| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8396305625524769, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0041981528127623844, | |
| "grad_norm": 5.267488956451416, | |
| "learning_rate": 9e-06, | |
| "loss": 1.0193, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.008396305625524769, | |
| "grad_norm": 2.3370044231414795, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.3362, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.012594458438287154, | |
| "grad_norm": 2.1865203380584717, | |
| "learning_rate": 2.9e-05, | |
| "loss": 0.2021, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.016792611251049538, | |
| "grad_norm": 1.625726342201233, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 0.1466, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.020990764063811923, | |
| "grad_norm": 1.1913143396377563, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.1184, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02518891687657431, | |
| "grad_norm": 1.6674305200576782, | |
| "learning_rate": 5.9e-05, | |
| "loss": 0.1028, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.02938706968933669, | |
| "grad_norm": 0.8989318013191223, | |
| "learning_rate": 6.9e-05, | |
| "loss": 0.0836, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.033585222502099076, | |
| "grad_norm": 0.8966989517211914, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 0.0851, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.037783375314861464, | |
| "grad_norm": 0.7644183039665222, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 0.0717, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.041981528127623846, | |
| "grad_norm": 1.272506833076477, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 0.0774, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.04617968094038623, | |
| "grad_norm": 1.4885015487670898, | |
| "learning_rate": 9.999446382823013e-05, | |
| "loss": 0.0712, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.05037783375314862, | |
| "grad_norm": 1.201401710510254, | |
| "learning_rate": 9.997532801828658e-05, | |
| "loss": 0.072, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.054575986565911, | |
| "grad_norm": 1.367532730102539, | |
| "learning_rate": 9.99425294526634e-05, | |
| "loss": 0.0682, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.05877413937867338, | |
| "grad_norm": 0.9127843976020813, | |
| "learning_rate": 9.989607709816091e-05, | |
| "loss": 0.0659, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.06297229219143577, | |
| "grad_norm": 0.7713141441345215, | |
| "learning_rate": 9.983598365438902e-05, | |
| "loss": 0.0528, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.06717044500419815, | |
| "grad_norm": 0.9186705946922302, | |
| "learning_rate": 9.976226555029522e-05, | |
| "loss": 0.0495, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.07136859781696053, | |
| "grad_norm": 0.5967724919319153, | |
| "learning_rate": 9.967494293967312e-05, | |
| "loss": 0.0522, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.07556675062972293, | |
| "grad_norm": 1.0857634544372559, | |
| "learning_rate": 9.95740396956525e-05, | |
| "loss": 0.0578, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.07976490344248531, | |
| "grad_norm": 0.5992318987846375, | |
| "learning_rate": 9.945958340417283e-05, | |
| "loss": 0.0478, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.08396305625524769, | |
| "grad_norm": 0.5689753293991089, | |
| "learning_rate": 9.93316053564413e-05, | |
| "loss": 0.0478, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.08816120906801007, | |
| "grad_norm": 0.9272230863571167, | |
| "learning_rate": 9.919014054037836e-05, | |
| "loss": 0.0527, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.09235936188077246, | |
| "grad_norm": 1.0262973308563232, | |
| "learning_rate": 9.903522763105218e-05, | |
| "loss": 0.051, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.09655751469353484, | |
| "grad_norm": 0.6636225581169128, | |
| "learning_rate": 9.886690898010535e-05, | |
| "loss": 0.0429, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.10075566750629723, | |
| "grad_norm": 0.6513682007789612, | |
| "learning_rate": 9.868523060417646e-05, | |
| "loss": 0.0443, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.10495382031905962, | |
| "grad_norm": 0.7679399251937866, | |
| "learning_rate": 9.849024217231935e-05, | |
| "loss": 0.0406, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.109151973131822, | |
| "grad_norm": 0.5520310401916504, | |
| "learning_rate": 9.82819969924244e-05, | |
| "loss": 0.0345, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.11335012594458438, | |
| "grad_norm": 0.7136860489845276, | |
| "learning_rate": 9.806055199664446e-05, | |
| "loss": 0.0487, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.11754827875734676, | |
| "grad_norm": 0.5643390417098999, | |
| "learning_rate": 9.782596772583026e-05, | |
| "loss": 0.0419, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.12174643157010916, | |
| "grad_norm": 0.6041691899299622, | |
| "learning_rate": 9.757830831297914e-05, | |
| "loss": 0.0479, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.12594458438287154, | |
| "grad_norm": 0.5313841700553894, | |
| "learning_rate": 9.731764146570173e-05, | |
| "loss": 0.037, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.13014273719563393, | |
| "grad_norm": 0.4854077100753784, | |
| "learning_rate": 9.704403844771128e-05, | |
| "loss": 0.0368, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.1343408900083963, | |
| "grad_norm": 0.44807952642440796, | |
| "learning_rate": 9.675757405934103e-05, | |
| "loss": 0.0363, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.1385390428211587, | |
| "grad_norm": 0.51119464635849, | |
| "learning_rate": 9.645832661709444e-05, | |
| "loss": 0.0308, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.14273719563392107, | |
| "grad_norm": 0.5140641927719116, | |
| "learning_rate": 9.614637793223425e-05, | |
| "loss": 0.0316, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.14693534844668346, | |
| "grad_norm": 0.5101410150527954, | |
| "learning_rate": 9.582181328841611e-05, | |
| "loss": 0.0307, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.15113350125944586, | |
| "grad_norm": 0.6274232864379883, | |
| "learning_rate": 9.548472141837286e-05, | |
| "loss": 0.0324, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.15533165407220823, | |
| "grad_norm": 0.6747106313705444, | |
| "learning_rate": 9.513519447965595e-05, | |
| "loss": 0.0353, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.15952980688497062, | |
| "grad_norm": 0.38412219285964966, | |
| "learning_rate": 9.477332802944044e-05, | |
| "loss": 0.0311, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.163727959697733, | |
| "grad_norm": 0.3586186468601227, | |
| "learning_rate": 9.439922099840054e-05, | |
| "loss": 0.032, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.16792611251049538, | |
| "grad_norm": 0.38126131892204285, | |
| "learning_rate": 9.401297566366318e-05, | |
| "loss": 0.0266, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.17212426532325778, | |
| "grad_norm": 0.3684011995792389, | |
| "learning_rate": 9.36146976208462e-05, | |
| "loss": 0.0375, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.17632241813602015, | |
| "grad_norm": 0.5006753206253052, | |
| "learning_rate": 9.320449575518972e-05, | |
| "loss": 0.028, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.18052057094878254, | |
| "grad_norm": 0.5954798460006714, | |
| "learning_rate": 9.278248221178798e-05, | |
| "loss": 0.0329, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.1847187237615449, | |
| "grad_norm": 0.745754063129425, | |
| "learning_rate": 9.234877236492997e-05, | |
| "loss": 0.0339, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.1889168765743073, | |
| "grad_norm": 0.6667306423187256, | |
| "learning_rate": 9.190348478655724e-05, | |
| "loss": 0.0316, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.19311502938706968, | |
| "grad_norm": 0.7456135153770447, | |
| "learning_rate": 9.144674121384757e-05, | |
| "loss": 0.0394, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.19731318219983207, | |
| "grad_norm": 0.4713476002216339, | |
| "learning_rate": 9.097866651593317e-05, | |
| "loss": 0.031, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.20151133501259447, | |
| "grad_norm": 0.9486730098724365, | |
| "learning_rate": 9.049938865976275e-05, | |
| "loss": 0.0335, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.20570948782535683, | |
| "grad_norm": 0.8080450296401978, | |
| "learning_rate": 9.000903867511666e-05, | |
| "loss": 0.0326, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 0.6074867844581604, | |
| "learning_rate": 8.950775061878453e-05, | |
| "loss": 0.0326, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.2141057934508816, | |
| "grad_norm": 0.4722858965396881, | |
| "learning_rate": 8.899566153791566e-05, | |
| "loss": 0.0362, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.218303946263644, | |
| "grad_norm": 0.3163154721260071, | |
| "learning_rate": 8.84729114325516e-05, | |
| "loss": 0.0312, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.2225020990764064, | |
| "grad_norm": 0.45854002237319946, | |
| "learning_rate": 8.79396432173515e-05, | |
| "loss": 0.0319, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.22670025188916876, | |
| "grad_norm": 0.5331682562828064, | |
| "learning_rate": 8.739600268252078e-05, | |
| "loss": 0.0302, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.23089840470193115, | |
| "grad_norm": 0.44824129343032837, | |
| "learning_rate": 8.684213845395339e-05, | |
| "loss": 0.0256, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.23509655751469352, | |
| "grad_norm": 0.38797256350517273, | |
| "learning_rate": 8.627820195259918e-05, | |
| "loss": 0.024, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.23929471032745592, | |
| "grad_norm": 0.44604647159576416, | |
| "learning_rate": 8.570434735306671e-05, | |
| "loss": 0.0265, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.2434928631402183, | |
| "grad_norm": 0.46292269229888916, | |
| "learning_rate": 8.512073154147362e-05, | |
| "loss": 0.0238, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.24769101595298068, | |
| "grad_norm": 0.6517325043678284, | |
| "learning_rate": 8.452751407255541e-05, | |
| "loss": 0.0283, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.2518891687657431, | |
| "grad_norm": 0.3638853430747986, | |
| "learning_rate": 8.392485712604483e-05, | |
| "loss": 0.0269, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.25608732157850544, | |
| "grad_norm": 0.3911242187023163, | |
| "learning_rate": 8.331292546233362e-05, | |
| "loss": 0.0298, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.26028547439126787, | |
| "grad_norm": 0.4468225836753845, | |
| "learning_rate": 8.269188637742846e-05, | |
| "loss": 0.0235, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.26448362720403024, | |
| "grad_norm": 0.4399245083332062, | |
| "learning_rate": 8.206190965721419e-05, | |
| "loss": 0.0217, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.2686817800167926, | |
| "grad_norm": 0.34588077664375305, | |
| "learning_rate": 8.14231675310358e-05, | |
| "loss": 0.0212, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.27287993282955497, | |
| "grad_norm": 0.3927046060562134, | |
| "learning_rate": 8.077583462461283e-05, | |
| "loss": 0.0228, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.2770780856423174, | |
| "grad_norm": 0.42579185962677, | |
| "learning_rate": 8.012008791229826e-05, | |
| "loss": 0.0239, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.28127623845507976, | |
| "grad_norm": 0.37947288155555725, | |
| "learning_rate": 7.945610666869568e-05, | |
| "loss": 0.0302, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.28547439126784213, | |
| "grad_norm": 0.46608966588974, | |
| "learning_rate": 7.878407241964729e-05, | |
| "loss": 0.0269, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.28967254408060455, | |
| "grad_norm": 0.4380934536457062, | |
| "learning_rate": 7.810416889260653e-05, | |
| "loss": 0.0259, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.2938706968933669, | |
| "grad_norm": 0.4817158281803131, | |
| "learning_rate": 7.741658196640892e-05, | |
| "loss": 0.0246, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.2980688497061293, | |
| "grad_norm": 0.3391236960887909, | |
| "learning_rate": 7.672149962045457e-05, | |
| "loss": 0.0285, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.3022670025188917, | |
| "grad_norm": 0.5921253561973572, | |
| "learning_rate": 7.60191118833165e-05, | |
| "loss": 0.0192, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.3064651553316541, | |
| "grad_norm": 0.3677942156791687, | |
| "learning_rate": 7.530961078078873e-05, | |
| "loss": 0.0259, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.31066330814441645, | |
| "grad_norm": 0.4081045985221863, | |
| "learning_rate": 7.45931902833884e-05, | |
| "loss": 0.025, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.3148614609571788, | |
| "grad_norm": 0.4675191342830658, | |
| "learning_rate": 7.387004625332608e-05, | |
| "loss": 0.0223, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.31905961376994124, | |
| "grad_norm": 0.40352970361709595, | |
| "learning_rate": 7.3140376390959e-05, | |
| "loss": 0.0177, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.3232577665827036, | |
| "grad_norm": 0.4271659851074219, | |
| "learning_rate": 7.240438018074189e-05, | |
| "loss": 0.0187, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.327455919395466, | |
| "grad_norm": 0.4077198803424835, | |
| "learning_rate": 7.166225883668969e-05, | |
| "loss": 0.0198, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.3316540722082284, | |
| "grad_norm": 0.24861399829387665, | |
| "learning_rate": 7.091421524736784e-05, | |
| "loss": 0.0182, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.33585222502099077, | |
| "grad_norm": 0.3340529501438141, | |
| "learning_rate": 7.016045392042452e-05, | |
| "loss": 0.0195, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.34005037783375314, | |
| "grad_norm": 0.30124542117118835, | |
| "learning_rate": 6.940118092668022e-05, | |
| "loss": 0.0202, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.34424853064651556, | |
| "grad_norm": 0.28010880947113037, | |
| "learning_rate": 6.863660384379017e-05, | |
| "loss": 0.0192, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.34844668345927793, | |
| "grad_norm": 0.34497565031051636, | |
| "learning_rate": 6.786693169949455e-05, | |
| "loss": 0.0188, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.3526448362720403, | |
| "grad_norm": 0.38020557165145874, | |
| "learning_rate": 6.709237491447249e-05, | |
| "loss": 0.0159, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.35684298908480266, | |
| "grad_norm": 0.677865207195282, | |
| "learning_rate": 6.631314524481513e-05, | |
| "loss": 0.0214, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.3610411418975651, | |
| "grad_norm": 0.741203784942627, | |
| "learning_rate": 6.552945572413358e-05, | |
| "loss": 0.0203, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.36523929471032746, | |
| "grad_norm": 0.35501229763031006, | |
| "learning_rate": 6.474152060531768e-05, | |
| "loss": 0.0193, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.3694374475230898, | |
| "grad_norm": 0.3309725522994995, | |
| "learning_rate": 6.394955530196147e-05, | |
| "loss": 0.0152, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.37363560033585225, | |
| "grad_norm": 0.36004316806793213, | |
| "learning_rate": 6.315377632947115e-05, | |
| "loss": 0.0151, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.3778337531486146, | |
| "grad_norm": 0.35332006216049194, | |
| "learning_rate": 6.235440124587198e-05, | |
| "loss": 0.0174, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.382031905961377, | |
| "grad_norm": 0.3157385587692261, | |
| "learning_rate": 6.155164859233012e-05, | |
| "loss": 0.0125, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.38623005877413935, | |
| "grad_norm": 0.2917327284812927, | |
| "learning_rate": 6.074573783340562e-05, | |
| "loss": 0.0185, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.3904282115869018, | |
| "grad_norm": 0.20623797178268433, | |
| "learning_rate": 5.9936889297052986e-05, | |
| "loss": 0.0141, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.39462636439966414, | |
| "grad_norm": 0.30338916182518005, | |
| "learning_rate": 5.912532411438576e-05, | |
| "loss": 0.0124, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.3988245172124265, | |
| "grad_norm": 0.46562105417251587, | |
| "learning_rate": 5.831126415922148e-05, | |
| "loss": 0.0187, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.40302267002518893, | |
| "grad_norm": 0.303715318441391, | |
| "learning_rate": 5.74949319874235e-05, | |
| "loss": 0.013, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.4072208228379513, | |
| "grad_norm": 0.33936807513237, | |
| "learning_rate": 5.667655077605659e-05, | |
| "loss": 0.0147, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.41141897565071367, | |
| "grad_norm": 0.26057168841362, | |
| "learning_rate": 5.585634426237246e-05, | |
| "loss": 0.0147, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.4156171284634761, | |
| "grad_norm": 0.30878400802612305, | |
| "learning_rate": 5.5034536682642224e-05, | |
| "loss": 0.0208, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 0.32619133591651917, | |
| "learning_rate": 5.4211352710852495e-05, | |
| "loss": 0.0147, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.42401343408900083, | |
| "grad_norm": 0.4049901068210602, | |
| "learning_rate": 5.3387017397281704e-05, | |
| "loss": 0.0136, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.4282115869017632, | |
| "grad_norm": 0.263507217168808, | |
| "learning_rate": 5.2561756106973656e-05, | |
| "loss": 0.0142, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.4324097397145256, | |
| "grad_norm": 0.3729311525821686, | |
| "learning_rate": 5.1735794458124956e-05, | |
| "loss": 0.0143, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.436607892527288, | |
| "grad_norm": 0.433525413274765, | |
| "learning_rate": 5.0909358260403186e-05, | |
| "loss": 0.0159, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.44080604534005036, | |
| "grad_norm": 0.20972338318824768, | |
| "learning_rate": 5.0082673453212914e-05, | |
| "loss": 0.0125, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.4450041981528128, | |
| "grad_norm": 0.44978395104408264, | |
| "learning_rate": 4.925596604392603e-05, | |
| "loss": 0.0201, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.44920235096557515, | |
| "grad_norm": 0.33291032910346985, | |
| "learning_rate": 4.8429462046093585e-05, | |
| "loss": 0.0145, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.4534005037783375, | |
| "grad_norm": 0.3551059663295746, | |
| "learning_rate": 4.7603387417656026e-05, | |
| "loss": 0.0164, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.45759865659109994, | |
| "grad_norm": 0.2902749478816986, | |
| "learning_rate": 4.677796799916845e-05, | |
| "loss": 0.0088, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.4617968094038623, | |
| "grad_norm": 0.4253701865673065, | |
| "learning_rate": 4.5953429452058135e-05, | |
| "loss": 0.0149, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.4659949622166247, | |
| "grad_norm": 0.25408294796943665, | |
| "learning_rate": 4.5129997196930845e-05, | |
| "loss": 0.014, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.47019311502938704, | |
| "grad_norm": 0.3801613450050354, | |
| "learning_rate": 4.430789635194324e-05, | |
| "loss": 0.0135, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.47439126784214947, | |
| "grad_norm": 0.3434455990791321, | |
| "learning_rate": 4.348735167125771e-05, | |
| "loss": 0.0175, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.47858942065491183, | |
| "grad_norm": 0.38031941652297974, | |
| "learning_rate": 4.2668587483596864e-05, | |
| "loss": 0.0131, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.4827875734676742, | |
| "grad_norm": 0.2573353350162506, | |
| "learning_rate": 4.1851827630914305e-05, | |
| "loss": 0.0137, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.4869857262804366, | |
| "grad_norm": 0.31360694766044617, | |
| "learning_rate": 4.103729540719847e-05, | |
| "loss": 0.0146, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.491183879093199, | |
| "grad_norm": 0.32134002447128296, | |
| "learning_rate": 4.0225213497426276e-05, | |
| "loss": 0.0133, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.49538203190596136, | |
| "grad_norm": 0.31462499499320984, | |
| "learning_rate": 3.9415803916683224e-05, | |
| "loss": 0.0124, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.4995801847187238, | |
| "grad_norm": 0.3015807569026947, | |
| "learning_rate": 3.860928794946682e-05, | |
| "loss": 0.0147, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.5037783375314862, | |
| "grad_norm": 0.26400843262672424, | |
| "learning_rate": 3.780588608918947e-05, | |
| "loss": 0.0125, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5079764903442485, | |
| "grad_norm": 0.27983877062797546, | |
| "learning_rate": 3.700581797789786e-05, | |
| "loss": 0.0139, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.5121746431570109, | |
| "grad_norm": 0.34057655930519104, | |
| "learning_rate": 3.6209302346225006e-05, | |
| "loss": 0.0125, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.5163727959697733, | |
| "grad_norm": 0.3145323097705841, | |
| "learning_rate": 3.541655695359142e-05, | |
| "loss": 0.0141, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.5205709487825357, | |
| "grad_norm": 0.2706829905509949, | |
| "learning_rate": 3.462779852867197e-05, | |
| "loss": 0.0112, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.5247691015952981, | |
| "grad_norm": 0.3848113417625427, | |
| "learning_rate": 3.384324271014429e-05, | |
| "loss": 0.0147, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.5289672544080605, | |
| "grad_norm": 0.32747682929039, | |
| "learning_rate": 3.3063103987735433e-05, | |
| "loss": 0.0114, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.5331654072208228, | |
| "grad_norm": 0.2991146445274353, | |
| "learning_rate": 3.228759564358248e-05, | |
| "loss": 0.0141, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.5373635600335852, | |
| "grad_norm": 0.3067378103733063, | |
| "learning_rate": 3.1516929693923315e-05, | |
| "loss": 0.0129, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.5415617128463476, | |
| "grad_norm": 0.15431927144527435, | |
| "learning_rate": 3.075131683113352e-05, | |
| "loss": 0.013, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.5457598656591099, | |
| "grad_norm": 0.1479986160993576, | |
| "learning_rate": 2.999096636612518e-05, | |
| "loss": 0.0121, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5499580184718724, | |
| "grad_norm": 0.1798468679189682, | |
| "learning_rate": 2.9236086171123404e-05, | |
| "loss": 0.0107, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.5541561712846348, | |
| "grad_norm": 0.21686813235282898, | |
| "learning_rate": 2.8486882622836026e-05, | |
| "loss": 0.0124, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.5583543240973972, | |
| "grad_norm": 0.40006551146507263, | |
| "learning_rate": 2.774356054603243e-05, | |
| "loss": 0.0103, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.5625524769101595, | |
| "grad_norm": 0.39905092120170593, | |
| "learning_rate": 2.7006323157546386e-05, | |
| "loss": 0.0146, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.5667506297229219, | |
| "grad_norm": 0.156532883644104, | |
| "learning_rate": 2.6275372010718635e-05, | |
| "loss": 0.0083, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.5709487825356843, | |
| "grad_norm": 0.418869286775589, | |
| "learning_rate": 2.555090694029421e-05, | |
| "loss": 0.0112, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.5751469353484466, | |
| "grad_norm": 0.3597627878189087, | |
| "learning_rate": 2.4833126007789653e-05, | |
| "loss": 0.0071, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.5793450881612091, | |
| "grad_norm": 0.3350473940372467, | |
| "learning_rate": 2.4122225447344875e-05, | |
| "loss": 0.0094, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.5835432409739715, | |
| "grad_norm": 0.2572272717952728, | |
| "learning_rate": 2.341839961207482e-05, | |
| "loss": 0.0086, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.5877413937867338, | |
| "grad_norm": 0.44453227519989014, | |
| "learning_rate": 2.2721840920935196e-05, | |
| "loss": 0.0104, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5919395465994962, | |
| "grad_norm": 0.23089422285556793, | |
| "learning_rate": 2.2032739806117058e-05, | |
| "loss": 0.0075, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.5961376994122586, | |
| "grad_norm": 0.2570338547229767, | |
| "learning_rate": 2.1351284660984572e-05, | |
| "loss": 0.0128, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.600335852225021, | |
| "grad_norm": 0.4198639690876007, | |
| "learning_rate": 2.067766178857013e-05, | |
| "loss": 0.0133, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.6045340050377834, | |
| "grad_norm": 0.4144093990325928, | |
| "learning_rate": 2.0012055350640986e-05, | |
| "loss": 0.0095, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.6087321578505458, | |
| "grad_norm": 0.25941231846809387, | |
| "learning_rate": 1.9354647317351188e-05, | |
| "loss": 0.0105, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.6129303106633082, | |
| "grad_norm": 0.25724703073501587, | |
| "learning_rate": 1.8705617417492883e-05, | |
| "loss": 0.0096, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.6171284634760705, | |
| "grad_norm": 0.2537188231945038, | |
| "learning_rate": 1.8065143089360172e-05, | |
| "loss": 0.0094, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.6213266162888329, | |
| "grad_norm": 0.25936058163642883, | |
| "learning_rate": 1.743339943223926e-05, | |
| "loss": 0.0132, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.6255247691015953, | |
| "grad_norm": 0.27446359395980835, | |
| "learning_rate": 1.6810559158538092e-05, | |
| "loss": 0.0084, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 0.27379536628723145, | |
| "learning_rate": 1.6196792546568472e-05, | |
| "loss": 0.0126, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6339210747271201, | |
| "grad_norm": 0.16973455250263214, | |
| "learning_rate": 1.5592267393993716e-05, | |
| "loss": 0.0082, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.6381192275398825, | |
| "grad_norm": 0.19682660698890686, | |
| "learning_rate": 1.4997148971954344e-05, | |
| "loss": 0.0119, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.6423173803526449, | |
| "grad_norm": 0.16340869665145874, | |
| "learning_rate": 1.4411599979884744e-05, | |
| "loss": 0.0101, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.6465155331654072, | |
| "grad_norm": 0.27933576703071594, | |
| "learning_rate": 1.383578050103268e-05, | |
| "loss": 0.0117, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.6507136859781696, | |
| "grad_norm": 0.10343077033758163, | |
| "learning_rate": 1.3269847958694148e-05, | |
| "loss": 0.0101, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.654911838790932, | |
| "grad_norm": 0.15868137776851654, | |
| "learning_rate": 1.2713957073175425e-05, | |
| "loss": 0.0084, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.6591099916036943, | |
| "grad_norm": 0.22004792094230652, | |
| "learning_rate": 1.2168259819494066e-05, | |
| "loss": 0.0087, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.6633081444164568, | |
| "grad_norm": 0.11181829124689102, | |
| "learning_rate": 1.1632905385830484e-05, | |
| "loss": 0.0098, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.6675062972292192, | |
| "grad_norm": 0.1479823738336563, | |
| "learning_rate": 1.1108040132741354e-05, | |
| "loss": 0.0093, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.6717044500419815, | |
| "grad_norm": 0.2433646023273468, | |
| "learning_rate": 1.059380755314613e-05, | |
| "loss": 0.0088, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6759026028547439, | |
| "grad_norm": 0.212657168507576, | |
| "learning_rate": 1.009034823309749e-05, | |
| "loss": 0.0061, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.6801007556675063, | |
| "grad_norm": 0.10653030127286911, | |
| "learning_rate": 9.597799813346525e-06, | |
| "loss": 0.0068, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.6842989084802686, | |
| "grad_norm": 0.10436981171369553, | |
| "learning_rate": 9.116296951713133e-06, | |
| "loss": 0.0064, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.6884970612930311, | |
| "grad_norm": 0.28674691915512085, | |
| "learning_rate": 8.645971286271904e-06, | |
| "loss": 0.0083, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.6926952141057935, | |
| "grad_norm": 0.2074134796857834, | |
| "learning_rate": 8.186951399363613e-06, | |
| "loss": 0.0103, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.6968933669185559, | |
| "grad_norm": 0.160618856549263, | |
| "learning_rate": 7.739362782442021e-06, | |
| "loss": 0.0051, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.7010915197313182, | |
| "grad_norm": 0.2147001475095749, | |
| "learning_rate": 7.30332780176588e-06, | |
| "loss": 0.0085, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.7052896725440806, | |
| "grad_norm": 0.27866339683532715, | |
| "learning_rate": 6.878965664945108e-06, | |
| "loss": 0.0089, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.709487825356843, | |
| "grad_norm": 0.22506844997406006, | |
| "learning_rate": 6.466392388350695e-06, | |
| "loss": 0.009, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.7136859781696053, | |
| "grad_norm": 0.2676888406276703, | |
| "learning_rate": 6.0657207653969315e-06, | |
| "loss": 0.0074, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.7178841309823678, | |
| "grad_norm": 0.19724851846694946, | |
| "learning_rate": 5.67706033570487e-06, | |
| "loss": 0.0074, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.7220822837951302, | |
| "grad_norm": 0.18323613703250885, | |
| "learning_rate": 5.300517355155215e-06, | |
| "loss": 0.0093, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.7262804366078925, | |
| "grad_norm": 0.2785090208053589, | |
| "learning_rate": 4.936194766839103e-06, | |
| "loss": 0.0093, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.7304785894206549, | |
| "grad_norm": 0.16792871057987213, | |
| "learning_rate": 4.5841921729144424e-06, | |
| "loss": 0.0063, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.7346767422334173, | |
| "grad_norm": 0.1223788782954216, | |
| "learning_rate": 4.244605807375679e-06, | |
| "loss": 0.009, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.7388748950461796, | |
| "grad_norm": 0.13132202625274658, | |
| "learning_rate": 3.917528509744412e-06, | |
| "loss": 0.0093, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.743073047858942, | |
| "grad_norm": 0.2555523216724396, | |
| "learning_rate": 3.60304969968796e-06, | |
| "loss": 0.0078, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.7472712006717045, | |
| "grad_norm": 0.16517667472362518, | |
| "learning_rate": 3.301255352572946e-06, | |
| "loss": 0.0074, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.7514693534844669, | |
| "grad_norm": 0.1717490404844284, | |
| "learning_rate": 3.0122279759604745e-06, | |
| "loss": 0.0116, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.7556675062972292, | |
| "grad_norm": 0.1151094138622284, | |
| "learning_rate": 2.73604658704939e-06, | |
| "loss": 0.0075, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7598656591099916, | |
| "grad_norm": 0.19842703640460968, | |
| "learning_rate": 2.4727866910737583e-06, | |
| "loss": 0.0065, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.764063811922754, | |
| "grad_norm": 0.2035660296678543, | |
| "learning_rate": 2.222520260660521e-06, | |
| "loss": 0.0057, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.7682619647355163, | |
| "grad_norm": 0.16106460988521576, | |
| "learning_rate": 1.985315716152847e-06, | |
| "loss": 0.0058, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.7724601175482787, | |
| "grad_norm": 0.20101721584796906, | |
| "learning_rate": 1.7612379069047335e-06, | |
| "loss": 0.0059, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.7766582703610412, | |
| "grad_norm": 0.11036290228366852, | |
| "learning_rate": 1.550348093551829e-06, | |
| "loss": 0.0069, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.7808564231738035, | |
| "grad_norm": 0.14676934480667114, | |
| "learning_rate": 1.3527039312633827e-06, | |
| "loss": 0.0082, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.7850545759865659, | |
| "grad_norm": 0.21743223071098328, | |
| "learning_rate": 1.1683594539798893e-06, | |
| "loss": 0.0093, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.7892527287993283, | |
| "grad_norm": 0.09522661566734314, | |
| "learning_rate": 9.97365059640787e-07, | |
| "loss": 0.0058, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.7934508816120907, | |
| "grad_norm": 0.12366088479757309, | |
| "learning_rate": 8.397674964061075e-07, | |
| "loss": 0.0077, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.797649034424853, | |
| "grad_norm": 0.17567159235477448, | |
| "learning_rate": 6.956098498760389e-07, | |
| "loss": 0.0105, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.8018471872376155, | |
| "grad_norm": 0.17301349341869354, | |
| "learning_rate": 5.64931531311741e-07, | |
| "loss": 0.0072, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.8060453400503779, | |
| "grad_norm": 0.13477115333080292, | |
| "learning_rate": 4.4776826686069305e-07, | |
| "loss": 0.008, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.8102434928631402, | |
| "grad_norm": 0.16665753722190857, | |
| "learning_rate": 3.441520877895288e-07, | |
| "loss": 0.0078, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.8144416456759026, | |
| "grad_norm": 0.20697897672653198, | |
| "learning_rate": 2.5411132172700194e-07, | |
| "loss": 0.0079, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.818639798488665, | |
| "grad_norm": 0.23099485039710999, | |
| "learning_rate": 1.776705849195037e-07, | |
| "loss": 0.0104, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.8228379513014273, | |
| "grad_norm": 0.16362063586711884, | |
| "learning_rate": 1.1485077550122402e-07, | |
| "loss": 0.0057, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.8270361041141897, | |
| "grad_norm": 0.12974146008491516, | |
| "learning_rate": 6.566906778079917e-08, | |
| "loss": 0.0079, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.8312342569269522, | |
| "grad_norm": 0.1990516036748886, | |
| "learning_rate": 3.01389075460512e-08, | |
| "loss": 0.0101, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.8354324097397146, | |
| "grad_norm": 0.15594905614852905, | |
| "learning_rate": 8.270008388022721e-09, | |
| "loss": 0.0065, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 0.07478626072406769, | |
| "learning_rate": 6.834904537900144e-11, | |
| "loss": 0.0054, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "step": 2000, | |
| "total_flos": 0.0, | |
| "train_loss": 0.030227677578106522, | |
| "train_runtime": 2758.5082, | |
| "train_samples_per_second": 23.201, | |
| "train_steps_per_second": 0.725 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |