
| "grad_norm": 4540.59326171875, | |
| "learning_rate": 5.0737463126843664e-05, | |
| "loss": 0.1003, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 5.596919127086007, | |
| "grad_norm": 5141.4697265625, | |
| "learning_rate": 5.0442477876106195e-05, | |
| "loss": 0.0993, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 5.622593068035943, | |
| "grad_norm": 8023.310546875, | |
| "learning_rate": 5.014749262536873e-05, | |
| "loss": 0.0923, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 5.648267008985879, | |
| "grad_norm": 6443.39404296875, | |
| "learning_rate": 4.985250737463127e-05, | |
| "loss": 0.1059, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.673940949935815, | |
| "grad_norm": 4546.185546875, | |
| "learning_rate": 4.955752212389381e-05, | |
| "loss": 0.1027, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 5.699614890885751, | |
| "grad_norm": 5331.25634765625, | |
| "learning_rate": 4.926253687315635e-05, | |
| "loss": 0.0994, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 5.7252888318356865, | |
| "grad_norm": 5486.52587890625, | |
| "learning_rate": 4.8967551622418886e-05, | |
| "loss": 0.1097, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 5.7509627727856225, | |
| "grad_norm": 5083.9794921875, | |
| "learning_rate": 4.867256637168142e-05, | |
| "loss": 0.0971, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 5.7766367137355585, | |
| "grad_norm": 5799.4931640625, | |
| "learning_rate": 4.8377581120943956e-05, | |
| "loss": 0.103, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.802310654685495, | |
| "grad_norm": 5407.1708984375, | |
| "learning_rate": 4.8082595870206494e-05, | |
| "loss": 0.1091, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 5.82798459563543, | |
| "grad_norm": 4890.9697265625, | |
| "learning_rate": 4.778761061946903e-05, | |
| "loss": 0.1013, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 5.853658536585366, | |
| "grad_norm": 5403.1416015625, | |
| "learning_rate": 4.749262536873156e-05, | |
| "loss": 0.1076, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 5.879332477535302, | |
| "grad_norm": 5159.65234375, | |
| "learning_rate": 4.71976401179941e-05, | |
| "loss": 0.0994, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 5.905006418485238, | |
| "grad_norm": 6055.45458984375, | |
| "learning_rate": 4.690265486725664e-05, | |
| "loss": 0.0998, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.930680359435174, | |
| "grad_norm": 5306.44677734375, | |
| "learning_rate": 4.660766961651918e-05, | |
| "loss": 0.0995, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 5.956354300385109, | |
| "grad_norm": 5193.0009765625, | |
| "learning_rate": 4.631268436578171e-05, | |
| "loss": 0.1011, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 5.982028241335045, | |
| "grad_norm": 6859.47509765625, | |
| "learning_rate": 4.601769911504425e-05, | |
| "loss": 0.1043, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 6.007702182284981, | |
| "grad_norm": 4973.0458984375, | |
| "learning_rate": 4.5722713864306786e-05, | |
| "loss": 0.1002, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 6.033376123234916, | |
| "grad_norm": 5652.50439453125, | |
| "learning_rate": 4.5427728613569324e-05, | |
| "loss": 0.0994, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 6.059050064184852, | |
| "grad_norm": 6935.865234375, | |
| "learning_rate": 4.5132743362831855e-05, | |
| "loss": 0.0998, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 6.084724005134788, | |
| "grad_norm": 4675.81982421875, | |
| "learning_rate": 4.48377581120944e-05, | |
| "loss": 0.1014, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 6.110397946084724, | |
| "grad_norm": 4515.3134765625, | |
| "learning_rate": 4.454277286135694e-05, | |
| "loss": 0.0968, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 6.13607188703466, | |
| "grad_norm": 5213.7578125, | |
| "learning_rate": 4.4247787610619477e-05, | |
| "loss": 0.0987, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 6.161745827984595, | |
| "grad_norm": 5425.05615234375, | |
| "learning_rate": 4.395280235988201e-05, | |
| "loss": 0.102, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 6.187419768934531, | |
| "grad_norm": 4345.66552734375, | |
| "learning_rate": 4.3657817109144546e-05, | |
| "loss": 0.0978, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 6.213093709884467, | |
| "grad_norm": 5057.90087890625, | |
| "learning_rate": 4.3362831858407084e-05, | |
| "loss": 0.1011, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 6.238767650834403, | |
| "grad_norm": 6916.2607421875, | |
| "learning_rate": 4.306784660766962e-05, | |
| "loss": 0.1023, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 6.264441591784339, | |
| "grad_norm": 6013.05126953125, | |
| "learning_rate": 4.2772861356932154e-05, | |
| "loss": 0.0995, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 6.290115532734275, | |
| "grad_norm": 4742.91357421875, | |
| "learning_rate": 4.247787610619469e-05, | |
| "loss": 0.0974, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 6.315789473684211, | |
| "grad_norm": 4979.93115234375, | |
| "learning_rate": 4.218289085545723e-05, | |
| "loss": 0.1019, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.341463414634147, | |
| "grad_norm": 5349.9130859375, | |
| "learning_rate": 4.188790560471977e-05, | |
| "loss": 0.1027, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 6.367137355584082, | |
| "grad_norm": 5003.3203125, | |
| "learning_rate": 4.15929203539823e-05, | |
| "loss": 0.1038, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 6.392811296534018, | |
| "grad_norm": 5897.6796875, | |
| "learning_rate": 4.129793510324484e-05, | |
| "loss": 0.0998, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 6.418485237483954, | |
| "grad_norm": 5018.42138671875, | |
| "learning_rate": 4.1002949852507376e-05, | |
| "loss": 0.0962, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.418485237483954, | |
| "eval_gen_len": 38.946, | |
| "eval_loss": 0.09220927208662033, | |
| "eval_rouge1": 0.6077, | |
| "eval_rouge2": 0.3279, | |
| "eval_rougeL": 0.5755, | |
| "eval_runtime": 65.0093, | |
| "eval_samples_per_second": 7.691, | |
| "eval_steps_per_second": 0.492, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.44415917843389, | |
| "grad_norm": 6092.09375, | |
| "learning_rate": 4.0707964601769914e-05, | |
| "loss": 0.0929, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 6.469833119383825, | |
| "grad_norm": 6269.76171875, | |
| "learning_rate": 4.0412979351032446e-05, | |
| "loss": 0.0972, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 6.495507060333761, | |
| "grad_norm": 4338.68896484375, | |
| "learning_rate": 4.0117994100294984e-05, | |
| "loss": 0.1009, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 6.521181001283697, | |
| "grad_norm": 4670.00537109375, | |
| "learning_rate": 3.982300884955752e-05, | |
| "loss": 0.1032, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 6.546854942233633, | |
| "grad_norm": 5199.564453125, | |
| "learning_rate": 3.952802359882006e-05, | |
| "loss": 0.0977, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 6.572528883183569, | |
| "grad_norm": 6262.904296875, | |
| "learning_rate": 3.92330383480826e-05, | |
| "loss": 0.0966, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 6.598202824133504, | |
| "grad_norm": 7214.66748046875, | |
| "learning_rate": 3.893805309734514e-05, | |
| "loss": 0.0942, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 6.62387676508344, | |
| "grad_norm": 5746.22705078125, | |
| "learning_rate": 3.8643067846607675e-05, | |
| "loss": 0.0946, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 6.649550706033376, | |
| "grad_norm": 6876.60986328125, | |
| "learning_rate": 3.834808259587021e-05, | |
| "loss": 0.1031, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 6.675224646983312, | |
| "grad_norm": 5216.8642578125, | |
| "learning_rate": 3.8053097345132744e-05, | |
| "loss": 0.0984, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.700898587933247, | |
| "grad_norm": 5965.8583984375, | |
| "learning_rate": 3.775811209439528e-05, | |
| "loss": 0.099, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 6.7265725288831835, | |
| "grad_norm": 7099.044921875, | |
| "learning_rate": 3.746312684365782e-05, | |
| "loss": 0.1052, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 6.7522464698331195, | |
| "grad_norm": 4748.5703125, | |
| "learning_rate": 3.716814159292036e-05, | |
| "loss": 0.1045, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 6.7779204107830555, | |
| "grad_norm": 5743.19921875, | |
| "learning_rate": 3.687315634218289e-05, | |
| "loss": 0.0937, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.803594351732991, | |
| "grad_norm": 5680.45068359375, | |
| "learning_rate": 3.657817109144543e-05, | |
| "loss": 0.0965, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.829268292682927, | |
| "grad_norm": 7245.03564453125, | |
| "learning_rate": 3.628318584070797e-05, | |
| "loss": 0.0909, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 6.854942233632863, | |
| "grad_norm": 5226.4365234375, | |
| "learning_rate": 3.5988200589970505e-05, | |
| "loss": 0.0987, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 6.880616174582799, | |
| "grad_norm": 5511.99853515625, | |
| "learning_rate": 3.5693215339233036e-05, | |
| "loss": 0.1066, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 6.906290115532734, | |
| "grad_norm": 5711.359375, | |
| "learning_rate": 3.5398230088495574e-05, | |
| "loss": 0.0995, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 6.93196405648267, | |
| "grad_norm": 5092.283203125, | |
| "learning_rate": 3.510324483775811e-05, | |
| "loss": 0.0971, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.957637997432606, | |
| "grad_norm": 6100.78271484375, | |
| "learning_rate": 3.480825958702065e-05, | |
| "loss": 0.1009, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 6.983311938382542, | |
| "grad_norm": 5600.61181640625, | |
| "learning_rate": 3.451327433628319e-05, | |
| "loss": 0.1026, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 7.008985879332478, | |
| "grad_norm": 5000.9541015625, | |
| "learning_rate": 3.421828908554573e-05, | |
| "loss": 0.101, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 7.034659820282413, | |
| "grad_norm": 5288.25048828125, | |
| "learning_rate": 3.3923303834808265e-05, | |
| "loss": 0.0961, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 7.060333761232349, | |
| "grad_norm": 5404.33837890625, | |
| "learning_rate": 3.3628318584070804e-05, | |
| "loss": 0.1074, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 7.086007702182285, | |
| "grad_norm": 4586.51708984375, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0978, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 7.111681643132221, | |
| "grad_norm": 5383.466796875, | |
| "learning_rate": 3.303834808259587e-05, | |
| "loss": 0.0983, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 7.137355584082156, | |
| "grad_norm": 5845.02294921875, | |
| "learning_rate": 3.274336283185841e-05, | |
| "loss": 0.0922, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 7.163029525032092, | |
| "grad_norm": 5654.388671875, | |
| "learning_rate": 3.244837758112095e-05, | |
| "loss": 0.0941, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 7.188703465982028, | |
| "grad_norm": 5832.42724609375, | |
| "learning_rate": 3.215339233038348e-05, | |
| "loss": 0.0949, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 7.214377406931964, | |
| "grad_norm": 5299.91015625, | |
| "learning_rate": 3.185840707964602e-05, | |
| "loss": 0.0988, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 7.2400513478818995, | |
| "grad_norm": 5652.5087890625, | |
| "learning_rate": 3.156342182890856e-05, | |
| "loss": 0.0982, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 7.2657252888318355, | |
| "grad_norm": 6181.0361328125, | |
| "learning_rate": 3.1268436578171095e-05, | |
| "loss": 0.098, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 7.291399229781772, | |
| "grad_norm": 13162.5078125, | |
| "learning_rate": 3.097345132743363e-05, | |
| "loss": 0.0951, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 7.317073170731708, | |
| "grad_norm": 5318.009765625, | |
| "learning_rate": 3.0678466076696165e-05, | |
| "loss": 0.0988, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 7.342747111681643, | |
| "grad_norm": 5820.310546875, | |
| "learning_rate": 3.0383480825958703e-05, | |
| "loss": 0.0983, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 7.368421052631579, | |
| "grad_norm": 5990.56640625, | |
| "learning_rate": 3.008849557522124e-05, | |
| "loss": 0.0982, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.394094993581515, | |
| "grad_norm": 5594.0703125, | |
| "learning_rate": 2.9793510324483776e-05, | |
| "loss": 0.0974, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 7.419768934531451, | |
| "grad_norm": 6317.15234375, | |
| "learning_rate": 2.9498525073746314e-05, | |
| "loss": 0.0932, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 7.445442875481387, | |
| "grad_norm": 8022.15185546875, | |
| "learning_rate": 2.9203539823008852e-05, | |
| "loss": 0.1041, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 7.471116816431322, | |
| "grad_norm": 5091.68310546875, | |
| "learning_rate": 2.890855457227139e-05, | |
| "loss": 0.0995, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 7.496790757381258, | |
| "grad_norm": 6386.40869140625, | |
| "learning_rate": 2.8613569321533922e-05, | |
| "loss": 0.0964, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 7.522464698331194, | |
| "grad_norm": 4850.58203125, | |
| "learning_rate": 2.831858407079646e-05, | |
| "loss": 0.1063, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 7.548138639281129, | |
| "grad_norm": 6846.75146484375, | |
| "learning_rate": 2.8023598820059e-05, | |
| "loss": 0.102, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 7.573812580231065, | |
| "grad_norm": 5613.95166015625, | |
| "learning_rate": 2.7728613569321537e-05, | |
| "loss": 0.0977, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 7.599486521181001, | |
| "grad_norm": 5055.47705078125, | |
| "learning_rate": 2.743362831858407e-05, | |
| "loss": 0.0937, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 7.625160462130937, | |
| "grad_norm": 5020.2568359375, | |
| "learning_rate": 2.713864306784661e-05, | |
| "loss": 0.0978, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 7.650834403080873, | |
| "grad_norm": 5974.265625, | |
| "learning_rate": 2.6843657817109148e-05, | |
| "loss": 0.098, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 7.676508344030808, | |
| "grad_norm": 6458.8662109375, | |
| "learning_rate": 2.6548672566371686e-05, | |
| "loss": 0.0964, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 7.702182284980744, | |
| "grad_norm": 5247.0791015625, | |
| "learning_rate": 2.6253687315634217e-05, | |
| "loss": 0.1029, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.702182284980744, | |
| "eval_gen_len": 38.946, | |
| "eval_loss": 0.09166006743907928, | |
| "eval_rouge1": 0.6133, | |
| "eval_rouge2": 0.3322, | |
| "eval_rougeL": 0.5794, | |
| "eval_runtime": 65.3195, | |
| "eval_samples_per_second": 7.655, | |
| "eval_steps_per_second": 0.49, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.7278562259306804, | |
| "grad_norm": 9114.529296875, | |
| "learning_rate": 2.5958702064896756e-05, | |
| "loss": 0.0958, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 7.7535301668806165, | |
| "grad_norm": 4675.69384765625, | |
| "learning_rate": 2.5663716814159294e-05, | |
| "loss": 0.0967, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 7.779204107830552, | |
| "grad_norm": 5986.85546875, | |
| "learning_rate": 2.5368731563421832e-05, | |
| "loss": 0.0963, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 7.804878048780488, | |
| "grad_norm": 5686.59716796875, | |
| "learning_rate": 2.5073746312684367e-05, | |
| "loss": 0.1, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 7.830551989730424, | |
| "grad_norm": 4628.58447265625, | |
| "learning_rate": 2.4778761061946905e-05, | |
| "loss": 0.0971, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 7.85622593068036, | |
| "grad_norm": 4568.95068359375, | |
| "learning_rate": 2.4483775811209443e-05, | |
| "loss": 0.0995, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 7.881899871630296, | |
| "grad_norm": 5026.5517578125, | |
| "learning_rate": 2.4188790560471978e-05, | |
| "loss": 0.0997, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 7.907573812580231, | |
| "grad_norm": 5142.33544921875, | |
| "learning_rate": 2.3893805309734516e-05, | |
| "loss": 0.0989, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 7.933247753530167, | |
| "grad_norm": 4715.99169921875, | |
| "learning_rate": 2.359882005899705e-05, | |
| "loss": 0.0982, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 7.958921694480103, | |
| "grad_norm": 7074.0263671875, | |
| "learning_rate": 2.330383480825959e-05, | |
| "loss": 0.0936, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.984595635430038, | |
| "grad_norm": 5483.7958984375, | |
| "learning_rate": 2.3008849557522124e-05, | |
| "loss": 0.1055, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 8.010269576379974, | |
| "grad_norm": 21462.302734375, | |
| "learning_rate": 2.2713864306784662e-05, | |
| "loss": 0.097, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 8.03594351732991, | |
| "grad_norm": 5375.9345703125, | |
| "learning_rate": 2.24188790560472e-05, | |
| "loss": 0.0945, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 8.061617458279846, | |
| "grad_norm": 5927.3203125, | |
| "learning_rate": 2.2123893805309738e-05, | |
| "loss": 0.0919, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 8.087291399229782, | |
| "grad_norm": 4952.16064453125, | |
| "learning_rate": 2.1828908554572273e-05, | |
| "loss": 0.1023, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 8.112965340179718, | |
| "grad_norm": 4753.6865234375, | |
| "learning_rate": 2.153392330383481e-05, | |
| "loss": 0.1038, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 8.138639281129654, | |
| "grad_norm": 4751.951171875, | |
| "learning_rate": 2.1238938053097346e-05, | |
| "loss": 0.0953, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 8.164313222079588, | |
| "grad_norm": 5049.07470703125, | |
| "learning_rate": 2.0943952802359884e-05, | |
| "loss": 0.0995, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 8.189987163029524, | |
| "grad_norm": 5914.5595703125, | |
| "learning_rate": 2.064896755162242e-05, | |
| "loss": 0.0994, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 8.21566110397946, | |
| "grad_norm": 5663.07568359375, | |
| "learning_rate": 2.0353982300884957e-05, | |
| "loss": 0.099, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 8.241335044929397, | |
| "grad_norm": 5172.39208984375, | |
| "learning_rate": 2.0058997050147492e-05, | |
| "loss": 0.1056, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 8.267008985879333, | |
| "grad_norm": 4296.75732421875, | |
| "learning_rate": 1.976401179941003e-05, | |
| "loss": 0.0917, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 8.292682926829269, | |
| "grad_norm": 5991.23046875, | |
| "learning_rate": 1.946902654867257e-05, | |
| "loss": 0.1005, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 8.318356867779205, | |
| "grad_norm": 4786.93017578125, | |
| "learning_rate": 1.9174041297935107e-05, | |
| "loss": 0.1013, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 8.34403080872914, | |
| "grad_norm": 5587.01416015625, | |
| "learning_rate": 1.887905604719764e-05, | |
| "loss": 0.1061, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 8.369704749679077, | |
| "grad_norm": 5002.3935546875, | |
| "learning_rate": 1.858407079646018e-05, | |
| "loss": 0.0981, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 8.39537869062901, | |
| "grad_norm": 5792.34814453125, | |
| "learning_rate": 1.8289085545722714e-05, | |
| "loss": 0.0982, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 8.421052631578947, | |
| "grad_norm": 5482.14501953125, | |
| "learning_rate": 1.7994100294985252e-05, | |
| "loss": 0.0977, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.446726572528883, | |
| "grad_norm": 5414.59326171875, | |
| "learning_rate": 1.7699115044247787e-05, | |
| "loss": 0.0959, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 8.472400513478819, | |
| "grad_norm": 6676.62548828125, | |
| "learning_rate": 1.7404129793510325e-05, | |
| "loss": 0.0964, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 8.498074454428755, | |
| "grad_norm": 5211.7705078125, | |
| "learning_rate": 1.7109144542772864e-05, | |
| "loss": 0.0936, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 8.523748395378691, | |
| "grad_norm": 5187.91015625, | |
| "learning_rate": 1.6814159292035402e-05, | |
| "loss": 0.092, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 8.549422336328627, | |
| "grad_norm": 16135.931640625, | |
| "learning_rate": 1.6519174041297937e-05, | |
| "loss": 0.093, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 8.575096277278563, | |
| "grad_norm": 5429.2236328125, | |
| "learning_rate": 1.6224188790560475e-05, | |
| "loss": 0.0957, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 8.600770218228497, | |
| "grad_norm": 5034.25732421875, | |
| "learning_rate": 1.592920353982301e-05, | |
| "loss": 0.0997, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 8.626444159178433, | |
| "grad_norm": 6611.349609375, | |
| "learning_rate": 1.5634218289085548e-05, | |
| "loss": 0.0963, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 8.65211810012837, | |
| "grad_norm": 5671.7568359375, | |
| "learning_rate": 1.5339233038348082e-05, | |
| "loss": 0.1065, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 8.677792041078305, | |
| "grad_norm": 8826.3564453125, | |
| "learning_rate": 1.504424778761062e-05, | |
| "loss": 0.0972, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 8.703465982028241, | |
| "grad_norm": 5669.00439453125, | |
| "learning_rate": 1.4749262536873157e-05, | |
| "loss": 0.0995, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 8.729139922978177, | |
| "grad_norm": 7719.87353515625, | |
| "learning_rate": 1.4454277286135695e-05, | |
| "loss": 0.0898, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.754813863928113, | |
| "grad_norm": 5668.51953125, | |
| "learning_rate": 1.415929203539823e-05, | |
| "loss": 0.1028, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 8.78048780487805, | |
| "grad_norm": 5719.044921875, | |
| "learning_rate": 1.3864306784660768e-05, | |
| "loss": 0.094, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 8.806161745827985, | |
| "grad_norm": 6085.166015625, | |
| "learning_rate": 1.3569321533923305e-05, | |
| "loss": 0.0938, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 8.83183568677792, | |
| "grad_norm": 5559.7431640625, | |
| "learning_rate": 1.3274336283185843e-05, | |
| "loss": 0.0962, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 8.857509627727856, | |
| "grad_norm": 7504.08349609375, | |
| "learning_rate": 1.2979351032448378e-05, | |
| "loss": 0.0964, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 8.883183568677792, | |
| "grad_norm": 6102.48486328125, | |
| "learning_rate": 1.2684365781710916e-05, | |
| "loss": 0.0967, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 8.908857509627728, | |
| "grad_norm": 5322.08251953125, | |
| "learning_rate": 1.2389380530973452e-05, | |
| "loss": 0.0952, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 8.934531450577664, | |
| "grad_norm": 5769.94091796875, | |
| "learning_rate": 1.2094395280235989e-05, | |
| "loss": 0.0957, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 8.9602053915276, | |
| "grad_norm": 4727.4755859375, | |
| "learning_rate": 1.1799410029498525e-05, | |
| "loss": 0.0942, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 8.985879332477536, | |
| "grad_norm": 4858.51416015625, | |
| "learning_rate": 1.1504424778761062e-05, | |
| "loss": 0.0977, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.985879332477536, | |
| "eval_gen_len": 38.946, | |
| "eval_loss": 0.09172539412975311, | |
| "eval_rouge1": 0.6126, | |
| "eval_rouge2": 0.3339, | |
| "eval_rougeL": 0.5795, | |
| "eval_runtime": 65.0875, | |
| "eval_samples_per_second": 7.682, | |
| "eval_steps_per_second": 0.492, | |
| "step": 3500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3890, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.517722961707008e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |