| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9247018197842845, | |
| "eval_steps": 500, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011558772747303555, | |
| "grad_norm": 166.625, | |
| "learning_rate": 1.1547344110854505e-07, | |
| "loss": 70.1991, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.002311754549460711, | |
| "grad_norm": 155.125, | |
| "learning_rate": 2.309468822170901e-07, | |
| "loss": 71.3333, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0034676318241910666, | |
| "grad_norm": 148.75, | |
| "learning_rate": 3.464203233256351e-07, | |
| "loss": 69.9043, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.004623509098921422, | |
| "grad_norm": 157.625, | |
| "learning_rate": 4.618937644341802e-07, | |
| "loss": 70.8317, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.005779386373651777, | |
| "grad_norm": 153.25, | |
| "learning_rate": 5.773672055427252e-07, | |
| "loss": 70.7937, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.006935263648382133, | |
| "grad_norm": 162.25, | |
| "learning_rate": 6.928406466512702e-07, | |
| "loss": 70.2267, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.00809114092311249, | |
| "grad_norm": 160.875, | |
| "learning_rate": 8.083140877598153e-07, | |
| "loss": 70.2227, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.009247018197842844, | |
| "grad_norm": 150.25, | |
| "learning_rate": 9.237875288683604e-07, | |
| "loss": 70.1021, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0104028954725732, | |
| "grad_norm": 156.125, | |
| "learning_rate": 1.0392609699769055e-06, | |
| "loss": 69.8542, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.011558772747303554, | |
| "grad_norm": 160.0, | |
| "learning_rate": 1.1547344110854503e-06, | |
| "loss": 69.067, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.01271465002203391, | |
| "grad_norm": 151.375, | |
| "learning_rate": 1.2702078521939956e-06, | |
| "loss": 69.1265, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.013870527296764266, | |
| "grad_norm": 155.75, | |
| "learning_rate": 1.3856812933025404e-06, | |
| "loss": 69.4999, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.015026404571494621, | |
| "grad_norm": 147.5, | |
| "learning_rate": 1.5011547344110855e-06, | |
| "loss": 69.5665, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.01618228184622498, | |
| "grad_norm": 144.125, | |
| "learning_rate": 1.6166281755196305e-06, | |
| "loss": 69.0875, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.01733815912095533, | |
| "grad_norm": 165.875, | |
| "learning_rate": 1.7321016166281756e-06, | |
| "loss": 70.327, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.018494036395685687, | |
| "grad_norm": 157.625, | |
| "learning_rate": 1.8475750577367208e-06, | |
| "loss": 68.3475, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.019649913670416044, | |
| "grad_norm": 159.0, | |
| "learning_rate": 1.9630484988452657e-06, | |
| "loss": 70.1929, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0208057909451464, | |
| "grad_norm": 163.0, | |
| "learning_rate": 2.078521939953811e-06, | |
| "loss": 69.7047, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.021961668219876756, | |
| "grad_norm": 155.5, | |
| "learning_rate": 2.1939953810623558e-06, | |
| "loss": 69.5095, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.02311754549460711, | |
| "grad_norm": 159.0, | |
| "learning_rate": 2.3094688221709006e-06, | |
| "loss": 70.2678, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.024273422769337465, | |
| "grad_norm": 159.125, | |
| "learning_rate": 2.424942263279446e-06, | |
| "loss": 68.9641, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.02542930004406782, | |
| "grad_norm": 164.5, | |
| "learning_rate": 2.540415704387991e-06, | |
| "loss": 68.6594, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.026585177318798177, | |
| "grad_norm": 159.0, | |
| "learning_rate": 2.6558891454965355e-06, | |
| "loss": 70.6938, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.027741054593528533, | |
| "grad_norm": 160.0, | |
| "learning_rate": 2.771362586605081e-06, | |
| "loss": 70.2793, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.02889693186825889, | |
| "grad_norm": 158.125, | |
| "learning_rate": 2.886836027713626e-06, | |
| "loss": 68.5085, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.030052809142989242, | |
| "grad_norm": 163.875, | |
| "learning_rate": 3.002309468822171e-06, | |
| "loss": 69.7674, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.031208686417719598, | |
| "grad_norm": 167.5, | |
| "learning_rate": 3.117782909930716e-06, | |
| "loss": 71.2427, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.03236456369244996, | |
| "grad_norm": 155.125, | |
| "learning_rate": 3.233256351039261e-06, | |
| "loss": 69.6699, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03352044096718031, | |
| "grad_norm": 167.875, | |
| "learning_rate": 3.3487297921478063e-06, | |
| "loss": 68.7639, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.03467631824191066, | |
| "grad_norm": 164.875, | |
| "learning_rate": 3.464203233256351e-06, | |
| "loss": 68.8589, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.03583219551664102, | |
| "grad_norm": 167.125, | |
| "learning_rate": 3.5796766743648964e-06, | |
| "loss": 69.8009, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.036988072791371375, | |
| "grad_norm": 161.5, | |
| "learning_rate": 3.6951501154734416e-06, | |
| "loss": 69.6292, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.03814395006610173, | |
| "grad_norm": 166.5, | |
| "learning_rate": 3.8106235565819865e-06, | |
| "loss": 70.7807, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.03929982734083209, | |
| "grad_norm": 163.875, | |
| "learning_rate": 3.926096997690531e-06, | |
| "loss": 72.3659, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.04045570461556244, | |
| "grad_norm": 176.75, | |
| "learning_rate": 4.041570438799077e-06, | |
| "loss": 69.9732, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.0416115818902928, | |
| "grad_norm": 162.125, | |
| "learning_rate": 4.157043879907622e-06, | |
| "loss": 68.9829, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.042767459165023156, | |
| "grad_norm": 166.875, | |
| "learning_rate": 4.272517321016167e-06, | |
| "loss": 70.6429, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.04392333643975351, | |
| "grad_norm": 158.0, | |
| "learning_rate": 4.3879907621247115e-06, | |
| "loss": 71.5107, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.04507921371448387, | |
| "grad_norm": 166.625, | |
| "learning_rate": 4.503464203233257e-06, | |
| "loss": 71.8921, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.04623509098921422, | |
| "grad_norm": 182.625, | |
| "learning_rate": 4.618937644341801e-06, | |
| "loss": 69.8707, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04739096826394457, | |
| "grad_norm": 152.0, | |
| "learning_rate": 4.7344110854503465e-06, | |
| "loss": 70.0987, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.04854684553867493, | |
| "grad_norm": 183.0, | |
| "learning_rate": 4.849884526558892e-06, | |
| "loss": 70.5575, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.049702722813405285, | |
| "grad_norm": 185.375, | |
| "learning_rate": 4.965357967667437e-06, | |
| "loss": 69.5014, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.05085860008813564, | |
| "grad_norm": 159.625, | |
| "learning_rate": 5.080831408775982e-06, | |
| "loss": 69.1703, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.052014477362866, | |
| "grad_norm": 169.375, | |
| "learning_rate": 5.196304849884527e-06, | |
| "loss": 71.4296, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.053170354637596354, | |
| "grad_norm": 161.125, | |
| "learning_rate": 5.311778290993071e-06, | |
| "loss": 70.7405, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.05432623191232671, | |
| "grad_norm": 170.5, | |
| "learning_rate": 5.427251732101617e-06, | |
| "loss": 71.3609, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.055482109187057066, | |
| "grad_norm": 160.125, | |
| "learning_rate": 5.542725173210162e-06, | |
| "loss": 71.3185, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.05663798646178742, | |
| "grad_norm": 166.875, | |
| "learning_rate": 5.658198614318708e-06, | |
| "loss": 70.5645, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.05779386373651778, | |
| "grad_norm": 184.5, | |
| "learning_rate": 5.773672055427252e-06, | |
| "loss": 71.7371, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05894974101124813, | |
| "grad_norm": 165.5, | |
| "learning_rate": 5.889145496535797e-06, | |
| "loss": 71.3129, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.060105618285978483, | |
| "grad_norm": 167.25, | |
| "learning_rate": 6.004618937644342e-06, | |
| "loss": 70.0941, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.06126149556070884, | |
| "grad_norm": 164.875, | |
| "learning_rate": 6.120092378752888e-06, | |
| "loss": 71.1921, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.062417372835439196, | |
| "grad_norm": 164.25, | |
| "learning_rate": 6.235565819861432e-06, | |
| "loss": 71.2645, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.06357325011016955, | |
| "grad_norm": 174.875, | |
| "learning_rate": 6.351039260969978e-06, | |
| "loss": 71.5271, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.06472912738489991, | |
| "grad_norm": 166.5, | |
| "learning_rate": 6.466512702078522e-06, | |
| "loss": 71.6864, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.06588500465963026, | |
| "grad_norm": 177.875, | |
| "learning_rate": 6.581986143187068e-06, | |
| "loss": 70.2806, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.06704088193436061, | |
| "grad_norm": 166.875, | |
| "learning_rate": 6.6974595842956126e-06, | |
| "loss": 71.3358, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.06819675920909098, | |
| "grad_norm": 184.5, | |
| "learning_rate": 6.812933025404158e-06, | |
| "loss": 71.4115, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.06935263648382133, | |
| "grad_norm": 164.125, | |
| "learning_rate": 6.928406466512702e-06, | |
| "loss": 71.433, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07050851375855169, | |
| "grad_norm": 181.5, | |
| "learning_rate": 7.0438799076212475e-06, | |
| "loss": 70.4249, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.07166439103328204, | |
| "grad_norm": 177.75, | |
| "learning_rate": 7.159353348729793e-06, | |
| "loss": 72.4122, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.0728202683080124, | |
| "grad_norm": 172.75, | |
| "learning_rate": 7.274826789838337e-06, | |
| "loss": 71.6193, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.07397614558274275, | |
| "grad_norm": 173.5, | |
| "learning_rate": 7.390300230946883e-06, | |
| "loss": 71.3936, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.07513202285747311, | |
| "grad_norm": 199.25, | |
| "learning_rate": 7.505773672055428e-06, | |
| "loss": 72.5941, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.07628790013220346, | |
| "grad_norm": 183.625, | |
| "learning_rate": 7.621247113163973e-06, | |
| "loss": 71.2533, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.07744377740693383, | |
| "grad_norm": 184.75, | |
| "learning_rate": 7.736720554272517e-06, | |
| "loss": 71.9591, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.07859965468166417, | |
| "grad_norm": 189.125, | |
| "learning_rate": 7.852193995381063e-06, | |
| "loss": 71.2283, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.07975553195639452, | |
| "grad_norm": 178.375, | |
| "learning_rate": 7.967667436489608e-06, | |
| "loss": 71.2926, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.08091140923112489, | |
| "grad_norm": 176.0, | |
| "learning_rate": 8.083140877598153e-06, | |
| "loss": 69.9807, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.08206728650585524, | |
| "grad_norm": 164.625, | |
| "learning_rate": 8.198614318706698e-06, | |
| "loss": 70.6723, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.0832231637805856, | |
| "grad_norm": 169.75, | |
| "learning_rate": 8.314087759815244e-06, | |
| "loss": 71.8471, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.08437904105531595, | |
| "grad_norm": 170.0, | |
| "learning_rate": 8.429561200923789e-06, | |
| "loss": 72.2858, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.08553491833004631, | |
| "grad_norm": 170.5, | |
| "learning_rate": 8.545034642032334e-06, | |
| "loss": 72.6236, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.08669079560477666, | |
| "grad_norm": 152.0, | |
| "learning_rate": 8.660508083140878e-06, | |
| "loss": 73.0968, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.08784667287950702, | |
| "grad_norm": 171.375, | |
| "learning_rate": 8.775981524249423e-06, | |
| "loss": 72.5791, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.08900255015423737, | |
| "grad_norm": 172.875, | |
| "learning_rate": 8.891454965357968e-06, | |
| "loss": 71.3901, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.09015842742896774, | |
| "grad_norm": 170.25, | |
| "learning_rate": 9.006928406466514e-06, | |
| "loss": 72.0083, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.09131430470369808, | |
| "grad_norm": 187.75, | |
| "learning_rate": 9.122401847575059e-06, | |
| "loss": 73.1241, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.09247018197842843, | |
| "grad_norm": 166.375, | |
| "learning_rate": 9.237875288683602e-06, | |
| "loss": 72.1686, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.0936260592531588, | |
| "grad_norm": 174.0, | |
| "learning_rate": 9.35334872979215e-06, | |
| "loss": 72.8243, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.09478193652788915, | |
| "grad_norm": 166.0, | |
| "learning_rate": 9.468822170900693e-06, | |
| "loss": 73.6437, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.09593781380261951, | |
| "grad_norm": 172.375, | |
| "learning_rate": 9.584295612009238e-06, | |
| "loss": 73.1887, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.09709369107734986, | |
| "grad_norm": 177.5, | |
| "learning_rate": 9.699769053117783e-06, | |
| "loss": 73.1708, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.09824956835208022, | |
| "grad_norm": 154.875, | |
| "learning_rate": 9.815242494226329e-06, | |
| "loss": 73.3365, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.09940544562681057, | |
| "grad_norm": 181.25, | |
| "learning_rate": 9.930715935334874e-06, | |
| "loss": 73.2845, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.10056132290154093, | |
| "grad_norm": 168.5, | |
| "learning_rate": 9.99486125385406e-06, | |
| "loss": 73.5392, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.10171720017627128, | |
| "grad_norm": 187.0, | |
| "learning_rate": 9.98201438848921e-06, | |
| "loss": 72.7898, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.10287307745100165, | |
| "grad_norm": 161.625, | |
| "learning_rate": 9.969167523124359e-06, | |
| "loss": 74.7973, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.104028954725732, | |
| "grad_norm": 163.375, | |
| "learning_rate": 9.956320657759508e-06, | |
| "loss": 74.4268, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.10518483200046234, | |
| "grad_norm": 172.25, | |
| "learning_rate": 9.943473792394657e-06, | |
| "loss": 73.9652, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.10634070927519271, | |
| "grad_norm": 162.875, | |
| "learning_rate": 9.930626927029806e-06, | |
| "loss": 73.0815, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.10749658654992306, | |
| "grad_norm": 164.625, | |
| "learning_rate": 9.917780061664954e-06, | |
| "loss": 72.6958, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.10865246382465342, | |
| "grad_norm": 170.625, | |
| "learning_rate": 9.904933196300103e-06, | |
| "loss": 73.5928, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.10980834109938377, | |
| "grad_norm": 189.25, | |
| "learning_rate": 9.892086330935252e-06, | |
| "loss": 73.1865, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.11096421837411413, | |
| "grad_norm": 181.875, | |
| "learning_rate": 9.879239465570401e-06, | |
| "loss": 73.976, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.11212009564884448, | |
| "grad_norm": 165.125, | |
| "learning_rate": 9.866392600205552e-06, | |
| "loss": 74.3113, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.11327597292357484, | |
| "grad_norm": 175.0, | |
| "learning_rate": 9.853545734840699e-06, | |
| "loss": 75.7309, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1144318501983052, | |
| "grad_norm": 190.25, | |
| "learning_rate": 9.84069886947585e-06, | |
| "loss": 73.5145, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.11558772747303556, | |
| "grad_norm": 166.0, | |
| "learning_rate": 9.827852004110998e-06, | |
| "loss": 74.2301, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.11558772747303556, | |
| "eval_loss": NaN, | |
| "eval_runtime": 384.5721, | |
| "eval_samples_per_second": 606.217, | |
| "eval_steps_per_second": 37.889, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1167436047477659, | |
| "grad_norm": 178.25, | |
| "learning_rate": 9.815005138746147e-06, | |
| "loss": 73.0971, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.11789948202249625, | |
| "grad_norm": 176.125, | |
| "learning_rate": 9.802158273381296e-06, | |
| "loss": 73.0252, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.11905535929722662, | |
| "grad_norm": 204.625, | |
| "learning_rate": 9.789311408016445e-06, | |
| "loss": 74.122, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.12021123657195697, | |
| "grad_norm": 175.25, | |
| "learning_rate": 9.776464542651593e-06, | |
| "loss": 75.304, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.12136711384668733, | |
| "grad_norm": 228.125, | |
| "learning_rate": 9.763617677286742e-06, | |
| "loss": 80.1305, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.12252299112141768, | |
| "grad_norm": 351.25, | |
| "learning_rate": 9.750770811921891e-06, | |
| "loss": 98.8667, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.12367886839614804, | |
| "grad_norm": 253.75, | |
| "learning_rate": 9.73792394655704e-06, | |
| "loss": 90.5137, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.12483474567087839, | |
| "grad_norm": 204.0, | |
| "learning_rate": 9.72507708119219e-06, | |
| "loss": 85.6656, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.12599062294560875, | |
| "grad_norm": 164.875, | |
| "learning_rate": 9.712230215827338e-06, | |
| "loss": 81.4311, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.1271465002203391, | |
| "grad_norm": 171.75, | |
| "learning_rate": 9.699383350462488e-06, | |
| "loss": 78.962, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.12830237749506945, | |
| "grad_norm": 181.375, | |
| "learning_rate": 9.686536485097637e-06, | |
| "loss": 78.2494, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.12945825476979983, | |
| "grad_norm": 167.125, | |
| "learning_rate": 9.673689619732786e-06, | |
| "loss": 76.5218, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.13061413204453018, | |
| "grad_norm": 174.625, | |
| "learning_rate": 9.660842754367935e-06, | |
| "loss": 76.3271, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.13177000931926053, | |
| "grad_norm": 163.125, | |
| "learning_rate": 9.647995889003084e-06, | |
| "loss": 77.2531, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.13292588659399088, | |
| "grad_norm": 185.875, | |
| "learning_rate": 9.635149023638232e-06, | |
| "loss": 76.3901, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.13408176386872123, | |
| "grad_norm": 160.5, | |
| "learning_rate": 9.622302158273383e-06, | |
| "loss": 76.1426, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.1352376411434516, | |
| "grad_norm": 170.75, | |
| "learning_rate": 9.609455292908532e-06, | |
| "loss": 75.3614, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.13639351841818195, | |
| "grad_norm": 186.125, | |
| "learning_rate": 9.59660842754368e-06, | |
| "loss": 75.3318, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.1375493956929123, | |
| "grad_norm": 176.5, | |
| "learning_rate": 9.58376156217883e-06, | |
| "loss": 73.9041, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.13870527296764265, | |
| "grad_norm": 170.25, | |
| "learning_rate": 9.570914696813978e-06, | |
| "loss": 74.9994, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.13986115024237303, | |
| "grad_norm": 222.625, | |
| "learning_rate": 9.558067831449127e-06, | |
| "loss": 76.4041, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.14101702751710338, | |
| "grad_norm": 174.625, | |
| "learning_rate": 9.545220966084276e-06, | |
| "loss": 76.0331, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.14217290479183373, | |
| "grad_norm": 169.875, | |
| "learning_rate": 9.532374100719425e-06, | |
| "loss": 74.6435, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.14332878206656408, | |
| "grad_norm": 188.75, | |
| "learning_rate": 9.519527235354574e-06, | |
| "loss": 76.2372, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.14448465934129442, | |
| "grad_norm": 164.25, | |
| "learning_rate": 9.506680369989724e-06, | |
| "loss": 75.1062, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.1456405366160248, | |
| "grad_norm": 165.125, | |
| "learning_rate": 9.493833504624871e-06, | |
| "loss": 75.9294, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.14679641389075515, | |
| "grad_norm": 189.75, | |
| "learning_rate": 9.480986639260022e-06, | |
| "loss": 75.5001, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.1479522911654855, | |
| "grad_norm": 228.5, | |
| "learning_rate": 9.468139773895171e-06, | |
| "loss": 74.1863, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.14910816844021585, | |
| "grad_norm": 222.0, | |
| "learning_rate": 9.45529290853032e-06, | |
| "loss": 76.5168, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.15026404571494623, | |
| "grad_norm": 187.875, | |
| "learning_rate": 9.442446043165469e-06, | |
| "loss": 74.2555, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.15141992298967658, | |
| "grad_norm": 159.0, | |
| "learning_rate": 9.429599177800617e-06, | |
| "loss": 75.0447, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.15257580026440692, | |
| "grad_norm": 158.75, | |
| "learning_rate": 9.416752312435766e-06, | |
| "loss": 73.3359, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.15373167753913727, | |
| "grad_norm": 170.625, | |
| "learning_rate": 9.403905447070915e-06, | |
| "loss": 73.3929, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.15488755481386765, | |
| "grad_norm": 202.5, | |
| "learning_rate": 9.391058581706064e-06, | |
| "loss": 75.0084, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.156043432088598, | |
| "grad_norm": 194.625, | |
| "learning_rate": 9.378211716341213e-06, | |
| "loss": 74.5239, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.15719930936332835, | |
| "grad_norm": 178.375, | |
| "learning_rate": 9.365364850976363e-06, | |
| "loss": 75.5832, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.1583551866380587, | |
| "grad_norm": 162.75, | |
| "learning_rate": 9.35251798561151e-06, | |
| "loss": 74.5449, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.15951106391278905, | |
| "grad_norm": 184.75, | |
| "learning_rate": 9.339671120246661e-06, | |
| "loss": 74.4716, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.16066694118751942, | |
| "grad_norm": 175.0, | |
| "learning_rate": 9.32682425488181e-06, | |
| "loss": 75.2684, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.16182281846224977, | |
| "grad_norm": 173.75, | |
| "learning_rate": 9.313977389516959e-06, | |
| "loss": 73.7116, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.16297869573698012, | |
| "grad_norm": 165.375, | |
| "learning_rate": 9.301130524152108e-06, | |
| "loss": 73.4956, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.16413457301171047, | |
| "grad_norm": 175.875, | |
| "learning_rate": 9.288283658787256e-06, | |
| "loss": 73.5028, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.16529045028644085, | |
| "grad_norm": 171.375, | |
| "learning_rate": 9.275436793422405e-06, | |
| "loss": 73.6011, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.1664463275611712, | |
| "grad_norm": 175.75, | |
| "learning_rate": 9.262589928057554e-06, | |
| "loss": 74.965, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.16760220483590155, | |
| "grad_norm": 160.625, | |
| "learning_rate": 9.249743062692705e-06, | |
| "loss": 74.7802, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.1687580821106319, | |
| "grad_norm": 163.875, | |
| "learning_rate": 9.236896197327852e-06, | |
| "loss": 74.4619, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.16991395938536225, | |
| "grad_norm": 173.0, | |
| "learning_rate": 9.224049331963002e-06, | |
| "loss": 74.9448, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.17106983666009262, | |
| "grad_norm": 193.375, | |
| "learning_rate": 9.21120246659815e-06, | |
| "loss": 74.6345, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.17222571393482297, | |
| "grad_norm": 168.25, | |
| "learning_rate": 9.1983556012333e-06, | |
| "loss": 74.5997, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.17338159120955332, | |
| "grad_norm": 181.375, | |
| "learning_rate": 9.185508735868449e-06, | |
| "loss": 72.9167, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.17453746848428367, | |
| "grad_norm": 209.875, | |
| "learning_rate": 9.172661870503598e-06, | |
| "loss": 75.2782, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.17569334575901405, | |
| "grad_norm": 165.25, | |
| "learning_rate": 9.159815005138747e-06, | |
| "loss": 73.2717, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.1768492230337444, | |
| "grad_norm": 168.75, | |
| "learning_rate": 9.146968139773897e-06, | |
| "loss": 75.1429, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.17800510030847475, | |
| "grad_norm": 161.0, | |
| "learning_rate": 9.134121274409044e-06, | |
| "loss": 74.8651, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.1791609775832051, | |
| "grad_norm": 186.375, | |
| "learning_rate": 9.121274409044195e-06, | |
| "loss": 75.679, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.18031685485793547, | |
| "grad_norm": 183.875, | |
| "learning_rate": 9.108427543679344e-06, | |
| "loss": 74.1174, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.18147273213266582, | |
| "grad_norm": 160.0, | |
| "learning_rate": 9.095580678314493e-06, | |
| "loss": 74.8214, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.18262860940739617, | |
| "grad_norm": 173.625, | |
| "learning_rate": 9.082733812949641e-06, | |
| "loss": 72.8824, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.18378448668212652, | |
| "grad_norm": 184.75, | |
| "learning_rate": 9.06988694758479e-06, | |
| "loss": 75.2577, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.18494036395685687, | |
| "grad_norm": 171.75, | |
| "learning_rate": 9.057040082219939e-06, | |
| "loss": 74.0234, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.18609624123158724, | |
| "grad_norm": 166.125, | |
| "learning_rate": 9.044193216855088e-06, | |
| "loss": 74.5028, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.1872521185063176, | |
| "grad_norm": 181.25, | |
| "learning_rate": 9.031346351490237e-06, | |
| "loss": 74.7906, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.18840799578104794, | |
| "grad_norm": 168.25, | |
| "learning_rate": 9.018499486125386e-06, | |
| "loss": 73.6598, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.1895638730557783, | |
| "grad_norm": 213.625, | |
| "learning_rate": 9.005652620760536e-06, | |
| "loss": 74.4191, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.19071975033050867, | |
| "grad_norm": 184.5, | |
| "learning_rate": 8.992805755395683e-06, | |
| "loss": 74.0181, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.19187562760523902, | |
| "grad_norm": 168.75, | |
| "learning_rate": 8.979958890030834e-06, | |
| "loss": 72.935, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.19303150487996937, | |
| "grad_norm": 176.625, | |
| "learning_rate": 8.967112024665983e-06, | |
| "loss": 74.5831, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.19418738215469972, | |
| "grad_norm": 168.625, | |
| "learning_rate": 8.954265159301132e-06, | |
| "loss": 75.029, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.1953432594294301, | |
| "grad_norm": 188.0, | |
| "learning_rate": 8.94141829393628e-06, | |
| "loss": 75.1383, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.19649913670416044, | |
| "grad_norm": 163.75, | |
| "learning_rate": 8.92857142857143e-06, | |
| "loss": 73.7028, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1976550139788908, | |
| "grad_norm": 168.75, | |
| "learning_rate": 8.915724563206578e-06, | |
| "loss": 73.8515, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.19881089125362114, | |
| "grad_norm": 199.875, | |
| "learning_rate": 8.902877697841727e-06, | |
| "loss": 74.1701, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.1999667685283515, | |
| "grad_norm": 157.5, | |
| "learning_rate": 8.890030832476876e-06, | |
| "loss": 75.171, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.20112264580308187, | |
| "grad_norm": 189.875, | |
| "learning_rate": 8.877183967112025e-06, | |
| "loss": 73.9979, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.20227852307781222, | |
| "grad_norm": 160.75, | |
| "learning_rate": 8.864337101747175e-06, | |
| "loss": 73.9971, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.20343440035254257, | |
| "grad_norm": 170.75, | |
| "learning_rate": 8.851490236382322e-06, | |
| "loss": 73.9474, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.20459027762727292, | |
| "grad_norm": 161.875, | |
| "learning_rate": 8.838643371017473e-06, | |
| "loss": 73.6239, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.2057461549020033, | |
| "grad_norm": 208.25, | |
| "learning_rate": 8.825796505652622e-06, | |
| "loss": 74.3579, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.20690203217673364, | |
| "grad_norm": 182.375, | |
| "learning_rate": 8.81294964028777e-06, | |
| "loss": 73.6577, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.208057909451464, | |
| "grad_norm": 167.375, | |
| "learning_rate": 8.80010277492292e-06, | |
| "loss": 73.492, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.20921378672619434, | |
| "grad_norm": 202.625, | |
| "learning_rate": 8.787255909558068e-06, | |
| "loss": 73.0436, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.2103696640009247, | |
| "grad_norm": 156.5, | |
| "learning_rate": 8.774409044193217e-06, | |
| "loss": 74.3569, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.21152554127565507, | |
| "grad_norm": 167.75, | |
| "learning_rate": 8.761562178828366e-06, | |
| "loss": 73.5566, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.21268141855038541, | |
| "grad_norm": 159.0, | |
| "learning_rate": 8.748715313463516e-06, | |
| "loss": 73.4377, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.21383729582511576, | |
| "grad_norm": 197.5, | |
| "learning_rate": 8.735868448098664e-06, | |
| "loss": 73.5647, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.2149931730998461, | |
| "grad_norm": 157.75, | |
| "learning_rate": 8.723021582733814e-06, | |
| "loss": 74.3935, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.2161490503745765, | |
| "grad_norm": 160.0, | |
| "learning_rate": 8.710174717368961e-06, | |
| "loss": 73.5235, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.21730492764930684, | |
| "grad_norm": 193.5, | |
| "learning_rate": 8.697327852004112e-06, | |
| "loss": 74.8152, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.2184608049240372, | |
| "grad_norm": 166.125, | |
| "learning_rate": 8.68448098663926e-06, | |
| "loss": 74.962, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.21961668219876754, | |
| "grad_norm": 187.75, | |
| "learning_rate": 8.67163412127441e-06, | |
| "loss": 73.3269, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.22077255947349791, | |
| "grad_norm": 177.625, | |
| "learning_rate": 8.658787255909558e-06, | |
| "loss": 72.6171, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.22192843674822826, | |
| "grad_norm": 207.875, | |
| "learning_rate": 8.645940390544709e-06, | |
| "loss": 73.4562, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.2230843140229586, | |
| "grad_norm": 197.0, | |
| "learning_rate": 8.633093525179856e-06, | |
| "loss": 72.9479, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.22424019129768896, | |
| "grad_norm": 178.75, | |
| "learning_rate": 8.620246659815007e-06, | |
| "loss": 73.2069, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.2253960685724193, | |
| "grad_norm": 184.375, | |
| "learning_rate": 8.607399794450156e-06, | |
| "loss": 72.4369, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.2265519458471497, | |
| "grad_norm": 193.625, | |
| "learning_rate": 8.594552929085304e-06, | |
| "loss": 75.2301, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.22770782312188004, | |
| "grad_norm": 160.75, | |
| "learning_rate": 8.581706063720453e-06, | |
| "loss": 74.4445, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.2288637003966104, | |
| "grad_norm": 188.0, | |
| "learning_rate": 8.568859198355602e-06, | |
| "loss": 73.1566, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.23001957767134074, | |
| "grad_norm": 166.0, | |
| "learning_rate": 8.556012332990751e-06, | |
| "loss": 72.9151, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.2311754549460711, | |
| "grad_norm": 165.5, | |
| "learning_rate": 8.5431654676259e-06, | |
| "loss": 72.4977, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2311754549460711, | |
| "eval_loss": NaN, | |
| "eval_runtime": 382.2843, | |
| "eval_samples_per_second": 609.845, | |
| "eval_steps_per_second": 38.116, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.23233133222080146, | |
| "grad_norm": 166.75, | |
| "learning_rate": 8.530318602261049e-06, | |
| "loss": 72.1535, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.2334872094955318, | |
| "grad_norm": 178.5, | |
| "learning_rate": 8.517471736896197e-06, | |
| "loss": 74.0213, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.23464308677026216, | |
| "grad_norm": 156.25, | |
| "learning_rate": 8.504624871531348e-06, | |
| "loss": 74.9422, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.2357989640449925, | |
| "grad_norm": 174.0, | |
| "learning_rate": 8.491778006166495e-06, | |
| "loss": 72.5082, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.2369548413197229, | |
| "grad_norm": 162.5, | |
| "learning_rate": 8.478931140801646e-06, | |
| "loss": 73.302, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.23811071859445324, | |
| "grad_norm": 182.625, | |
| "learning_rate": 8.466084275436795e-06, | |
| "loss": 73.0291, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.23926659586918358, | |
| "grad_norm": 162.875, | |
| "learning_rate": 8.453237410071943e-06, | |
| "loss": 73.237, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.24042247314391393, | |
| "grad_norm": 181.75, | |
| "learning_rate": 8.440390544707092e-06, | |
| "loss": 73.0543, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.2415783504186443, | |
| "grad_norm": 194.25, | |
| "learning_rate": 8.427543679342241e-06, | |
| "loss": 72.1511, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.24273422769337466, | |
| "grad_norm": 197.625, | |
| "learning_rate": 8.41469681397739e-06, | |
| "loss": 72.7044, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.243890104968105, | |
| "grad_norm": 205.75, | |
| "learning_rate": 8.401849948612539e-06, | |
| "loss": 73.4772, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.24504598224283536, | |
| "grad_norm": 179.375, | |
| "learning_rate": 8.38900308324769e-06, | |
| "loss": 73.4765, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.24620185951756574, | |
| "grad_norm": 163.25, | |
| "learning_rate": 8.376156217882836e-06, | |
| "loss": 71.4076, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.24735773679229608, | |
| "grad_norm": 159.75, | |
| "learning_rate": 8.363309352517987e-06, | |
| "loss": 74.3268, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.24851361406702643, | |
| "grad_norm": 174.625, | |
| "learning_rate": 8.350462487153134e-06, | |
| "loss": 71.9133, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.24966949134175678, | |
| "grad_norm": 224.125, | |
| "learning_rate": 8.337615621788285e-06, | |
| "loss": 72.5681, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.25082536861648713, | |
| "grad_norm": 196.25, | |
| "learning_rate": 8.324768756423434e-06, | |
| "loss": 72.7169, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.2519812458912175, | |
| "grad_norm": 163.5, | |
| "learning_rate": 8.311921891058582e-06, | |
| "loss": 73.2148, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.25313712316594783, | |
| "grad_norm": 181.5, | |
| "learning_rate": 8.299075025693731e-06, | |
| "loss": 74.4619, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.2542930004406782, | |
| "grad_norm": 201.125, | |
| "learning_rate": 8.28622816032888e-06, | |
| "loss": 72.7498, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2554488777154086, | |
| "grad_norm": 191.75, | |
| "learning_rate": 8.273381294964029e-06, | |
| "loss": 73.7849, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.2566047549901389, | |
| "grad_norm": 187.375, | |
| "learning_rate": 8.260534429599178e-06, | |
| "loss": 73.536, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.2577606322648693, | |
| "grad_norm": 161.25, | |
| "learning_rate": 8.247687564234328e-06, | |
| "loss": 73.2744, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.25891650953959966, | |
| "grad_norm": 170.875, | |
| "learning_rate": 8.234840698869475e-06, | |
| "loss": 74.2862, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.26007238681433, | |
| "grad_norm": 175.375, | |
| "learning_rate": 8.221993833504626e-06, | |
| "loss": 73.767, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.26122826408906036, | |
| "grad_norm": 175.25, | |
| "learning_rate": 8.209146968139775e-06, | |
| "loss": 72.3037, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.2623841413637907, | |
| "grad_norm": 180.125, | |
| "learning_rate": 8.196300102774924e-06, | |
| "loss": 73.3881, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.26354001863852106, | |
| "grad_norm": 165.25, | |
| "learning_rate": 8.183453237410073e-06, | |
| "loss": 74.2604, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.26469589591325143, | |
| "grad_norm": 208.875, | |
| "learning_rate": 8.170606372045221e-06, | |
| "loss": 73.6266, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.26585177318798175, | |
| "grad_norm": 164.375, | |
| "learning_rate": 8.15775950668037e-06, | |
| "loss": 71.7306, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.26700765046271213, | |
| "grad_norm": 180.5, | |
| "learning_rate": 8.14491264131552e-06, | |
| "loss": 73.3496, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.26816352773744245, | |
| "grad_norm": 161.75, | |
| "learning_rate": 8.132065775950668e-06, | |
| "loss": 71.9542, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.26931940501217283, | |
| "grad_norm": 160.375, | |
| "learning_rate": 8.119218910585818e-06, | |
| "loss": 72.852, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.2704752822869032, | |
| "grad_norm": 166.875, | |
| "learning_rate": 8.106372045220967e-06, | |
| "loss": 73.498, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.27163115956163353, | |
| "grad_norm": 152.5, | |
| "learning_rate": 8.093525179856116e-06, | |
| "loss": 72.89, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.2727870368363639, | |
| "grad_norm": 179.625, | |
| "learning_rate": 8.080678314491265e-06, | |
| "loss": 73.7572, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.2739429141110943, | |
| "grad_norm": 178.625, | |
| "learning_rate": 8.067831449126414e-06, | |
| "loss": 72.1453, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.2750987913858246, | |
| "grad_norm": 165.0, | |
| "learning_rate": 8.054984583761563e-06, | |
| "loss": 73.9004, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.276254668660555, | |
| "grad_norm": 170.875, | |
| "learning_rate": 8.042137718396712e-06, | |
| "loss": 72.7976, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.2774105459352853, | |
| "grad_norm": 163.125, | |
| "learning_rate": 8.029290853031862e-06, | |
| "loss": 73.1718, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.2785664232100157, | |
| "grad_norm": 179.25, | |
| "learning_rate": 8.01644398766701e-06, | |
| "loss": 72.7056, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.27972230048474606, | |
| "grad_norm": 190.625, | |
| "learning_rate": 8.00359712230216e-06, | |
| "loss": 73.2504, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.2808781777594764, | |
| "grad_norm": 171.375, | |
| "learning_rate": 7.990750256937307e-06, | |
| "loss": 72.886, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.28203405503420675, | |
| "grad_norm": 168.25, | |
| "learning_rate": 7.977903391572457e-06, | |
| "loss": 73.0636, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.2831899323089371, | |
| "grad_norm": 170.875, | |
| "learning_rate": 7.965056526207606e-06, | |
| "loss": 73.4683, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.28434580958366745, | |
| "grad_norm": 170.375, | |
| "learning_rate": 7.952209660842755e-06, | |
| "loss": 72.7721, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.28550168685839783, | |
| "grad_norm": 187.875, | |
| "learning_rate": 7.939362795477904e-06, | |
| "loss": 74.0967, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.28665756413312815, | |
| "grad_norm": 206.0, | |
| "learning_rate": 7.926515930113053e-06, | |
| "loss": 72.2294, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.28781344140785853, | |
| "grad_norm": 174.625, | |
| "learning_rate": 7.913669064748202e-06, | |
| "loss": 71.7166, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.28896931868258885, | |
| "grad_norm": 195.0, | |
| "learning_rate": 7.90082219938335e-06, | |
| "loss": 72.4936, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.2901251959573192, | |
| "grad_norm": 174.25, | |
| "learning_rate": 7.887975334018501e-06, | |
| "loss": 71.8098, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.2912810732320496, | |
| "grad_norm": 162.25, | |
| "learning_rate": 7.875128468653648e-06, | |
| "loss": 73.6741, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.2924369505067799, | |
| "grad_norm": 163.0, | |
| "learning_rate": 7.862281603288799e-06, | |
| "loss": 74.2419, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.2935928277815103, | |
| "grad_norm": 156.625, | |
| "learning_rate": 7.849434737923948e-06, | |
| "loss": 71.8441, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.2947487050562407, | |
| "grad_norm": 159.125, | |
| "learning_rate": 7.836587872559096e-06, | |
| "loss": 72.4638, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.295904582330971, | |
| "grad_norm": 163.75, | |
| "learning_rate": 7.823741007194245e-06, | |
| "loss": 72.4716, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.2970604596057014, | |
| "grad_norm": 149.25, | |
| "learning_rate": 7.810894141829394e-06, | |
| "loss": 71.0848, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.2982163368804317, | |
| "grad_norm": 157.625, | |
| "learning_rate": 7.798047276464543e-06, | |
| "loss": 71.6886, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.2993722141551621, | |
| "grad_norm": 161.125, | |
| "learning_rate": 7.785200411099692e-06, | |
| "loss": 72.6646, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.30052809142989245, | |
| "grad_norm": 189.25, | |
| "learning_rate": 7.77235354573484e-06, | |
| "loss": 72.4807, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3016839687046228, | |
| "grad_norm": 207.0, | |
| "learning_rate": 7.75950668036999e-06, | |
| "loss": 72.7943, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.30283984597935315, | |
| "grad_norm": 160.625, | |
| "learning_rate": 7.74665981500514e-06, | |
| "loss": 72.9901, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.30399572325408347, | |
| "grad_norm": 218.375, | |
| "learning_rate": 7.733812949640287e-06, | |
| "loss": 73.4908, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.30515160052881385, | |
| "grad_norm": 166.125, | |
| "learning_rate": 7.720966084275438e-06, | |
| "loss": 72.9456, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.3063074778035442, | |
| "grad_norm": 171.0, | |
| "learning_rate": 7.708119218910587e-06, | |
| "loss": 72.6809, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.30746335507827455, | |
| "grad_norm": 162.75, | |
| "learning_rate": 7.695272353545736e-06, | |
| "loss": 71.7408, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.3086192323530049, | |
| "grad_norm": 155.875, | |
| "learning_rate": 7.682425488180884e-06, | |
| "loss": 73.5375, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.3097751096277353, | |
| "grad_norm": 155.625, | |
| "learning_rate": 7.669578622816033e-06, | |
| "loss": 72.8354, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.3109309869024656, | |
| "grad_norm": 176.75, | |
| "learning_rate": 7.656731757451182e-06, | |
| "loss": 72.5769, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.312086864177196, | |
| "grad_norm": 187.75, | |
| "learning_rate": 7.643884892086333e-06, | |
| "loss": 72.8171, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.3132427414519263, | |
| "grad_norm": 159.25, | |
| "learning_rate": 7.63103802672148e-06, | |
| "loss": 72.2052, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.3143986187266567, | |
| "grad_norm": 163.125, | |
| "learning_rate": 7.6181911613566294e-06, | |
| "loss": 71.9153, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.3155544960013871, | |
| "grad_norm": 171.375, | |
| "learning_rate": 7.605344295991779e-06, | |
| "loss": 70.9257, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.3167103732761174, | |
| "grad_norm": 180.0, | |
| "learning_rate": 7.592497430626927e-06, | |
| "loss": 72.8817, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.3178662505508478, | |
| "grad_norm": 173.625, | |
| "learning_rate": 7.579650565262077e-06, | |
| "loss": 72.5966, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.3190221278255781, | |
| "grad_norm": 160.125, | |
| "learning_rate": 7.566803699897226e-06, | |
| "loss": 72.0935, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.32017800510030847, | |
| "grad_norm": 158.875, | |
| "learning_rate": 7.5539568345323745e-06, | |
| "loss": 72.1024, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.32133388237503885, | |
| "grad_norm": 155.375, | |
| "learning_rate": 7.541109969167524e-06, | |
| "loss": 71.8397, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.32248975964976917, | |
| "grad_norm": 165.5, | |
| "learning_rate": 7.528263103802673e-06, | |
| "loss": 71.5119, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.32364563692449955, | |
| "grad_norm": 170.875, | |
| "learning_rate": 7.515416238437822e-06, | |
| "loss": 73.3594, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.3248015141992299, | |
| "grad_norm": 184.125, | |
| "learning_rate": 7.502569373072971e-06, | |
| "loss": 73.504, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.32595739147396025, | |
| "grad_norm": 165.25, | |
| "learning_rate": 7.4897225077081205e-06, | |
| "loss": 71.1564, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.3271132687486906, | |
| "grad_norm": 175.5, | |
| "learning_rate": 7.4768756423432685e-06, | |
| "loss": 72.8789, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.32826914602342094, | |
| "grad_norm": 168.375, | |
| "learning_rate": 7.464028776978418e-06, | |
| "loss": 73.2979, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.3294250232981513, | |
| "grad_norm": 177.5, | |
| "learning_rate": 7.451181911613566e-06, | |
| "loss": 71.187, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.3305809005728817, | |
| "grad_norm": 160.375, | |
| "learning_rate": 7.438335046248716e-06, | |
| "loss": 73.9314, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.331736777847612, | |
| "grad_norm": 167.625, | |
| "learning_rate": 7.4254881808838655e-06, | |
| "loss": 72.4073, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.3328926551223424, | |
| "grad_norm": 167.0, | |
| "learning_rate": 7.4126413155190135e-06, | |
| "loss": 72.5102, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.3340485323970727, | |
| "grad_norm": 173.5, | |
| "learning_rate": 7.399794450154163e-06, | |
| "loss": 72.4247, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.3352044096718031, | |
| "grad_norm": 169.125, | |
| "learning_rate": 7.386947584789312e-06, | |
| "loss": 71.9371, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.33636028694653347, | |
| "grad_norm": 157.375, | |
| "learning_rate": 7.374100719424461e-06, | |
| "loss": 71.7102, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.3375161642212638, | |
| "grad_norm": 173.625, | |
| "learning_rate": 7.36125385405961e-06, | |
| "loss": 72.5739, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.33867204149599417, | |
| "grad_norm": 196.125, | |
| "learning_rate": 7.3484069886947595e-06, | |
| "loss": 72.1263, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.3398279187707245, | |
| "grad_norm": 154.125, | |
| "learning_rate": 7.3355601233299075e-06, | |
| "loss": 72.4053, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.34098379604545487, | |
| "grad_norm": 182.125, | |
| "learning_rate": 7.322713257965057e-06, | |
| "loss": 72.6574, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.34213967332018524, | |
| "grad_norm": 182.375, | |
| "learning_rate": 7.309866392600207e-06, | |
| "loss": 72.0569, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.34329555059491557, | |
| "grad_norm": 161.375, | |
| "learning_rate": 7.297019527235355e-06, | |
| "loss": 72.4482, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.34445142786964594, | |
| "grad_norm": 165.25, | |
| "learning_rate": 7.2841726618705045e-06, | |
| "loss": 73.3294, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.3456073051443763, | |
| "grad_norm": 162.625, | |
| "learning_rate": 7.2713257965056525e-06, | |
| "loss": 72.414, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.34676318241910664, | |
| "grad_norm": 165.875, | |
| "learning_rate": 7.258478931140802e-06, | |
| "loss": 70.1396, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.34676318241910664, | |
| "eval_loss": NaN, | |
| "eval_runtime": 382.2578, | |
| "eval_samples_per_second": 609.887, | |
| "eval_steps_per_second": 38.118, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.347919059693837, | |
| "grad_norm": 167.875, | |
| "learning_rate": 7.245632065775952e-06, | |
| "loss": 71.7923, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.34907493696856734, | |
| "grad_norm": 180.625, | |
| "learning_rate": 7.2327852004111e-06, | |
| "loss": 72.9641, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.3502308142432977, | |
| "grad_norm": 161.875, | |
| "learning_rate": 7.21993833504625e-06, | |
| "loss": 71.7615, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.3513866915180281, | |
| "grad_norm": 183.5, | |
| "learning_rate": 7.2070914696813985e-06, | |
| "loss": 71.8146, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.3525425687927584, | |
| "grad_norm": 162.0, | |
| "learning_rate": 7.194244604316547e-06, | |
| "loss": 71.0128, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.3536984460674888, | |
| "grad_norm": 163.375, | |
| "learning_rate": 7.181397738951696e-06, | |
| "loss": 71.2744, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.3548543233422191, | |
| "grad_norm": 173.75, | |
| "learning_rate": 7.168550873586846e-06, | |
| "loss": 72.7448, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.3560102006169495, | |
| "grad_norm": 182.625, | |
| "learning_rate": 7.155704008221994e-06, | |
| "loss": 72.3184, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.35716607789167987, | |
| "grad_norm": 169.625, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 73.2466, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.3583219551664102, | |
| "grad_norm": 169.375, | |
| "learning_rate": 7.1300102774922916e-06, | |
| "loss": 72.9034, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.35947783244114057, | |
| "grad_norm": 155.0, | |
| "learning_rate": 7.117163412127441e-06, | |
| "loss": 70.7591, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.36063370971587094, | |
| "grad_norm": 192.25, | |
| "learning_rate": 7.104316546762591e-06, | |
| "loss": 72.9981, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.36178958699060126, | |
| "grad_norm": 167.375, | |
| "learning_rate": 7.091469681397739e-06, | |
| "loss": 71.7857, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 0.36294546426533164, | |
| "grad_norm": 160.5, | |
| "learning_rate": 7.078622816032889e-06, | |
| "loss": 71.0528, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.36410134154006196, | |
| "grad_norm": 178.75, | |
| "learning_rate": 7.0657759506680375e-06, | |
| "loss": 72.1775, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.36525721881479234, | |
| "grad_norm": 161.5, | |
| "learning_rate": 7.052929085303186e-06, | |
| "loss": 71.9378, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.3664130960895227, | |
| "grad_norm": 158.25, | |
| "learning_rate": 7.040082219938336e-06, | |
| "loss": 71.7813, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 0.36756897336425304, | |
| "grad_norm": 160.0, | |
| "learning_rate": 7.027235354573485e-06, | |
| "loss": 71.8137, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.3687248506389834, | |
| "grad_norm": 165.0, | |
| "learning_rate": 7.014388489208634e-06, | |
| "loss": 73.0897, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 0.36988072791371374, | |
| "grad_norm": 166.125, | |
| "learning_rate": 7.0015416238437826e-06, | |
| "loss": 71.867, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.3710366051884441, | |
| "grad_norm": 152.375, | |
| "learning_rate": 6.988694758478932e-06, | |
| "loss": 72.0306, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 0.3721924824631745, | |
| "grad_norm": 153.625, | |
| "learning_rate": 6.97584789311408e-06, | |
| "loss": 71.2797, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.3733483597379048, | |
| "grad_norm": 157.75, | |
| "learning_rate": 6.96300102774923e-06, | |
| "loss": 70.6556, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 0.3745042370126352, | |
| "grad_norm": 206.25, | |
| "learning_rate": 6.950154162384378e-06, | |
| "loss": 70.427, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.37566011428736557, | |
| "grad_norm": 198.0, | |
| "learning_rate": 6.937307297019528e-06, | |
| "loss": 70.8553, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.3768159915620959, | |
| "grad_norm": 172.375, | |
| "learning_rate": 6.924460431654677e-06, | |
| "loss": 72.5536, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.37797186883682626, | |
| "grad_norm": 162.25, | |
| "learning_rate": 6.911613566289825e-06, | |
| "loss": 71.3941, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 0.3791277461115566, | |
| "grad_norm": 154.75, | |
| "learning_rate": 6.898766700924975e-06, | |
| "loss": 70.9165, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.38028362338628696, | |
| "grad_norm": 181.75, | |
| "learning_rate": 6.885919835560124e-06, | |
| "loss": 72.0093, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 0.38143950066101734, | |
| "grad_norm": 150.0, | |
| "learning_rate": 6.873072970195273e-06, | |
| "loss": 71.9246, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.38259537793574766, | |
| "grad_norm": 164.875, | |
| "learning_rate": 6.860226104830422e-06, | |
| "loss": 71.8933, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 0.38375125521047804, | |
| "grad_norm": 157.0, | |
| "learning_rate": 6.847379239465571e-06, | |
| "loss": 72.0258, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.38490713248520836, | |
| "grad_norm": 177.375, | |
| "learning_rate": 6.834532374100719e-06, | |
| "loss": 72.3605, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 0.38606300975993874, | |
| "grad_norm": 175.625, | |
| "learning_rate": 6.821685508735869e-06, | |
| "loss": 71.7816, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.3872188870346691, | |
| "grad_norm": 178.375, | |
| "learning_rate": 6.808838643371019e-06, | |
| "loss": 71.8166, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.38837476430939943, | |
| "grad_norm": 164.0, | |
| "learning_rate": 6.795991778006167e-06, | |
| "loss": 72.0027, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.3895306415841298, | |
| "grad_norm": 168.0, | |
| "learning_rate": 6.783144912641316e-06, | |
| "loss": 71.8431, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 0.3906865188588602, | |
| "grad_norm": 165.375, | |
| "learning_rate": 6.770298047276464e-06, | |
| "loss": 71.9873, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.3918423961335905, | |
| "grad_norm": 161.25, | |
| "learning_rate": 6.757451181911614e-06, | |
| "loss": 72.0379, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 0.3929982734083209, | |
| "grad_norm": 175.125, | |
| "learning_rate": 6.744604316546764e-06, | |
| "loss": 72.1231, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.3941541506830512, | |
| "grad_norm": 177.375, | |
| "learning_rate": 6.731757451181912e-06, | |
| "loss": 70.1524, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 0.3953100279577816, | |
| "grad_norm": 167.75, | |
| "learning_rate": 6.7189105858170614e-06, | |
| "loss": 71.1926, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.39646590523251196, | |
| "grad_norm": 148.125, | |
| "learning_rate": 6.70606372045221e-06, | |
| "loss": 72.7071, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 0.3976217825072423, | |
| "grad_norm": 162.875, | |
| "learning_rate": 6.693216855087359e-06, | |
| "loss": 71.857, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.39877765978197266, | |
| "grad_norm": 198.625, | |
| "learning_rate": 6.680369989722508e-06, | |
| "loss": 72.0683, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.399933537056703, | |
| "grad_norm": 155.625, | |
| "learning_rate": 6.667523124357658e-06, | |
| "loss": 70.8161, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.40108941433143336, | |
| "grad_norm": 164.0, | |
| "learning_rate": 6.654676258992806e-06, | |
| "loss": 70.9434, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 0.40224529160616374, | |
| "grad_norm": 159.375, | |
| "learning_rate": 6.641829393627955e-06, | |
| "loss": 71.407, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.40340116888089406, | |
| "grad_norm": 161.875, | |
| "learning_rate": 6.628982528263105e-06, | |
| "loss": 71.7674, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 0.40455704615562443, | |
| "grad_norm": 162.625, | |
| "learning_rate": 6.616135662898253e-06, | |
| "loss": 72.8434, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.40571292343035475, | |
| "grad_norm": 205.75, | |
| "learning_rate": 6.603288797533403e-06, | |
| "loss": 71.8435, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 0.40686880070508513, | |
| "grad_norm": 149.625, | |
| "learning_rate": 6.590441932168551e-06, | |
| "loss": 71.5685, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.4080246779798155, | |
| "grad_norm": 144.625, | |
| "learning_rate": 6.5775950668037005e-06, | |
| "loss": 71.5258, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 0.40918055525454583, | |
| "grad_norm": 171.5, | |
| "learning_rate": 6.564748201438849e-06, | |
| "loss": 70.664, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.4103364325292762, | |
| "grad_norm": 177.875, | |
| "learning_rate": 6.551901336073998e-06, | |
| "loss": 71.853, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.4114923098040066, | |
| "grad_norm": 168.5, | |
| "learning_rate": 6.539054470709148e-06, | |
| "loss": 71.6207, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.4126481870787369, | |
| "grad_norm": 166.375, | |
| "learning_rate": 6.526207605344297e-06, | |
| "loss": 71.5577, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 0.4138040643534673, | |
| "grad_norm": 187.375, | |
| "learning_rate": 6.5133607399794455e-06, | |
| "loss": 71.5951, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.4149599416281976, | |
| "grad_norm": 167.625, | |
| "learning_rate": 6.500513874614594e-06, | |
| "loss": 72.7333, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 0.416115818902928, | |
| "grad_norm": 177.375, | |
| "learning_rate": 6.487667009249744e-06, | |
| "loss": 71.4623, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.41727169617765836, | |
| "grad_norm": 176.25, | |
| "learning_rate": 6.474820143884892e-06, | |
| "loss": 70.88, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 0.4184275734523887, | |
| "grad_norm": 178.25, | |
| "learning_rate": 6.461973278520042e-06, | |
| "loss": 70.7191, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.41958345072711906, | |
| "grad_norm": 190.75, | |
| "learning_rate": 6.4491264131551915e-06, | |
| "loss": 71.4613, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 0.4207393280018494, | |
| "grad_norm": 175.75, | |
| "learning_rate": 6.4362795477903395e-06, | |
| "loss": 71.9579, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.42189520527657975, | |
| "grad_norm": 159.625, | |
| "learning_rate": 6.423432682425489e-06, | |
| "loss": 70.9512, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.42305108255131013, | |
| "grad_norm": 171.0, | |
| "learning_rate": 6.410585817060637e-06, | |
| "loss": 71.0018, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.42420695982604045, | |
| "grad_norm": 168.625, | |
| "learning_rate": 6.397738951695787e-06, | |
| "loss": 69.9553, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 0.42536283710077083, | |
| "grad_norm": 161.875, | |
| "learning_rate": 6.384892086330936e-06, | |
| "loss": 71.8582, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.4265187143755012, | |
| "grad_norm": 159.375, | |
| "learning_rate": 6.3720452209660846e-06, | |
| "loss": 71.7597, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 0.42767459165023153, | |
| "grad_norm": 168.75, | |
| "learning_rate": 6.359198355601233e-06, | |
| "loss": 72.3329, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4288304689249619, | |
| "grad_norm": 227.625, | |
| "learning_rate": 6.346351490236383e-06, | |
| "loss": 70.8971, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 0.4299863461996922, | |
| "grad_norm": 162.125, | |
| "learning_rate": 6.333504624871531e-06, | |
| "loss": 72.014, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.4311422234744226, | |
| "grad_norm": 167.375, | |
| "learning_rate": 6.320657759506681e-06, | |
| "loss": 70.4521, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 0.432298100749153, | |
| "grad_norm": 176.625, | |
| "learning_rate": 6.3078108941418305e-06, | |
| "loss": 71.6845, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.4334539780238833, | |
| "grad_norm": 160.5, | |
| "learning_rate": 6.2949640287769785e-06, | |
| "loss": 70.4146, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.4346098552986137, | |
| "grad_norm": 169.0, | |
| "learning_rate": 6.282117163412128e-06, | |
| "loss": 70.7766, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.435765732573344, | |
| "grad_norm": 166.125, | |
| "learning_rate": 6.269270298047278e-06, | |
| "loss": 70.6441, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 0.4369216098480744, | |
| "grad_norm": 169.875, | |
| "learning_rate": 6.256423432682426e-06, | |
| "loss": 70.6906, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.43807748712280475, | |
| "grad_norm": 189.0, | |
| "learning_rate": 6.2435765673175756e-06, | |
| "loss": 69.9695, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 0.4392333643975351, | |
| "grad_norm": 163.5, | |
| "learning_rate": 6.2307297019527236e-06, | |
| "loss": 72.2346, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.44038924167226545, | |
| "grad_norm": 168.25, | |
| "learning_rate": 6.217882836587873e-06, | |
| "loss": 70.2958, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 0.44154511894699583, | |
| "grad_norm": 187.875, | |
| "learning_rate": 6.205035971223022e-06, | |
| "loss": 69.8792, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.44270099622172615, | |
| "grad_norm": 149.625, | |
| "learning_rate": 6.192189105858171e-06, | |
| "loss": 72.0271, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 0.4438568734964565, | |
| "grad_norm": 160.625, | |
| "learning_rate": 6.17934224049332e-06, | |
| "loss": 71.8988, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.44501275077118685, | |
| "grad_norm": 186.875, | |
| "learning_rate": 6.1664953751284695e-06, | |
| "loss": 71.6762, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.4461686280459172, | |
| "grad_norm": 166.125, | |
| "learning_rate": 6.1536485097636175e-06, | |
| "loss": 70.4196, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.4473245053206476, | |
| "grad_norm": 169.625, | |
| "learning_rate": 6.140801644398767e-06, | |
| "loss": 70.9007, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 0.4484803825953779, | |
| "grad_norm": 166.25, | |
| "learning_rate": 6.127954779033917e-06, | |
| "loss": 69.9268, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.4496362598701083, | |
| "grad_norm": 151.625, | |
| "learning_rate": 6.115107913669065e-06, | |
| "loss": 71.6101, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 0.4507921371448386, | |
| "grad_norm": 169.125, | |
| "learning_rate": 6.102261048304215e-06, | |
| "loss": 72.2735, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.451948014419569, | |
| "grad_norm": 161.375, | |
| "learning_rate": 6.0894141829393634e-06, | |
| "loss": 71.2896, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 0.4531038916942994, | |
| "grad_norm": 174.875, | |
| "learning_rate": 6.076567317574512e-06, | |
| "loss": 71.3177, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.4542597689690297, | |
| "grad_norm": 170.5, | |
| "learning_rate": 6.063720452209661e-06, | |
| "loss": 70.5251, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 0.4554156462437601, | |
| "grad_norm": 156.875, | |
| "learning_rate": 6.05087358684481e-06, | |
| "loss": 70.9012, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.45657152351849045, | |
| "grad_norm": 156.0, | |
| "learning_rate": 6.03802672147996e-06, | |
| "loss": 70.7252, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.4577274007932208, | |
| "grad_norm": 152.75, | |
| "learning_rate": 6.0251798561151085e-06, | |
| "loss": 70.9783, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.45888327806795115, | |
| "grad_norm": 159.5, | |
| "learning_rate": 6.012332990750257e-06, | |
| "loss": 70.2681, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 0.46003915534268147, | |
| "grad_norm": 178.0, | |
| "learning_rate": 5.999486125385406e-06, | |
| "loss": 70.291, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.46119503261741185, | |
| "grad_norm": 176.875, | |
| "learning_rate": 5.986639260020556e-06, | |
| "loss": 70.9205, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 0.4623509098921422, | |
| "grad_norm": 188.75, | |
| "learning_rate": 5.973792394655704e-06, | |
| "loss": 72.2977, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4623509098921422, | |
| "eval_loss": NaN, | |
| "eval_runtime": 382.3957, | |
| "eval_samples_per_second": 609.667, | |
| "eval_steps_per_second": 38.105, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.46350678716687255, | |
| "grad_norm": 176.5, | |
| "learning_rate": 5.960945529290854e-06, | |
| "loss": 71.4161, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 0.4646626644416029, | |
| "grad_norm": 179.375, | |
| "learning_rate": 5.948098663926003e-06, | |
| "loss": 70.9138, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.46581854171633325, | |
| "grad_norm": 160.375, | |
| "learning_rate": 5.935251798561151e-06, | |
| "loss": 72.0052, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 0.4669744189910636, | |
| "grad_norm": 164.5, | |
| "learning_rate": 5.922404933196301e-06, | |
| "loss": 70.5761, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.468130296265794, | |
| "grad_norm": 164.5, | |
| "learning_rate": 5.90955806783145e-06, | |
| "loss": 69.3542, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.4692861735405243, | |
| "grad_norm": 159.5, | |
| "learning_rate": 5.896711202466599e-06, | |
| "loss": 70.1756, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.4704420508152547, | |
| "grad_norm": 205.125, | |
| "learning_rate": 5.8838643371017475e-06, | |
| "loss": 70.4389, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 0.471597928089985, | |
| "grad_norm": 176.375, | |
| "learning_rate": 5.871017471736896e-06, | |
| "loss": 70.7157, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.4727538053647154, | |
| "grad_norm": 164.25, | |
| "learning_rate": 5.858170606372045e-06, | |
| "loss": 71.9851, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 0.4739096826394458, | |
| "grad_norm": 183.875, | |
| "learning_rate": 5.845323741007195e-06, | |
| "loss": 71.0546, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.4750655599141761, | |
| "grad_norm": 211.375, | |
| "learning_rate": 5.832476875642343e-06, | |
| "loss": 71.8851, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 0.47622143718890647, | |
| "grad_norm": 182.75, | |
| "learning_rate": 5.819630010277493e-06, | |
| "loss": 71.1175, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.47737731446363685, | |
| "grad_norm": 167.25, | |
| "learning_rate": 5.806783144912642e-06, | |
| "loss": 71.7044, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 0.47853319173836717, | |
| "grad_norm": 158.375, | |
| "learning_rate": 5.79393627954779e-06, | |
| "loss": 69.464, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.47968906901309755, | |
| "grad_norm": 169.625, | |
| "learning_rate": 5.78108941418294e-06, | |
| "loss": 71.0951, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.48084494628782787, | |
| "grad_norm": 186.75, | |
| "learning_rate": 5.76824254881809e-06, | |
| "loss": 69.9219, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.48200082356255824, | |
| "grad_norm": 168.5, | |
| "learning_rate": 5.755395683453238e-06, | |
| "loss": 71.1934, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 0.4831567008372886, | |
| "grad_norm": 155.125, | |
| "learning_rate": 5.742548818088387e-06, | |
| "loss": 71.4351, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.48431257811201894, | |
| "grad_norm": 154.0, | |
| "learning_rate": 5.729701952723536e-06, | |
| "loss": 70.3735, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 0.4854684553867493, | |
| "grad_norm": 158.0, | |
| "learning_rate": 5.716855087358685e-06, | |
| "loss": 69.8907, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.48662433266147964, | |
| "grad_norm": 160.375, | |
| "learning_rate": 5.704008221993834e-06, | |
| "loss": 71.9331, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 0.48778020993621, | |
| "grad_norm": 159.875, | |
| "learning_rate": 5.691161356628983e-06, | |
| "loss": 70.6931, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.4889360872109404, | |
| "grad_norm": 208.25, | |
| "learning_rate": 5.678314491264132e-06, | |
| "loss": 70.8729, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 0.4900919644856707, | |
| "grad_norm": 169.875, | |
| "learning_rate": 5.665467625899281e-06, | |
| "loss": 71.4016, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.4912478417604011, | |
| "grad_norm": 172.875, | |
| "learning_rate": 5.652620760534429e-06, | |
| "loss": 70.1855, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.49240371903513147, | |
| "grad_norm": 161.25, | |
| "learning_rate": 5.639773895169579e-06, | |
| "loss": 69.836, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.4935595963098618, | |
| "grad_norm": 154.625, | |
| "learning_rate": 5.626927029804729e-06, | |
| "loss": 70.073, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 0.49471547358459217, | |
| "grad_norm": 157.25, | |
| "learning_rate": 5.614080164439877e-06, | |
| "loss": 69.7913, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.4958713508593225, | |
| "grad_norm": 176.25, | |
| "learning_rate": 5.601233299075026e-06, | |
| "loss": 69.8387, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 0.49702722813405287, | |
| "grad_norm": 161.125, | |
| "learning_rate": 5.588386433710175e-06, | |
| "loss": 69.4486, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.49818310540878324, | |
| "grad_norm": 180.875, | |
| "learning_rate": 5.575539568345324e-06, | |
| "loss": 69.8849, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 0.49933898268351357, | |
| "grad_norm": 179.25, | |
| "learning_rate": 5.562692702980474e-06, | |
| "loss": 70.4132, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.5004948599582439, | |
| "grad_norm": 151.375, | |
| "learning_rate": 5.549845837615623e-06, | |
| "loss": 70.3948, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 0.5016507372329743, | |
| "grad_norm": 163.0, | |
| "learning_rate": 5.5369989722507715e-06, | |
| "loss": 70.1088, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.5028066145077047, | |
| "grad_norm": 150.0, | |
| "learning_rate": 5.52415210688592e-06, | |
| "loss": 70.178, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.503962491782435, | |
| "grad_norm": 159.625, | |
| "learning_rate": 5.511305241521069e-06, | |
| "loss": 71.454, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.5051183690571653, | |
| "grad_norm": 192.25, | |
| "learning_rate": 5.498458376156218e-06, | |
| "loss": 70.7199, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 0.5062742463318957, | |
| "grad_norm": 190.375, | |
| "learning_rate": 5.485611510791368e-06, | |
| "loss": 70.7635, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.5074301236066261, | |
| "grad_norm": 202.5, | |
| "learning_rate": 5.472764645426516e-06, | |
| "loss": 70.3444, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 0.5085860008813564, | |
| "grad_norm": 169.875, | |
| "learning_rate": 5.459917780061665e-06, | |
| "loss": 70.8077, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.5097418781560867, | |
| "grad_norm": 164.125, | |
| "learning_rate": 5.447070914696815e-06, | |
| "loss": 70.1035, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 0.5108977554308172, | |
| "grad_norm": 156.75, | |
| "learning_rate": 5.434224049331963e-06, | |
| "loss": 70.4719, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.5120536327055475, | |
| "grad_norm": 185.625, | |
| "learning_rate": 5.421377183967113e-06, | |
| "loss": 71.3653, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 0.5132095099802778, | |
| "grad_norm": 161.125, | |
| "learning_rate": 5.408530318602262e-06, | |
| "loss": 69.8378, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.5143653872550082, | |
| "grad_norm": 159.5, | |
| "learning_rate": 5.3956834532374105e-06, | |
| "loss": 70.8915, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.5155212645297386, | |
| "grad_norm": 157.375, | |
| "learning_rate": 5.382836587872559e-06, | |
| "loss": 69.4553, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.5166771418044689, | |
| "grad_norm": 157.625, | |
| "learning_rate": 5.369989722507709e-06, | |
| "loss": 69.6028, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 0.5178330190791993, | |
| "grad_norm": 174.0, | |
| "learning_rate": 5.357142857142857e-06, | |
| "loss": 69.8295, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.5189888963539296, | |
| "grad_norm": 172.0, | |
| "learning_rate": 5.344295991778007e-06, | |
| "loss": 70.2746, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 0.52014477362866, | |
| "grad_norm": 156.625, | |
| "learning_rate": 5.331449126413155e-06, | |
| "loss": 70.6405, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.5213006509033903, | |
| "grad_norm": 186.0, | |
| "learning_rate": 5.3186022610483044e-06, | |
| "loss": 69.7524, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 0.5224565281781207, | |
| "grad_norm": 177.375, | |
| "learning_rate": 5.305755395683454e-06, | |
| "loss": 69.9427, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.523612405452851, | |
| "grad_norm": 162.0, | |
| "learning_rate": 5.292908530318602e-06, | |
| "loss": 69.921, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 0.5247682827275814, | |
| "grad_norm": 149.25, | |
| "learning_rate": 5.280061664953752e-06, | |
| "loss": 69.2333, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.5259241600023118, | |
| "grad_norm": 156.125, | |
| "learning_rate": 5.2672147995889015e-06, | |
| "loss": 69.6589, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.5270800372770421, | |
| "grad_norm": 165.25, | |
| "learning_rate": 5.2543679342240495e-06, | |
| "loss": 71.545, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.5282359145517724, | |
| "grad_norm": 159.75, | |
| "learning_rate": 5.241521068859199e-06, | |
| "loss": 69.826, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 0.5293917918265029, | |
| "grad_norm": 166.25, | |
| "learning_rate": 5.228674203494348e-06, | |
| "loss": 69.6276, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.5305476691012332, | |
| "grad_norm": 170.125, | |
| "learning_rate": 5.215827338129497e-06, | |
| "loss": 69.4714, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 0.5317035463759635, | |
| "grad_norm": 153.5, | |
| "learning_rate": 5.202980472764646e-06, | |
| "loss": 70.1795, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.5328594236506939, | |
| "grad_norm": 157.5, | |
| "learning_rate": 5.190133607399795e-06, | |
| "loss": 69.2672, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 0.5340153009254243, | |
| "grad_norm": 161.375, | |
| "learning_rate": 5.1772867420349434e-06, | |
| "loss": 69.4437, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.5351711782001546, | |
| "grad_norm": 167.25, | |
| "learning_rate": 5.164439876670093e-06, | |
| "loss": 70.2732, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 0.5363270554748849, | |
| "grad_norm": 158.0, | |
| "learning_rate": 5.151593011305241e-06, | |
| "loss": 70.8601, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.5374829327496153, | |
| "grad_norm": 189.625, | |
| "learning_rate": 5.138746145940391e-06, | |
| "loss": 70.938, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.5386388100243457, | |
| "grad_norm": 174.75, | |
| "learning_rate": 5.1258992805755405e-06, | |
| "loss": 69.4389, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.539794687299076, | |
| "grad_norm": 193.875, | |
| "learning_rate": 5.1130524152106885e-06, | |
| "loss": 70.4369, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 0.5409505645738064, | |
| "grad_norm": 164.0, | |
| "learning_rate": 5.100205549845838e-06, | |
| "loss": 69.9866, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.5421064418485367, | |
| "grad_norm": 148.125, | |
| "learning_rate": 5.087358684480987e-06, | |
| "loss": 69.0706, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 0.5432623191232671, | |
| "grad_norm": 167.5, | |
| "learning_rate": 5.074511819116136e-06, | |
| "loss": 70.4756, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.5444181963979975, | |
| "grad_norm": 161.75, | |
| "learning_rate": 5.061664953751286e-06, | |
| "loss": 69.9491, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 0.5455740736727278, | |
| "grad_norm": 174.0, | |
| "learning_rate": 5.0488180883864345e-06, | |
| "loss": 69.262, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.5467299509474581, | |
| "grad_norm": 172.0, | |
| "learning_rate": 5.035971223021583e-06, | |
| "loss": 70.5688, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 0.5478858282221886, | |
| "grad_norm": 181.5, | |
| "learning_rate": 5.023124357656732e-06, | |
| "loss": 69.6968, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.5490417054969189, | |
| "grad_norm": 177.25, | |
| "learning_rate": 5.010277492291881e-06, | |
| "loss": 71.718, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.5501975827716492, | |
| "grad_norm": 166.625, | |
| "learning_rate": 4.99743062692703e-06, | |
| "loss": 70.3106, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.5513534600463795, | |
| "grad_norm": 152.25, | |
| "learning_rate": 4.9845837615621795e-06, | |
| "loss": 69.8569, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 0.55250933732111, | |
| "grad_norm": 170.625, | |
| "learning_rate": 4.971736896197328e-06, | |
| "loss": 69.7388, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.5536652145958403, | |
| "grad_norm": 151.5, | |
| "learning_rate": 4.958890030832477e-06, | |
| "loss": 69.3465, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 0.5548210918705706, | |
| "grad_norm": 162.0, | |
| "learning_rate": 4.946043165467626e-06, | |
| "loss": 69.7658, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.555976969145301, | |
| "grad_norm": 170.25, | |
| "learning_rate": 4.933196300102776e-06, | |
| "loss": 68.5618, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 0.5571328464200314, | |
| "grad_norm": 168.0, | |
| "learning_rate": 4.920349434737925e-06, | |
| "loss": 69.0478, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.5582887236947617, | |
| "grad_norm": 163.75, | |
| "learning_rate": 4.9075025693730735e-06, | |
| "loss": 70.553, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 0.5594446009694921, | |
| "grad_norm": 187.25, | |
| "learning_rate": 4.894655704008222e-06, | |
| "loss": 70.2036, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.5606004782442224, | |
| "grad_norm": 169.125, | |
| "learning_rate": 4.881808838643371e-06, | |
| "loss": 69.2802, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.5617563555189528, | |
| "grad_norm": 156.75, | |
| "learning_rate": 4.86896197327852e-06, | |
| "loss": 69.5526, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.5629122327936831, | |
| "grad_norm": 161.75, | |
| "learning_rate": 4.856115107913669e-06, | |
| "loss": 70.8698, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 0.5640681100684135, | |
| "grad_norm": 162.75, | |
| "learning_rate": 4.8432682425488185e-06, | |
| "loss": 68.8072, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.5652239873431438, | |
| "grad_norm": 179.25, | |
| "learning_rate": 4.830421377183967e-06, | |
| "loss": 70.1462, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 0.5663798646178742, | |
| "grad_norm": 162.625, | |
| "learning_rate": 4.817574511819116e-06, | |
| "loss": 70.458, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.5675357418926046, | |
| "grad_norm": 151.125, | |
| "learning_rate": 4.804727646454266e-06, | |
| "loss": 70.8089, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 0.5686916191673349, | |
| "grad_norm": 166.0, | |
| "learning_rate": 4.791880781089415e-06, | |
| "loss": 69.5729, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.5698474964420652, | |
| "grad_norm": 153.125, | |
| "learning_rate": 4.779033915724564e-06, | |
| "loss": 69.4479, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 0.5710033737167957, | |
| "grad_norm": 153.0, | |
| "learning_rate": 4.7661870503597125e-06, | |
| "loss": 67.9093, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.572159250991526, | |
| "grad_norm": 155.125, | |
| "learning_rate": 4.753340184994862e-06, | |
| "loss": 70.8675, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.5733151282662563, | |
| "grad_norm": 173.375, | |
| "learning_rate": 4.740493319630011e-06, | |
| "loss": 69.6522, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.5744710055409867, | |
| "grad_norm": 167.125, | |
| "learning_rate": 4.72764645426516e-06, | |
| "loss": 69.2384, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 0.5756268828157171, | |
| "grad_norm": 163.25, | |
| "learning_rate": 4.714799588900309e-06, | |
| "loss": 69.2969, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.5767827600904474, | |
| "grad_norm": 154.875, | |
| "learning_rate": 4.7019527235354576e-06, | |
| "loss": 69.5717, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 0.5779386373651777, | |
| "grad_norm": 162.375, | |
| "learning_rate": 4.689105858170606e-06, | |
| "loss": 69.8666, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5779386373651777, | |
| "eval_loss": NaN, | |
| "eval_runtime": 381.701, | |
| "eval_samples_per_second": 610.776, | |
| "eval_steps_per_second": 38.174, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5790945146399081, | |
| "grad_norm": 172.125, | |
| "learning_rate": 4.676258992805755e-06, | |
| "loss": 69.669, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 0.5802503919146385, | |
| "grad_norm": 157.25, | |
| "learning_rate": 4.663412127440905e-06, | |
| "loss": 70.5345, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.5814062691893688, | |
| "grad_norm": 152.375, | |
| "learning_rate": 4.650565262076054e-06, | |
| "loss": 68.8704, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 0.5825621464640992, | |
| "grad_norm": 182.375, | |
| "learning_rate": 4.637718396711203e-06, | |
| "loss": 68.8443, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.5837180237388295, | |
| "grad_norm": 174.125, | |
| "learning_rate": 4.624871531346352e-06, | |
| "loss": 69.324, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.5848739010135598, | |
| "grad_norm": 170.375, | |
| "learning_rate": 4.612024665981501e-06, | |
| "loss": 69.2429, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.5860297782882903, | |
| "grad_norm": 162.875, | |
| "learning_rate": 4.59917780061665e-06, | |
| "loss": 68.8715, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 0.5871856555630206, | |
| "grad_norm": 154.875, | |
| "learning_rate": 4.586330935251799e-06, | |
| "loss": 70.5938, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.5883415328377509, | |
| "grad_norm": 164.125, | |
| "learning_rate": 4.5734840698869486e-06, | |
| "loss": 69.1381, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 0.5894974101124814, | |
| "grad_norm": 168.5, | |
| "learning_rate": 4.560637204522097e-06, | |
| "loss": 68.9723, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.5906532873872117, | |
| "grad_norm": 157.125, | |
| "learning_rate": 4.547790339157246e-06, | |
| "loss": 68.8563, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 0.591809164661942, | |
| "grad_norm": 170.125, | |
| "learning_rate": 4.534943473792395e-06, | |
| "loss": 69.5333, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.5929650419366723, | |
| "grad_norm": 168.75, | |
| "learning_rate": 4.522096608427544e-06, | |
| "loss": 68.3743, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 0.5941209192114028, | |
| "grad_norm": 161.5, | |
| "learning_rate": 4.509249743062693e-06, | |
| "loss": 68.4732, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.5952767964861331, | |
| "grad_norm": 177.375, | |
| "learning_rate": 4.496402877697842e-06, | |
| "loss": 69.9702, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.5964326737608634, | |
| "grad_norm": 160.5, | |
| "learning_rate": 4.483556012332991e-06, | |
| "loss": 69.5753, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.5975885510355938, | |
| "grad_norm": 170.625, | |
| "learning_rate": 4.47070914696814e-06, | |
| "loss": 69.2494, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 0.5987444283103242, | |
| "grad_norm": 160.5, | |
| "learning_rate": 4.457862281603289e-06, | |
| "loss": 69.5917, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.5999003055850545, | |
| "grad_norm": 157.875, | |
| "learning_rate": 4.445015416238438e-06, | |
| "loss": 69.0619, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 0.6010561828597849, | |
| "grad_norm": 150.875, | |
| "learning_rate": 4.432168550873588e-06, | |
| "loss": 69.3003, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.6022120601345152, | |
| "grad_norm": 163.125, | |
| "learning_rate": 4.4193216855087364e-06, | |
| "loss": 68.8558, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 0.6033679374092455, | |
| "grad_norm": 168.375, | |
| "learning_rate": 4.406474820143885e-06, | |
| "loss": 70.8878, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.604523814683976, | |
| "grad_norm": 157.25, | |
| "learning_rate": 4.393627954779034e-06, | |
| "loss": 69.8354, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 0.6056796919587063, | |
| "grad_norm": 154.625, | |
| "learning_rate": 4.380781089414183e-06, | |
| "loss": 69.2094, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.6068355692334366, | |
| "grad_norm": 178.875, | |
| "learning_rate": 4.367934224049332e-06, | |
| "loss": 69.4816, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.6079914465081669, | |
| "grad_norm": 148.5, | |
| "learning_rate": 4.355087358684481e-06, | |
| "loss": 69.5445, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.6091473237828974, | |
| "grad_norm": 189.375, | |
| "learning_rate": 4.34224049331963e-06, | |
| "loss": 69.7445, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 0.6103032010576277, | |
| "grad_norm": 163.375, | |
| "learning_rate": 4.329393627954779e-06, | |
| "loss": 70.1707, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.611459078332358, | |
| "grad_norm": 159.625, | |
| "learning_rate": 4.316546762589928e-06, | |
| "loss": 70.575, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 0.6126149556070885, | |
| "grad_norm": 163.875, | |
| "learning_rate": 4.303699897225078e-06, | |
| "loss": 68.2103, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.6137708328818188, | |
| "grad_norm": 173.25, | |
| "learning_rate": 4.290853031860227e-06, | |
| "loss": 68.2952, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 0.6149267101565491, | |
| "grad_norm": 173.75, | |
| "learning_rate": 4.2780061664953754e-06, | |
| "loss": 68.7904, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.6160825874312795, | |
| "grad_norm": 163.875, | |
| "learning_rate": 4.265159301130524e-06, | |
| "loss": 70.0719, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 0.6172384647060098, | |
| "grad_norm": 180.625, | |
| "learning_rate": 4.252312435765674e-06, | |
| "loss": 68.3693, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.6183943419807402, | |
| "grad_norm": 154.5, | |
| "learning_rate": 4.239465570400823e-06, | |
| "loss": 68.8033, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.6195502192554706, | |
| "grad_norm": 160.25, | |
| "learning_rate": 4.226618705035972e-06, | |
| "loss": 67.7923, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.6207060965302009, | |
| "grad_norm": 159.5, | |
| "learning_rate": 4.2137718396711205e-06, | |
| "loss": 69.2746, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 0.6218619738049312, | |
| "grad_norm": 164.0, | |
| "learning_rate": 4.200924974306269e-06, | |
| "loss": 68.537, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.6230178510796616, | |
| "grad_norm": 156.25, | |
| "learning_rate": 4.188078108941418e-06, | |
| "loss": 68.6553, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 0.624173728354392, | |
| "grad_norm": 158.625, | |
| "learning_rate": 4.175231243576567e-06, | |
| "loss": 68.8911, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.6253296056291223, | |
| "grad_norm": 166.25, | |
| "learning_rate": 4.162384378211717e-06, | |
| "loss": 69.7975, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 0.6264854829038526, | |
| "grad_norm": 160.25, | |
| "learning_rate": 4.149537512846866e-06, | |
| "loss": 69.9966, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.6276413601785831, | |
| "grad_norm": 167.0, | |
| "learning_rate": 4.1366906474820145e-06, | |
| "loss": 70.2918, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 0.6287972374533134, | |
| "grad_norm": 156.5, | |
| "learning_rate": 4.123843782117164e-06, | |
| "loss": 67.3517, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.6299531147280437, | |
| "grad_norm": 171.25, | |
| "learning_rate": 4.110996916752313e-06, | |
| "loss": 68.5519, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.6311089920027741, | |
| "grad_norm": 178.75, | |
| "learning_rate": 4.098150051387462e-06, | |
| "loss": 69.3377, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.6322648692775045, | |
| "grad_norm": 167.875, | |
| "learning_rate": 4.085303186022611e-06, | |
| "loss": 68.7498, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 0.6334207465522348, | |
| "grad_norm": 164.875, | |
| "learning_rate": 4.07245632065776e-06, | |
| "loss": 68.4089, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.6345766238269652, | |
| "grad_norm": 171.5, | |
| "learning_rate": 4.059609455292909e-06, | |
| "loss": 68.4604, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 0.6357325011016955, | |
| "grad_norm": 169.5, | |
| "learning_rate": 4.046762589928058e-06, | |
| "loss": 69.7823, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.6368883783764259, | |
| "grad_norm": 167.125, | |
| "learning_rate": 4.033915724563207e-06, | |
| "loss": 68.461, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 0.6380442556511562, | |
| "grad_norm": 157.75, | |
| "learning_rate": 4.021068859198356e-06, | |
| "loss": 69.8009, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.6392001329258866, | |
| "grad_norm": 153.25, | |
| "learning_rate": 4.008221993833505e-06, | |
| "loss": 69.0233, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 0.6403560102006169, | |
| "grad_norm": 153.375, | |
| "learning_rate": 3.9953751284686535e-06, | |
| "loss": 69.8202, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.6415118874753473, | |
| "grad_norm": 152.25, | |
| "learning_rate": 3.982528263103803e-06, | |
| "loss": 70.0482, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.6426677647500777, | |
| "grad_norm": 151.0, | |
| "learning_rate": 3.969681397738952e-06, | |
| "loss": 68.4574, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.643823642024808, | |
| "grad_norm": 153.75, | |
| "learning_rate": 3.956834532374101e-06, | |
| "loss": 68.9972, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 0.6449795192995383, | |
| "grad_norm": 155.0, | |
| "learning_rate": 3.9439876670092506e-06, | |
| "loss": 69.0547, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.6461353965742688, | |
| "grad_norm": 165.75, | |
| "learning_rate": 3.931140801644399e-06, | |
| "loss": 69.5258, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 0.6472912738489991, | |
| "grad_norm": 148.625, | |
| "learning_rate": 3.918293936279548e-06, | |
| "loss": 68.9722, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.6484471511237294, | |
| "grad_norm": 146.125, | |
| "learning_rate": 3.905447070914697e-06, | |
| "loss": 69.1489, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 0.6496030283984598, | |
| "grad_norm": 167.625, | |
| "learning_rate": 3.892600205549846e-06, | |
| "loss": 67.9903, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.6507589056731902, | |
| "grad_norm": 154.0, | |
| "learning_rate": 3.879753340184995e-06, | |
| "loss": 69.16, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 0.6519147829479205, | |
| "grad_norm": 157.25, | |
| "learning_rate": 3.866906474820144e-06, | |
| "loss": 70.0419, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.6530706602226508, | |
| "grad_norm": 156.625, | |
| "learning_rate": 3.854059609455293e-06, | |
| "loss": 69.2703, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.6542265374973812, | |
| "grad_norm": 153.875, | |
| "learning_rate": 3.841212744090442e-06, | |
| "loss": 69.3849, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.6553824147721116, | |
| "grad_norm": 161.125, | |
| "learning_rate": 3.828365878725591e-06, | |
| "loss": 69.6766, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 0.6565382920468419, | |
| "grad_norm": 147.875, | |
| "learning_rate": 3.81551901336074e-06, | |
| "loss": 67.732, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.6576941693215723, | |
| "grad_norm": 168.0, | |
| "learning_rate": 3.8026721479958896e-06, | |
| "loss": 69.0334, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 0.6588500465963026, | |
| "grad_norm": 153.75, | |
| "learning_rate": 3.7898252826310384e-06, | |
| "loss": 68.3081, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.660005923871033, | |
| "grad_norm": 184.25, | |
| "learning_rate": 3.7769784172661873e-06, | |
| "loss": 69.0602, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 0.6611618011457634, | |
| "grad_norm": 189.75, | |
| "learning_rate": 3.7641315519013365e-06, | |
| "loss": 67.2991, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.6623176784204937, | |
| "grad_norm": 169.0, | |
| "learning_rate": 3.7512846865364854e-06, | |
| "loss": 70.8647, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 0.663473555695224, | |
| "grad_norm": 157.75, | |
| "learning_rate": 3.7384378211716342e-06, | |
| "loss": 68.4063, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.6646294329699545, | |
| "grad_norm": 176.375, | |
| "learning_rate": 3.725590955806783e-06, | |
| "loss": 68.3025, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.6657853102446848, | |
| "grad_norm": 180.0, | |
| "learning_rate": 3.7127440904419328e-06, | |
| "loss": 68.6778, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.6669411875194151, | |
| "grad_norm": 168.5, | |
| "learning_rate": 3.6998972250770816e-06, | |
| "loss": 67.2759, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 0.6680970647941454, | |
| "grad_norm": 164.25, | |
| "learning_rate": 3.6870503597122305e-06, | |
| "loss": 69.538, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.6692529420688759, | |
| "grad_norm": 157.625, | |
| "learning_rate": 3.6742034943473797e-06, | |
| "loss": 68.8391, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 0.6704088193436062, | |
| "grad_norm": 171.625, | |
| "learning_rate": 3.6613566289825286e-06, | |
| "loss": 68.3289, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.6715646966183365, | |
| "grad_norm": 161.375, | |
| "learning_rate": 3.6485097636176774e-06, | |
| "loss": 67.6343, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 0.6727205738930669, | |
| "grad_norm": 159.0, | |
| "learning_rate": 3.6356628982528263e-06, | |
| "loss": 67.9106, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.6738764511677973, | |
| "grad_norm": 148.625, | |
| "learning_rate": 3.622816032887976e-06, | |
| "loss": 68.7488, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 0.6750323284425276, | |
| "grad_norm": 172.0, | |
| "learning_rate": 3.609969167523125e-06, | |
| "loss": 68.8937, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.676188205717258, | |
| "grad_norm": 169.875, | |
| "learning_rate": 3.5971223021582737e-06, | |
| "loss": 68.0938, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.6773440829919883, | |
| "grad_norm": 161.25, | |
| "learning_rate": 3.584275436793423e-06, | |
| "loss": 68.6723, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.6784999602667187, | |
| "grad_norm": 159.0, | |
| "learning_rate": 3.5714285714285718e-06, | |
| "loss": 68.2037, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 0.679655837541449, | |
| "grad_norm": 165.875, | |
| "learning_rate": 3.5585817060637206e-06, | |
| "loss": 68.3861, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.6808117148161794, | |
| "grad_norm": 160.625, | |
| "learning_rate": 3.5457348406988695e-06, | |
| "loss": 67.6351, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 0.6819675920909097, | |
| "grad_norm": 172.625, | |
| "learning_rate": 3.5328879753340187e-06, | |
| "loss": 68.8567, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.6831234693656401, | |
| "grad_norm": 185.125, | |
| "learning_rate": 3.520041109969168e-06, | |
| "loss": 66.5128, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 0.6842793466403705, | |
| "grad_norm": 157.75, | |
| "learning_rate": 3.507194244604317e-06, | |
| "loss": 68.0714, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.6854352239151008, | |
| "grad_norm": 151.375, | |
| "learning_rate": 3.494347379239466e-06, | |
| "loss": 67.9339, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 0.6865911011898311, | |
| "grad_norm": 164.25, | |
| "learning_rate": 3.481500513874615e-06, | |
| "loss": 68.5463, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.6877469784645616, | |
| "grad_norm": 138.0, | |
| "learning_rate": 3.468653648509764e-06, | |
| "loss": 69.1578, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.6889028557392919, | |
| "grad_norm": 200.75, | |
| "learning_rate": 3.4558067831449127e-06, | |
| "loss": 68.674, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.6900587330140222, | |
| "grad_norm": 149.75, | |
| "learning_rate": 3.442959917780062e-06, | |
| "loss": 68.8552, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 0.6912146102887526, | |
| "grad_norm": 165.125, | |
| "learning_rate": 3.430113052415211e-06, | |
| "loss": 68.541, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.692370487563483, | |
| "grad_norm": 175.625, | |
| "learning_rate": 3.4172661870503596e-06, | |
| "loss": 68.8684, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 0.6935263648382133, | |
| "grad_norm": 165.5, | |
| "learning_rate": 3.4044193216855093e-06, | |
| "loss": 68.3212, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6935263648382133, | |
| "eval_loss": NaN, | |
| "eval_runtime": 380.967, | |
| "eval_samples_per_second": 611.953, | |
| "eval_steps_per_second": 38.247, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6946822421129436, | |
| "grad_norm": 171.0, | |
| "learning_rate": 3.391572456320658e-06, | |
| "loss": 69.3591, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 0.695838119387674, | |
| "grad_norm": 178.875, | |
| "learning_rate": 3.378725590955807e-06, | |
| "loss": 69.3434, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.6969939966624044, | |
| "grad_norm": 188.375, | |
| "learning_rate": 3.365878725590956e-06, | |
| "loss": 68.8737, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 0.6981498739371347, | |
| "grad_norm": 169.875, | |
| "learning_rate": 3.353031860226105e-06, | |
| "loss": 68.2949, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.6993057512118651, | |
| "grad_norm": 154.25, | |
| "learning_rate": 3.340184994861254e-06, | |
| "loss": 67.784, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.7004616284865954, | |
| "grad_norm": 159.25, | |
| "learning_rate": 3.327338129496403e-06, | |
| "loss": 68.2709, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.7016175057613258, | |
| "grad_norm": 148.75, | |
| "learning_rate": 3.3144912641315525e-06, | |
| "loss": 69.2274, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 0.7027733830360562, | |
| "grad_norm": 157.375, | |
| "learning_rate": 3.3016443987667014e-06, | |
| "loss": 67.7498, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.7039292603107865, | |
| "grad_norm": 153.0, | |
| "learning_rate": 3.2887975334018502e-06, | |
| "loss": 67.9523, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 0.7050851375855168, | |
| "grad_norm": 162.375, | |
| "learning_rate": 3.275950668036999e-06, | |
| "loss": 67.1861, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.7062410148602473, | |
| "grad_norm": 154.5, | |
| "learning_rate": 3.2631038026721483e-06, | |
| "loss": 68.3158, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 0.7073968921349776, | |
| "grad_norm": 145.875, | |
| "learning_rate": 3.250256937307297e-06, | |
| "loss": 68.1404, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.7085527694097079, | |
| "grad_norm": 163.375, | |
| "learning_rate": 3.237410071942446e-06, | |
| "loss": 68.4577, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 0.7097086466844382, | |
| "grad_norm": 167.0, | |
| "learning_rate": 3.2245632065775957e-06, | |
| "loss": 67.7858, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.7108645239591687, | |
| "grad_norm": 148.25, | |
| "learning_rate": 3.2117163412127446e-06, | |
| "loss": 67.727, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.712020401233899, | |
| "grad_norm": 151.375, | |
| "learning_rate": 3.1988694758478934e-06, | |
| "loss": 68.0403, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.7131762785086293, | |
| "grad_norm": 158.25, | |
| "learning_rate": 3.1860226104830423e-06, | |
| "loss": 68.9341, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 0.7143321557833597, | |
| "grad_norm": 159.875, | |
| "learning_rate": 3.1731757451181915e-06, | |
| "loss": 68.1696, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.7154880330580901, | |
| "grad_norm": 172.0, | |
| "learning_rate": 3.1603288797533404e-06, | |
| "loss": 67.7838, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 0.7166439103328204, | |
| "grad_norm": 166.875, | |
| "learning_rate": 3.1474820143884892e-06, | |
| "loss": 68.0554, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.7177997876075508, | |
| "grad_norm": 150.375, | |
| "learning_rate": 3.134635149023639e-06, | |
| "loss": 68.3312, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 0.7189556648822811, | |
| "grad_norm": 154.375, | |
| "learning_rate": 3.1217882836587878e-06, | |
| "loss": 68.1626, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.7201115421570115, | |
| "grad_norm": 175.125, | |
| "learning_rate": 3.1089414182939366e-06, | |
| "loss": 68.838, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 0.7212674194317419, | |
| "grad_norm": 162.125, | |
| "learning_rate": 3.0960945529290855e-06, | |
| "loss": 68.3526, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.7224232967064722, | |
| "grad_norm": 163.625, | |
| "learning_rate": 3.0832476875642348e-06, | |
| "loss": 67.3081, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.7235791739812025, | |
| "grad_norm": 150.5, | |
| "learning_rate": 3.0704008221993836e-06, | |
| "loss": 67.0766, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.7247350512559328, | |
| "grad_norm": 159.25, | |
| "learning_rate": 3.0575539568345324e-06, | |
| "loss": 68.1058, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 0.7258909285306633, | |
| "grad_norm": 156.625, | |
| "learning_rate": 3.0447070914696817e-06, | |
| "loss": 68.7516, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.7270468058053936, | |
| "grad_norm": 154.875, | |
| "learning_rate": 3.0318602261048306e-06, | |
| "loss": 69.4498, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 0.7282026830801239, | |
| "grad_norm": 146.5, | |
| "learning_rate": 3.01901336073998e-06, | |
| "loss": 68.6958, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.7293585603548544, | |
| "grad_norm": 158.75, | |
| "learning_rate": 3.0061664953751287e-06, | |
| "loss": 67.5718, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 0.7305144376295847, | |
| "grad_norm": 159.0, | |
| "learning_rate": 2.993319630010278e-06, | |
| "loss": 67.7357, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.731670314904315, | |
| "grad_norm": 164.625, | |
| "learning_rate": 2.980472764645427e-06, | |
| "loss": 68.8242, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 0.7328261921790454, | |
| "grad_norm": 162.875, | |
| "learning_rate": 2.9676258992805756e-06, | |
| "loss": 67.6763, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.7339820694537758, | |
| "grad_norm": 152.625, | |
| "learning_rate": 2.954779033915725e-06, | |
| "loss": 67.596, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.7351379467285061, | |
| "grad_norm": 160.125, | |
| "learning_rate": 2.9419321685508738e-06, | |
| "loss": 67.8334, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.7362938240032365, | |
| "grad_norm": 176.875, | |
| "learning_rate": 2.9290853031860226e-06, | |
| "loss": 68.2346, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 0.7374497012779668, | |
| "grad_norm": 158.625, | |
| "learning_rate": 2.9162384378211715e-06, | |
| "loss": 67.566, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.7386055785526972, | |
| "grad_norm": 171.625, | |
| "learning_rate": 2.903391572456321e-06, | |
| "loss": 67.5561, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 0.7397614558274275, | |
| "grad_norm": 172.375, | |
| "learning_rate": 2.89054470709147e-06, | |
| "loss": 66.9553, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.7409173331021579, | |
| "grad_norm": 158.875, | |
| "learning_rate": 2.877697841726619e-06, | |
| "loss": 67.6197, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 0.7420732103768882, | |
| "grad_norm": 164.625, | |
| "learning_rate": 2.864850976361768e-06, | |
| "loss": 67.4917, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.7432290876516185, | |
| "grad_norm": 171.5, | |
| "learning_rate": 2.852004110996917e-06, | |
| "loss": 68.1053, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 0.744384964926349, | |
| "grad_norm": 169.5, | |
| "learning_rate": 2.839157245632066e-06, | |
| "loss": 67.5385, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.7455408422010793, | |
| "grad_norm": 163.0, | |
| "learning_rate": 2.8263103802672147e-06, | |
| "loss": 67.5871, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.7466967194758096, | |
| "grad_norm": 155.0, | |
| "learning_rate": 2.8134635149023644e-06, | |
| "loss": 67.678, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.74785259675054, | |
| "grad_norm": 171.5, | |
| "learning_rate": 2.800616649537513e-06, | |
| "loss": 68.6263, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 0.7490084740252704, | |
| "grad_norm": 161.625, | |
| "learning_rate": 2.787769784172662e-06, | |
| "loss": 66.3458, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.7501643513000007, | |
| "grad_norm": 154.0, | |
| "learning_rate": 2.7749229188078113e-06, | |
| "loss": 68.3403, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 0.7513202285747311, | |
| "grad_norm": 151.25, | |
| "learning_rate": 2.76207605344296e-06, | |
| "loss": 67.5946, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.7524761058494615, | |
| "grad_norm": 159.875, | |
| "learning_rate": 2.749229188078109e-06, | |
| "loss": 67.0883, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 0.7536319831241918, | |
| "grad_norm": 153.0, | |
| "learning_rate": 2.736382322713258e-06, | |
| "loss": 68.4621, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.7547878603989221, | |
| "grad_norm": 160.625, | |
| "learning_rate": 2.7235354573484076e-06, | |
| "loss": 67.5966, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 0.7559437376736525, | |
| "grad_norm": 160.0, | |
| "learning_rate": 2.7106885919835564e-06, | |
| "loss": 69.2813, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.7570996149483828, | |
| "grad_norm": 156.75, | |
| "learning_rate": 2.6978417266187052e-06, | |
| "loss": 67.1882, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.7582554922231132, | |
| "grad_norm": 162.75, | |
| "learning_rate": 2.6849948612538545e-06, | |
| "loss": 67.7005, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.7594113694978436, | |
| "grad_norm": 180.875, | |
| "learning_rate": 2.6721479958890034e-06, | |
| "loss": 68.8234, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 0.7605672467725739, | |
| "grad_norm": 184.5, | |
| "learning_rate": 2.6593011305241522e-06, | |
| "loss": 68.6867, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.7617231240473042, | |
| "grad_norm": 157.5, | |
| "learning_rate": 2.646454265159301e-06, | |
| "loss": 67.4642, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 0.7628790013220347, | |
| "grad_norm": 156.125, | |
| "learning_rate": 2.6336073997944508e-06, | |
| "loss": 67.4047, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.764034878596765, | |
| "grad_norm": 149.0, | |
| "learning_rate": 2.6207605344295996e-06, | |
| "loss": 67.1334, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 0.7651907558714953, | |
| "grad_norm": 157.5, | |
| "learning_rate": 2.6079136690647484e-06, | |
| "loss": 66.8244, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.7663466331462258, | |
| "grad_norm": 160.125, | |
| "learning_rate": 2.5950668036998973e-06, | |
| "loss": 68.4834, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 0.7675025104209561, | |
| "grad_norm": 163.125, | |
| "learning_rate": 2.5822199383350466e-06, | |
| "loss": 67.9042, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.7686583876956864, | |
| "grad_norm": 164.25, | |
| "learning_rate": 2.5693730729701954e-06, | |
| "loss": 67.7928, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.7698142649704167, | |
| "grad_norm": 161.875, | |
| "learning_rate": 2.5565262076053443e-06, | |
| "loss": 69.5935, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.7709701422451471, | |
| "grad_norm": 142.125, | |
| "learning_rate": 2.5436793422404935e-06, | |
| "loss": 66.6378, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 0.7721260195198775, | |
| "grad_norm": 156.125, | |
| "learning_rate": 2.530832476875643e-06, | |
| "loss": 67.4647, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.7732818967946078, | |
| "grad_norm": 155.875, | |
| "learning_rate": 2.5179856115107916e-06, | |
| "loss": 68.0196, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 0.7744377740693382, | |
| "grad_norm": 163.125, | |
| "learning_rate": 2.5051387461459405e-06, | |
| "loss": 66.556, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.7755936513440685, | |
| "grad_norm": 173.375, | |
| "learning_rate": 2.4922918807810898e-06, | |
| "loss": 68.6675, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 0.7767495286187989, | |
| "grad_norm": 158.25, | |
| "learning_rate": 2.4794450154162386e-06, | |
| "loss": 66.8891, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.7779054058935293, | |
| "grad_norm": 152.25, | |
| "learning_rate": 2.466598150051388e-06, | |
| "loss": 67.2771, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 0.7790612831682596, | |
| "grad_norm": 160.125, | |
| "learning_rate": 2.4537512846865367e-06, | |
| "loss": 66.9648, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.7802171604429899, | |
| "grad_norm": 164.75, | |
| "learning_rate": 2.4409044193216856e-06, | |
| "loss": 68.0911, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.7813730377177204, | |
| "grad_norm": 162.125, | |
| "learning_rate": 2.4280575539568344e-06, | |
| "loss": 67.2639, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.7825289149924507, | |
| "grad_norm": 161.375, | |
| "learning_rate": 2.4152106885919837e-06, | |
| "loss": 66.6423, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 0.783684792267181, | |
| "grad_norm": 149.5, | |
| "learning_rate": 2.402363823227133e-06, | |
| "loss": 67.8673, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.7848406695419113, | |
| "grad_norm": 151.5, | |
| "learning_rate": 2.389516957862282e-06, | |
| "loss": 66.1444, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 0.7859965468166418, | |
| "grad_norm": 156.625, | |
| "learning_rate": 2.376670092497431e-06, | |
| "loss": 67.8318, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.7871524240913721, | |
| "grad_norm": 152.75, | |
| "learning_rate": 2.36382322713258e-06, | |
| "loss": 68.5987, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 0.7883083013661024, | |
| "grad_norm": 163.25, | |
| "learning_rate": 2.3509763617677288e-06, | |
| "loss": 67.008, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.7894641786408328, | |
| "grad_norm": 167.375, | |
| "learning_rate": 2.3381294964028776e-06, | |
| "loss": 67.2252, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 0.7906200559155632, | |
| "grad_norm": 157.25, | |
| "learning_rate": 2.325282631038027e-06, | |
| "loss": 66.6725, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.7917759331902935, | |
| "grad_norm": 155.875, | |
| "learning_rate": 2.312435765673176e-06, | |
| "loss": 67.9088, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.7929318104650239, | |
| "grad_norm": 157.375, | |
| "learning_rate": 2.299588900308325e-06, | |
| "loss": 66.1468, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.7940876877397542, | |
| "grad_norm": 152.875, | |
| "learning_rate": 2.2867420349434743e-06, | |
| "loss": 66.662, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 0.7952435650144846, | |
| "grad_norm": 164.0, | |
| "learning_rate": 2.273895169578623e-06, | |
| "loss": 66.9803, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.7963994422892149, | |
| "grad_norm": 159.5, | |
| "learning_rate": 2.261048304213772e-06, | |
| "loss": 67.1713, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 0.7975553195639453, | |
| "grad_norm": 162.0, | |
| "learning_rate": 2.248201438848921e-06, | |
| "loss": 67.7926, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.7987111968386756, | |
| "grad_norm": 163.875, | |
| "learning_rate": 2.23535457348407e-06, | |
| "loss": 67.5221, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 0.799867074113406, | |
| "grad_norm": 161.75, | |
| "learning_rate": 2.222507708119219e-06, | |
| "loss": 69.4966, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.8010229513881364, | |
| "grad_norm": 209.375, | |
| "learning_rate": 2.2096608427543682e-06, | |
| "loss": 66.1803, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 0.8021788286628667, | |
| "grad_norm": 168.0, | |
| "learning_rate": 2.196813977389517e-06, | |
| "loss": 66.1049, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.803334705937597, | |
| "grad_norm": 157.75, | |
| "learning_rate": 2.183967112024666e-06, | |
| "loss": 66.1451, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 0.8044905832123275, | |
| "grad_norm": 150.75, | |
| "learning_rate": 2.171120246659815e-06, | |
| "loss": 66.0184, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.8056464604870578, | |
| "grad_norm": 151.375, | |
| "learning_rate": 2.158273381294964e-06, | |
| "loss": 67.9845, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 0.8068023377617881, | |
| "grad_norm": 160.25, | |
| "learning_rate": 2.1454265159301133e-06, | |
| "loss": 66.5661, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.8079582150365185, | |
| "grad_norm": 153.875, | |
| "learning_rate": 2.132579650565262e-06, | |
| "loss": 67.4468, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 0.8091140923112489, | |
| "grad_norm": 150.25, | |
| "learning_rate": 2.1197327852004114e-06, | |
| "loss": 66.7717, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8091140923112489, | |
| "eval_loss": NaN, | |
| "eval_runtime": 383.0157, | |
| "eval_samples_per_second": 608.68, | |
| "eval_steps_per_second": 38.043, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8102699695859792, | |
| "grad_norm": 164.375, | |
| "learning_rate": 2.1068859198355603e-06, | |
| "loss": 67.4157, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 0.8114258468607095, | |
| "grad_norm": 159.5, | |
| "learning_rate": 2.094039054470709e-06, | |
| "loss": 67.4029, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.8125817241354399, | |
| "grad_norm": 143.25, | |
| "learning_rate": 2.0811921891058584e-06, | |
| "loss": 67.8529, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 0.8137376014101703, | |
| "grad_norm": 156.25, | |
| "learning_rate": 2.0683453237410072e-06, | |
| "loss": 66.8855, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.8148934786849006, | |
| "grad_norm": 168.75, | |
| "learning_rate": 2.0554984583761565e-06, | |
| "loss": 68.2299, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 0.816049355959631, | |
| "grad_norm": 149.0, | |
| "learning_rate": 2.0426515930113053e-06, | |
| "loss": 66.2473, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.8172052332343613, | |
| "grad_norm": 149.0, | |
| "learning_rate": 2.0298047276464546e-06, | |
| "loss": 66.8619, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 0.8183611105090917, | |
| "grad_norm": 171.375, | |
| "learning_rate": 2.0169578622816035e-06, | |
| "loss": 67.1076, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.8195169877838221, | |
| "grad_norm": 153.875, | |
| "learning_rate": 2.0041109969167523e-06, | |
| "loss": 68.1818, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 0.8206728650585524, | |
| "grad_norm": 169.25, | |
| "learning_rate": 1.9912641315519016e-06, | |
| "loss": 67.3567, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.8218287423332827, | |
| "grad_norm": 175.75, | |
| "learning_rate": 1.9784172661870504e-06, | |
| "loss": 68.2824, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 0.8229846196080132, | |
| "grad_norm": 157.125, | |
| "learning_rate": 1.9655704008221997e-06, | |
| "loss": 68.0355, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.8241404968827435, | |
| "grad_norm": 166.875, | |
| "learning_rate": 1.9527235354573485e-06, | |
| "loss": 65.4001, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 0.8252963741574738, | |
| "grad_norm": 161.125, | |
| "learning_rate": 1.9398766700924974e-06, | |
| "loss": 66.6461, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.8264522514322041, | |
| "grad_norm": 158.375, | |
| "learning_rate": 1.9270298047276467e-06, | |
| "loss": 66.6772, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 0.8276081287069346, | |
| "grad_norm": 147.75, | |
| "learning_rate": 1.9141829393627955e-06, | |
| "loss": 66.801, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.8287640059816649, | |
| "grad_norm": 155.25, | |
| "learning_rate": 1.9013360739979448e-06, | |
| "loss": 66.6397, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 0.8299198832563952, | |
| "grad_norm": 167.625, | |
| "learning_rate": 1.8884892086330936e-06, | |
| "loss": 67.2453, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.8310757605311256, | |
| "grad_norm": 158.5, | |
| "learning_rate": 1.8756423432682427e-06, | |
| "loss": 67.7861, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 0.832231637805856, | |
| "grad_norm": 155.625, | |
| "learning_rate": 1.8627954779033915e-06, | |
| "loss": 68.5794, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.8333875150805863, | |
| "grad_norm": 158.625, | |
| "learning_rate": 1.8499486125385408e-06, | |
| "loss": 68.7008, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 0.8345433923553167, | |
| "grad_norm": 168.25, | |
| "learning_rate": 1.8371017471736899e-06, | |
| "loss": 67.8454, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.835699269630047, | |
| "grad_norm": 155.0, | |
| "learning_rate": 1.8242548818088387e-06, | |
| "loss": 67.5842, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 0.8368551469047774, | |
| "grad_norm": 169.0, | |
| "learning_rate": 1.811408016443988e-06, | |
| "loss": 66.7899, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.8380110241795078, | |
| "grad_norm": 165.25, | |
| "learning_rate": 1.7985611510791368e-06, | |
| "loss": 66.9411, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 0.8391669014542381, | |
| "grad_norm": 158.125, | |
| "learning_rate": 1.7857142857142859e-06, | |
| "loss": 65.2401, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.8403227787289684, | |
| "grad_norm": 152.5, | |
| "learning_rate": 1.7728674203494347e-06, | |
| "loss": 66.357, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 0.8414786560036988, | |
| "grad_norm": 166.75, | |
| "learning_rate": 1.760020554984584e-06, | |
| "loss": 66.9579, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.8426345332784292, | |
| "grad_norm": 151.875, | |
| "learning_rate": 1.747173689619733e-06, | |
| "loss": 68.2974, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 0.8437904105531595, | |
| "grad_norm": 151.75, | |
| "learning_rate": 1.734326824254882e-06, | |
| "loss": 66.9873, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.8449462878278898, | |
| "grad_norm": 168.125, | |
| "learning_rate": 1.721479958890031e-06, | |
| "loss": 67.4954, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 0.8461021651026203, | |
| "grad_norm": 148.5, | |
| "learning_rate": 1.7086330935251798e-06, | |
| "loss": 67.3915, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.8472580423773506, | |
| "grad_norm": 159.125, | |
| "learning_rate": 1.695786228160329e-06, | |
| "loss": 67.0368, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 0.8484139196520809, | |
| "grad_norm": 158.0, | |
| "learning_rate": 1.682939362795478e-06, | |
| "loss": 67.0658, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.8495697969268113, | |
| "grad_norm": 152.625, | |
| "learning_rate": 1.670092497430627e-06, | |
| "loss": 65.5299, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 0.8507256742015417, | |
| "grad_norm": 151.125, | |
| "learning_rate": 1.6572456320657763e-06, | |
| "loss": 66.0855, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.851881551476272, | |
| "grad_norm": 153.5, | |
| "learning_rate": 1.6443987667009251e-06, | |
| "loss": 67.1155, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 0.8530374287510024, | |
| "grad_norm": 146.25, | |
| "learning_rate": 1.6315519013360742e-06, | |
| "loss": 66.6861, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.8541933060257327, | |
| "grad_norm": 166.0, | |
| "learning_rate": 1.618705035971223e-06, | |
| "loss": 67.9181, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 0.8553491833004631, | |
| "grad_norm": 150.625, | |
| "learning_rate": 1.6058581706063723e-06, | |
| "loss": 67.422, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.8565050605751934, | |
| "grad_norm": 159.0, | |
| "learning_rate": 1.5930113052415211e-06, | |
| "loss": 66.2905, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 0.8576609378499238, | |
| "grad_norm": 159.125, | |
| "learning_rate": 1.5801644398766702e-06, | |
| "loss": 67.4452, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.8588168151246541, | |
| "grad_norm": 151.875, | |
| "learning_rate": 1.5673175745118195e-06, | |
| "loss": 66.2737, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 0.8599726923993845, | |
| "grad_norm": 166.5, | |
| "learning_rate": 1.5544707091469683e-06, | |
| "loss": 67.6893, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.8611285696741149, | |
| "grad_norm": 158.375, | |
| "learning_rate": 1.5416238437821174e-06, | |
| "loss": 67.26, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 0.8622844469488452, | |
| "grad_norm": 164.25, | |
| "learning_rate": 1.5287769784172662e-06, | |
| "loss": 67.7357, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.8634403242235755, | |
| "grad_norm": 156.125, | |
| "learning_rate": 1.5159301130524153e-06, | |
| "loss": 65.816, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 0.864596201498306, | |
| "grad_norm": 158.0, | |
| "learning_rate": 1.5030832476875643e-06, | |
| "loss": 66.2453, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.8657520787730363, | |
| "grad_norm": 146.75, | |
| "learning_rate": 1.4902363823227134e-06, | |
| "loss": 66.5201, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 0.8669079560477666, | |
| "grad_norm": 161.125, | |
| "learning_rate": 1.4773895169578625e-06, | |
| "loss": 67.2529, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.868063833322497, | |
| "grad_norm": 180.5, | |
| "learning_rate": 1.4645426515930113e-06, | |
| "loss": 66.3079, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 0.8692197105972274, | |
| "grad_norm": 149.375, | |
| "learning_rate": 1.4516957862281606e-06, | |
| "loss": 68.2998, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.8703755878719577, | |
| "grad_norm": 165.875, | |
| "learning_rate": 1.4388489208633094e-06, | |
| "loss": 66.4161, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 0.871531465146688, | |
| "grad_norm": 164.375, | |
| "learning_rate": 1.4260020554984585e-06, | |
| "loss": 66.5728, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.8726873424214184, | |
| "grad_norm": 150.125, | |
| "learning_rate": 1.4131551901336073e-06, | |
| "loss": 66.1363, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 0.8738432196961488, | |
| "grad_norm": 143.75, | |
| "learning_rate": 1.4003083247687566e-06, | |
| "loss": 67.2544, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.8749990969708791, | |
| "grad_norm": 148.75, | |
| "learning_rate": 1.3874614594039057e-06, | |
| "loss": 65.8916, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 0.8761549742456095, | |
| "grad_norm": 163.375, | |
| "learning_rate": 1.3746145940390545e-06, | |
| "loss": 65.9047, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.8773108515203398, | |
| "grad_norm": 153.625, | |
| "learning_rate": 1.3617677286742038e-06, | |
| "loss": 66.0558, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 0.8784667287950702, | |
| "grad_norm": 168.75, | |
| "learning_rate": 1.3489208633093526e-06, | |
| "loss": 67.7262, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.8796226060698006, | |
| "grad_norm": 154.5, | |
| "learning_rate": 1.3360739979445017e-06, | |
| "loss": 66.8116, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 0.8807784833445309, | |
| "grad_norm": 141.875, | |
| "learning_rate": 1.3232271325796505e-06, | |
| "loss": 66.8746, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.8819343606192612, | |
| "grad_norm": 162.0, | |
| "learning_rate": 1.3103802672147998e-06, | |
| "loss": 66.7104, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 0.8830902378939917, | |
| "grad_norm": 147.75, | |
| "learning_rate": 1.2975334018499486e-06, | |
| "loss": 66.5884, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.884246115168722, | |
| "grad_norm": 156.125, | |
| "learning_rate": 1.2846865364850977e-06, | |
| "loss": 68.1134, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 0.8854019924434523, | |
| "grad_norm": 159.5, | |
| "learning_rate": 1.2718396711202468e-06, | |
| "loss": 66.4509, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.8865578697181826, | |
| "grad_norm": 178.5, | |
| "learning_rate": 1.2589928057553958e-06, | |
| "loss": 65.9568, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 0.887713746992913, | |
| "grad_norm": 177.75, | |
| "learning_rate": 1.2461459403905449e-06, | |
| "loss": 68.2432, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.8888696242676434, | |
| "grad_norm": 145.0, | |
| "learning_rate": 1.233299075025694e-06, | |
| "loss": 65.4491, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 0.8900255015423737, | |
| "grad_norm": 150.25, | |
| "learning_rate": 1.2204522096608428e-06, | |
| "loss": 66.6217, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.8911813788171041, | |
| "grad_norm": 167.25, | |
| "learning_rate": 1.2076053442959918e-06, | |
| "loss": 66.3795, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 0.8923372560918345, | |
| "grad_norm": 159.125, | |
| "learning_rate": 1.194758478931141e-06, | |
| "loss": 67.1028, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.8934931333665648, | |
| "grad_norm": 152.5, | |
| "learning_rate": 1.18191161356629e-06, | |
| "loss": 66.2524, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 0.8946490106412952, | |
| "grad_norm": 155.25, | |
| "learning_rate": 1.1690647482014388e-06, | |
| "loss": 65.7507, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.8958048879160255, | |
| "grad_norm": 162.25, | |
| "learning_rate": 1.156217882836588e-06, | |
| "loss": 66.8517, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 0.8969607651907558, | |
| "grad_norm": 180.875, | |
| "learning_rate": 1.1433710174717371e-06, | |
| "loss": 66.3441, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.8981166424654863, | |
| "grad_norm": 156.75, | |
| "learning_rate": 1.130524152106886e-06, | |
| "loss": 65.7155, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 0.8992725197402166, | |
| "grad_norm": 163.0, | |
| "learning_rate": 1.117677286742035e-06, | |
| "loss": 66.562, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.9004283970149469, | |
| "grad_norm": 155.0, | |
| "learning_rate": 1.1048304213771841e-06, | |
| "loss": 66.4161, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 0.9015842742896772, | |
| "grad_norm": 148.625, | |
| "learning_rate": 1.091983556012333e-06, | |
| "loss": 65.2983, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.9027401515644077, | |
| "grad_norm": 155.25, | |
| "learning_rate": 1.079136690647482e-06, | |
| "loss": 64.831, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 0.903896028839138, | |
| "grad_norm": 160.75, | |
| "learning_rate": 1.066289825282631e-06, | |
| "loss": 66.6982, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.9050519061138683, | |
| "grad_norm": 159.625, | |
| "learning_rate": 1.0534429599177801e-06, | |
| "loss": 66.8952, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 0.9062077833885988, | |
| "grad_norm": 151.5, | |
| "learning_rate": 1.0405960945529292e-06, | |
| "loss": 66.3558, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.9073636606633291, | |
| "grad_norm": 162.25, | |
| "learning_rate": 1.0277492291880783e-06, | |
| "loss": 66.1818, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 0.9085195379380594, | |
| "grad_norm": 146.125, | |
| "learning_rate": 1.0149023638232273e-06, | |
| "loss": 66.7638, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.9096754152127898, | |
| "grad_norm": 158.0, | |
| "learning_rate": 1.0020554984583762e-06, | |
| "loss": 65.9911, | |
| "step": 3935 | |
| }, | |
| { | |
| "epoch": 0.9108312924875201, | |
| "grad_norm": 156.5, | |
| "learning_rate": 9.892086330935252e-07, | |
| "loss": 66.3479, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.9119871697622505, | |
| "grad_norm": 153.75, | |
| "learning_rate": 9.763617677286743e-07, | |
| "loss": 66.4355, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 0.9131430470369809, | |
| "grad_norm": 164.875, | |
| "learning_rate": 9.635149023638233e-07, | |
| "loss": 65.8444, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.9142989243117112, | |
| "grad_norm": 157.875, | |
| "learning_rate": 9.506680369989724e-07, | |
| "loss": 65.7399, | |
| "step": 3955 | |
| }, | |
| { | |
| "epoch": 0.9154548015864415, | |
| "grad_norm": 159.0, | |
| "learning_rate": 9.378211716341213e-07, | |
| "loss": 67.6389, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.9166106788611719, | |
| "grad_norm": 165.625, | |
| "learning_rate": 9.249743062692704e-07, | |
| "loss": 67.185, | |
| "step": 3965 | |
| }, | |
| { | |
| "epoch": 0.9177665561359023, | |
| "grad_norm": 154.5, | |
| "learning_rate": 9.121274409044194e-07, | |
| "loss": 66.45, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.9189224334106326, | |
| "grad_norm": 150.125, | |
| "learning_rate": 8.992805755395684e-07, | |
| "loss": 65.8415, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 0.9200783106853629, | |
| "grad_norm": 155.125, | |
| "learning_rate": 8.864337101747174e-07, | |
| "loss": 65.8874, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.9212341879600934, | |
| "grad_norm": 158.375, | |
| "learning_rate": 8.735868448098665e-07, | |
| "loss": 66.4925, | |
| "step": 3985 | |
| }, | |
| { | |
| "epoch": 0.9223900652348237, | |
| "grad_norm": 155.25, | |
| "learning_rate": 8.607399794450155e-07, | |
| "loss": 65.5232, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.923545942509554, | |
| "grad_norm": 147.5, | |
| "learning_rate": 8.478931140801645e-07, | |
| "loss": 64.7553, | |
| "step": 3995 | |
| }, | |
| { | |
| "epoch": 0.9247018197842845, | |
| "grad_norm": 161.75, | |
| "learning_rate": 8.350462487153135e-07, | |
| "loss": 66.8576, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.9247018197842845, | |
| "eval_loss": NaN, | |
| "eval_runtime": 383.9543, | |
| "eval_samples_per_second": 607.192, | |
| "eval_steps_per_second": 37.95, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 4325, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 1.7329431971155149e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |