{ "best_global_step": 4000, "best_metric": 0.6836733584774409, "best_model_checkpoint": "./SALAMA_NEW6/checkpoint-4000", "epoch": 1.597563411224286, "eval_steps": 2000, "global_step": 4000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.003994407829039345, "grad_norm": 3.5260863304138184, "learning_rate": 1.8e-07, "loss": 0.0171, "step": 10 }, { "epoch": 0.00798881565807869, "grad_norm": 3.8677453994750977, "learning_rate": 3.8e-07, "loss": 0.0168, "step": 20 }, { "epoch": 0.011983223487118035, "grad_norm": 2.2200398445129395, "learning_rate": 5.800000000000001e-07, "loss": 0.0143, "step": 30 }, { "epoch": 0.01597763131615738, "grad_norm": 2.8865885734558105, "learning_rate": 7.8e-07, "loss": 0.0148, "step": 40 }, { "epoch": 0.019972039145196723, "grad_norm": 2.37961483001709, "learning_rate": 9.800000000000001e-07, "loss": 0.0278, "step": 50 }, { "epoch": 0.02396644697423607, "grad_norm": 0.550510585308075, "learning_rate": 1.1800000000000001e-06, "loss": 0.01, "step": 60 }, { "epoch": 0.027960854803275415, "grad_norm": 0.5346644520759583, "learning_rate": 1.3800000000000001e-06, "loss": 0.0095, "step": 70 }, { "epoch": 0.03195526263231476, "grad_norm": 0.3471826910972595, "learning_rate": 1.5800000000000001e-06, "loss": 0.0102, "step": 80 }, { "epoch": 0.035949670461354104, "grad_norm": 0.6894211769104004, "learning_rate": 1.7800000000000001e-06, "loss": 0.0097, "step": 90 }, { "epoch": 0.039944078290393446, "grad_norm": 1.4083952903747559, "learning_rate": 1.98e-06, "loss": 0.0145, "step": 100 }, { "epoch": 0.043938486119432796, "grad_norm": 0.9561022520065308, "learning_rate": 2.1800000000000003e-06, "loss": 0.0099, "step": 110 }, { "epoch": 0.04793289394847214, "grad_norm": 0.5757145285606384, "learning_rate": 2.38e-06, "loss": 0.0078, "step": 120 }, { "epoch": 0.05192730177751148, "grad_norm": 0.9376910924911499, "learning_rate": 2.5800000000000003e-06, "loss": 0.0135, "step": 130 }, { "epoch": 0.05592170960655083, "grad_norm": 1.6390388011932373, "learning_rate": 2.7800000000000005e-06, "loss": 0.0157, "step": 140 }, { "epoch": 0.05991611743559017, "grad_norm": 1.0386844873428345, "learning_rate": 2.9800000000000003e-06, "loss": 0.0099, "step": 150 }, { "epoch": 0.06391052526462952, "grad_norm": 1.7207807302474976, "learning_rate": 3.1800000000000005e-06, "loss": 0.0096, "step": 160 }, { "epoch": 0.06790493309366886, "grad_norm": 0.7114217877388, "learning_rate": 3.3800000000000007e-06, "loss": 0.0137, "step": 170 }, { "epoch": 0.07189934092270821, "grad_norm": 2.062455654144287, "learning_rate": 3.58e-06, "loss": 0.0181, "step": 180 }, { "epoch": 0.07589374875174755, "grad_norm": 3.8748466968536377, "learning_rate": 3.7800000000000002e-06, "loss": 0.0155, "step": 190 }, { "epoch": 0.07988815658078689, "grad_norm": 2.8697941303253174, "learning_rate": 3.980000000000001e-06, "loss": 0.0206, "step": 200 }, { "epoch": 0.08388256440982625, "grad_norm": 1.5111762285232544, "learning_rate": 4.18e-06, "loss": 0.014, "step": 210 }, { "epoch": 0.08787697223886559, "grad_norm": 2.7273306846618652, "learning_rate": 4.38e-06, "loss": 0.0115, "step": 220 }, { "epoch": 0.09187138006790493, "grad_norm": 1.604333758354187, "learning_rate": 4.58e-06, "loss": 0.0111, "step": 230 }, { "epoch": 0.09586578789694428, "grad_norm": 1.6656010150909424, "learning_rate": 4.78e-06, "loss": 0.0103, "step": 240 }, { "epoch": 0.09986019572598362, "grad_norm": 1.9466540813446045, "learning_rate": 4.980000000000001e-06, "loss": 0.0118, "step": 250 }, { "epoch": 0.10385460355502296, "grad_norm": 0.7741212844848633, "learning_rate": 5.18e-06, "loss": 0.0203, "step": 260 }, { "epoch": 0.10784901138406232, "grad_norm": 1.5594321489334106, "learning_rate": 5.380000000000001e-06, "loss": 0.0132, "step": 270 }, { "epoch": 0.11184341921310166, "grad_norm": 2.1854960918426514, "learning_rate": 5.580000000000001e-06, "loss": 0.0227, "step": 280 }, { "epoch": 0.115837827042141, "grad_norm": 1.809208869934082, "learning_rate": 5.78e-06, "loss": 0.014, "step": 290 }, { "epoch": 0.11983223487118035, "grad_norm": 2.003018617630005, "learning_rate": 5.98e-06, "loss": 0.0253, "step": 300 }, { "epoch": 0.12382664270021969, "grad_norm": 2.226447582244873, "learning_rate": 6.18e-06, "loss": 0.0141, "step": 310 }, { "epoch": 0.12782105052925904, "grad_norm": 3.213719367980957, "learning_rate": 6.380000000000001e-06, "loss": 0.0222, "step": 320 }, { "epoch": 0.1318154583582984, "grad_norm": 2.7083969116210938, "learning_rate": 6.5800000000000005e-06, "loss": 0.0165, "step": 330 }, { "epoch": 0.13580986618733773, "grad_norm": 2.327775001525879, "learning_rate": 6.780000000000001e-06, "loss": 0.0127, "step": 340 }, { "epoch": 0.13980427401637707, "grad_norm": 2.199615478515625, "learning_rate": 6.98e-06, "loss": 0.0092, "step": 350 }, { "epoch": 0.14379868184541642, "grad_norm": 2.6200337409973145, "learning_rate": 7.180000000000001e-06, "loss": 0.0229, "step": 360 }, { "epoch": 0.14779308967445576, "grad_norm": 2.0899276733398438, "learning_rate": 7.3800000000000005e-06, "loss": 0.0123, "step": 370 }, { "epoch": 0.1517874975034951, "grad_norm": 3.313321113586426, "learning_rate": 7.58e-06, "loss": 0.0225, "step": 380 }, { "epoch": 0.15578190533253444, "grad_norm": 2.0713999271392822, "learning_rate": 7.78e-06, "loss": 0.0166, "step": 390 }, { "epoch": 0.15977631316157379, "grad_norm": 3.2806386947631836, "learning_rate": 7.980000000000002e-06, "loss": 0.0143, "step": 400 }, { "epoch": 0.16377072099061313, "grad_norm": 2.053285598754883, "learning_rate": 8.18e-06, "loss": 0.0183, "step": 410 }, { "epoch": 0.1677651288196525, "grad_norm": 1.4635854959487915, "learning_rate": 8.380000000000001e-06, "loss": 0.0198, "step": 420 }, { "epoch": 0.17175953664869184, "grad_norm": 1.461671233177185, "learning_rate": 8.580000000000001e-06, "loss": 0.0223, "step": 430 }, { "epoch": 0.17575394447773118, "grad_norm": 1.910568118095398, "learning_rate": 8.78e-06, "loss": 0.0127, "step": 440 }, { "epoch": 0.17974835230677053, "grad_norm": 2.9844930171966553, "learning_rate": 8.98e-06, "loss": 0.0198, "step": 450 }, { "epoch": 0.18374276013580987, "grad_norm": 3.3782155513763428, "learning_rate": 9.180000000000002e-06, "loss": 0.0277, "step": 460 }, { "epoch": 0.1877371679648492, "grad_norm": 2.6012775897979736, "learning_rate": 9.38e-06, "loss": 0.0154, "step": 470 }, { "epoch": 0.19173157579388855, "grad_norm": 2.880096197128296, "learning_rate": 9.58e-06, "loss": 0.0219, "step": 480 }, { "epoch": 0.1957259836229279, "grad_norm": 3.002333879470825, "learning_rate": 9.780000000000001e-06, "loss": 0.0179, "step": 490 }, { "epoch": 0.19972039145196724, "grad_norm": 6.162044525146484, "learning_rate": 9.980000000000001e-06, "loss": 0.0165, "step": 500 }, { "epoch": 0.20371479928100658, "grad_norm": 3.4659605026245117, "learning_rate": 9.98716486023959e-06, "loss": 0.0239, "step": 510 }, { "epoch": 0.20770920711004592, "grad_norm": 3.2337770462036133, "learning_rate": 9.972903593839133e-06, "loss": 0.017, "step": 520 }, { "epoch": 0.2117036149390853, "grad_norm": 1.7297512292861938, "learning_rate": 9.958642327438678e-06, "loss": 0.0197, "step": 530 }, { "epoch": 0.21569802276812464, "grad_norm": 3.336660385131836, "learning_rate": 9.944381061038221e-06, "loss": 0.0215, "step": 540 }, { "epoch": 0.21969243059716398, "grad_norm": 1.850265383720398, "learning_rate": 9.930119794637765e-06, "loss": 0.0259, "step": 550 }, { "epoch": 0.22368683842620332, "grad_norm": 2.2153711318969727, "learning_rate": 9.91585852823731e-06, "loss": 0.0192, "step": 560 }, { "epoch": 0.22768124625524266, "grad_norm": 1.2140111923217773, "learning_rate": 9.901597261836851e-06, "loss": 0.0214, "step": 570 }, { "epoch": 0.231675654084282, "grad_norm": 2.983187198638916, "learning_rate": 9.887335995436396e-06, "loss": 0.0212, "step": 580 }, { "epoch": 0.23567006191332135, "grad_norm": 1.3873775005340576, "learning_rate": 9.87307472903594e-06, "loss": 0.0278, "step": 590 }, { "epoch": 0.2396644697423607, "grad_norm": 2.8896610736846924, "learning_rate": 9.858813462635483e-06, "loss": 0.0188, "step": 600 }, { "epoch": 0.24365887757140003, "grad_norm": 3.1558310985565186, "learning_rate": 9.844552196235026e-06, "loss": 0.0268, "step": 610 }, { "epoch": 0.24765328540043938, "grad_norm": 1.7268836498260498, "learning_rate": 9.83029092983457e-06, "loss": 0.0214, "step": 620 }, { "epoch": 0.2516476932294787, "grad_norm": 3.3516862392425537, "learning_rate": 9.816029663434114e-06, "loss": 0.0264, "step": 630 }, { "epoch": 0.2556421010585181, "grad_norm": 2.8272225856781006, "learning_rate": 9.801768397033657e-06, "loss": 0.0279, "step": 640 }, { "epoch": 0.2596365088875574, "grad_norm": 2.9199869632720947, "learning_rate": 9.787507130633202e-06, "loss": 0.0225, "step": 650 }, { "epoch": 0.2636309167165968, "grad_norm": 2.80952787399292, "learning_rate": 9.773245864232744e-06, "loss": 0.0188, "step": 660 }, { "epoch": 0.2676253245456361, "grad_norm": 2.8590474128723145, "learning_rate": 9.758984597832289e-06, "loss": 0.0214, "step": 670 }, { "epoch": 0.27161973237467546, "grad_norm": 2.1354668140411377, "learning_rate": 9.744723331431832e-06, "loss": 0.0241, "step": 680 }, { "epoch": 0.2756141402037148, "grad_norm": 1.54200279712677, "learning_rate": 9.730462065031375e-06, "loss": 0.0185, "step": 690 }, { "epoch": 0.27960854803275415, "grad_norm": 2.3145532608032227, "learning_rate": 9.71620079863092e-06, "loss": 0.0162, "step": 700 }, { "epoch": 0.2836029558617935, "grad_norm": 2.7305352687835693, "learning_rate": 9.701939532230463e-06, "loss": 0.0246, "step": 710 }, { "epoch": 0.28759736369083283, "grad_norm": 1.983603596687317, "learning_rate": 9.687678265830007e-06, "loss": 0.0236, "step": 720 }, { "epoch": 0.2915917715198722, "grad_norm": 1.5119189023971558, "learning_rate": 9.67341699942955e-06, "loss": 0.02, "step": 730 }, { "epoch": 0.2955861793489115, "grad_norm": 2.9574663639068604, "learning_rate": 9.659155733029095e-06, "loss": 0.024, "step": 740 }, { "epoch": 0.2995805871779509, "grad_norm": 2.4619083404541016, "learning_rate": 9.644894466628636e-06, "loss": 0.0267, "step": 750 }, { "epoch": 0.3035749950069902, "grad_norm": 1.9842376708984375, "learning_rate": 9.630633200228181e-06, "loss": 0.0183, "step": 760 }, { "epoch": 0.30756940283602957, "grad_norm": 2.300018310546875, "learning_rate": 9.616371933827725e-06, "loss": 0.022, "step": 770 }, { "epoch": 0.3115638106650689, "grad_norm": 1.9513983726501465, "learning_rate": 9.602110667427268e-06, "loss": 0.0183, "step": 780 }, { "epoch": 0.31555821849410826, "grad_norm": 1.9509272575378418, "learning_rate": 9.587849401026813e-06, "loss": 0.0239, "step": 790 }, { "epoch": 0.31955262632314757, "grad_norm": 2.2186226844787598, "learning_rate": 9.573588134626356e-06, "loss": 0.0244, "step": 800 }, { "epoch": 0.32354703415218694, "grad_norm": 3.058155059814453, "learning_rate": 9.5593268682259e-06, "loss": 0.0225, "step": 810 }, { "epoch": 0.32754144198122626, "grad_norm": 2.130378484725952, "learning_rate": 9.545065601825442e-06, "loss": 0.0302, "step": 820 }, { "epoch": 0.3315358498102656, "grad_norm": 2.9102509021759033, "learning_rate": 9.530804335424987e-06, "loss": 0.0223, "step": 830 }, { "epoch": 0.335530257639305, "grad_norm": 3.1255991458892822, "learning_rate": 9.516543069024529e-06, "loss": 0.0199, "step": 840 }, { "epoch": 0.3395246654683443, "grad_norm": 1.9874292612075806, "learning_rate": 9.502281802624074e-06, "loss": 0.0214, "step": 850 }, { "epoch": 0.3435190732973837, "grad_norm": 1.9070899486541748, "learning_rate": 9.488020536223617e-06, "loss": 0.0283, "step": 860 }, { "epoch": 0.347513481126423, "grad_norm": 2.783308744430542, "learning_rate": 9.47375926982316e-06, "loss": 0.0172, "step": 870 }, { "epoch": 0.35150788895546237, "grad_norm": 1.1049232482910156, "learning_rate": 9.459498003422705e-06, "loss": 0.0261, "step": 880 }, { "epoch": 0.3555022967845017, "grad_norm": 3.386366844177246, "learning_rate": 9.445236737022249e-06, "loss": 0.0214, "step": 890 }, { "epoch": 0.35949670461354105, "grad_norm": 1.623195767402649, "learning_rate": 9.430975470621792e-06, "loss": 0.0221, "step": 900 }, { "epoch": 0.36349111244258037, "grad_norm": 3.33766508102417, "learning_rate": 9.416714204221335e-06, "loss": 0.0245, "step": 910 }, { "epoch": 0.36748552027161974, "grad_norm": 3.1946351528167725, "learning_rate": 9.40245293782088e-06, "loss": 0.0225, "step": 920 }, { "epoch": 0.37147992810065905, "grad_norm": 2.7116000652313232, "learning_rate": 9.388191671420423e-06, "loss": 0.0224, "step": 930 }, { "epoch": 0.3754743359296984, "grad_norm": 1.7920254468917847, "learning_rate": 9.373930405019966e-06, "loss": 0.024, "step": 940 }, { "epoch": 0.3794687437587378, "grad_norm": 3.112621545791626, "learning_rate": 9.35966913861951e-06, "loss": 0.0179, "step": 950 }, { "epoch": 0.3834631515877771, "grad_norm": 2.081601142883301, "learning_rate": 9.345407872219053e-06, "loss": 0.0237, "step": 960 }, { "epoch": 0.3874575594168165, "grad_norm": 2.1594202518463135, "learning_rate": 9.331146605818598e-06, "loss": 0.0286, "step": 970 }, { "epoch": 0.3914519672458558, "grad_norm": 2.277575969696045, "learning_rate": 9.316885339418141e-06, "loss": 0.0199, "step": 980 }, { "epoch": 0.39544637507489516, "grad_norm": 2.457400321960449, "learning_rate": 9.302624073017684e-06, "loss": 0.0148, "step": 990 }, { "epoch": 0.3994407829039345, "grad_norm": 2.731627941131592, "learning_rate": 9.288362806617228e-06, "loss": 0.026, "step": 1000 }, { "epoch": 0.40343519073297385, "grad_norm": 2.26240873336792, "learning_rate": 9.274101540216773e-06, "loss": 0.0205, "step": 1010 }, { "epoch": 0.40742959856201316, "grad_norm": 3.270246982574463, "learning_rate": 9.259840273816316e-06, "loss": 0.0245, "step": 1020 }, { "epoch": 0.41142400639105253, "grad_norm": 2.118734836578369, "learning_rate": 9.245579007415859e-06, "loss": 0.0244, "step": 1030 }, { "epoch": 0.41541841422009185, "grad_norm": 2.5819344520568848, "learning_rate": 9.231317741015402e-06, "loss": 0.0204, "step": 1040 }, { "epoch": 0.4194128220491312, "grad_norm": 3.1689200401306152, "learning_rate": 9.217056474614946e-06, "loss": 0.0283, "step": 1050 }, { "epoch": 0.4234072298781706, "grad_norm": 2.6276655197143555, "learning_rate": 9.20279520821449e-06, "loss": 0.022, "step": 1060 }, { "epoch": 0.4274016377072099, "grad_norm": 3.010298013687134, "learning_rate": 9.188533941814034e-06, "loss": 0.0355, "step": 1070 }, { "epoch": 0.4313960455362493, "grad_norm": 2.780543327331543, "learning_rate": 9.174272675413579e-06, "loss": 0.0341, "step": 1080 }, { "epoch": 0.4353904533652886, "grad_norm": 2.277057647705078, "learning_rate": 9.16001140901312e-06, "loss": 0.0239, "step": 1090 }, { "epoch": 0.43938486119432796, "grad_norm": 1.640129566192627, "learning_rate": 9.145750142612665e-06, "loss": 0.0228, "step": 1100 }, { "epoch": 0.4433792690233673, "grad_norm": 3.9390289783477783, "learning_rate": 9.131488876212208e-06, "loss": 0.0299, "step": 1110 }, { "epoch": 0.44737367685240664, "grad_norm": 1.726510763168335, "learning_rate": 9.117227609811752e-06, "loss": 0.0207, "step": 1120 }, { "epoch": 0.45136808468144596, "grad_norm": 1.583237648010254, "learning_rate": 9.102966343411297e-06, "loss": 0.018, "step": 1130 }, { "epoch": 0.45536249251048533, "grad_norm": 3.3998148441314697, "learning_rate": 9.08870507701084e-06, "loss": 0.0333, "step": 1140 }, { "epoch": 0.45935690033952464, "grad_norm": 1.7249724864959717, "learning_rate": 9.074443810610383e-06, "loss": 0.0244, "step": 1150 }, { "epoch": 0.463351308168564, "grad_norm": 2.0827536582946777, "learning_rate": 9.060182544209926e-06, "loss": 0.0295, "step": 1160 }, { "epoch": 0.46734571599760333, "grad_norm": 2.095423698425293, "learning_rate": 9.045921277809471e-06, "loss": 0.025, "step": 1170 }, { "epoch": 0.4713401238266427, "grad_norm": 2.7629899978637695, "learning_rate": 9.031660011409013e-06, "loss": 0.0139, "step": 1180 }, { "epoch": 0.47533453165568207, "grad_norm": 1.6056689023971558, "learning_rate": 9.017398745008558e-06, "loss": 0.0379, "step": 1190 }, { "epoch": 0.4793289394847214, "grad_norm": 2.5653021335601807, "learning_rate": 9.003137478608101e-06, "loss": 0.0236, "step": 1200 }, { "epoch": 0.48332334731376075, "grad_norm": 2.9385814666748047, "learning_rate": 8.988876212207644e-06, "loss": 0.035, "step": 1210 }, { "epoch": 0.48731775514280007, "grad_norm": 3.0189478397369385, "learning_rate": 8.97461494580719e-06, "loss": 0.0187, "step": 1220 }, { "epoch": 0.49131216297183944, "grad_norm": 1.1368498802185059, "learning_rate": 8.960353679406733e-06, "loss": 0.0205, "step": 1230 }, { "epoch": 0.49530657080087875, "grad_norm": 2.0715737342834473, "learning_rate": 8.946092413006276e-06, "loss": 0.021, "step": 1240 }, { "epoch": 0.4993009786299181, "grad_norm": 2.2337212562561035, "learning_rate": 8.931831146605819e-06, "loss": 0.0216, "step": 1250 }, { "epoch": 0.5032953864589574, "grad_norm": 1.4322441816329956, "learning_rate": 8.917569880205364e-06, "loss": 0.0188, "step": 1260 }, { "epoch": 0.5072897942879968, "grad_norm": 1.9030108451843262, "learning_rate": 8.903308613804906e-06, "loss": 0.0177, "step": 1270 }, { "epoch": 0.5112842021170362, "grad_norm": 2.806227684020996, "learning_rate": 8.88904734740445e-06, "loss": 0.0166, "step": 1280 }, { "epoch": 0.5152786099460755, "grad_norm": 3.692470073699951, "learning_rate": 8.874786081003994e-06, "loss": 0.032, "step": 1290 }, { "epoch": 0.5192730177751148, "grad_norm": 2.1846556663513184, "learning_rate": 8.860524814603537e-06, "loss": 0.0203, "step": 1300 }, { "epoch": 0.5232674256041542, "grad_norm": 2.7069506645202637, "learning_rate": 8.846263548203082e-06, "loss": 0.0216, "step": 1310 }, { "epoch": 0.5272618334331935, "grad_norm": 3.010519027709961, "learning_rate": 8.832002281802625e-06, "loss": 0.0291, "step": 1320 }, { "epoch": 0.5312562412622329, "grad_norm": 2.396578073501587, "learning_rate": 8.817741015402168e-06, "loss": 0.0315, "step": 1330 }, { "epoch": 0.5352506490912722, "grad_norm": 3.0465903282165527, "learning_rate": 8.803479749001712e-06, "loss": 0.0229, "step": 1340 }, { "epoch": 0.5392450569203115, "grad_norm": 3.0167880058288574, "learning_rate": 8.789218482601257e-06, "loss": 0.0237, "step": 1350 }, { "epoch": 0.5432394647493509, "grad_norm": 2.1383025646209717, "learning_rate": 8.7749572162008e-06, "loss": 0.0204, "step": 1360 }, { "epoch": 0.5472338725783903, "grad_norm": 4.080812931060791, "learning_rate": 8.760695949800343e-06, "loss": 0.0289, "step": 1370 }, { "epoch": 0.5512282804074295, "grad_norm": 1.5765025615692139, "learning_rate": 8.746434683399886e-06, "loss": 0.0266, "step": 1380 }, { "epoch": 0.5552226882364689, "grad_norm": 2.26448917388916, "learning_rate": 8.73217341699943e-06, "loss": 0.0209, "step": 1390 }, { "epoch": 0.5592170960655083, "grad_norm": 2.3659119606018066, "learning_rate": 8.717912150598975e-06, "loss": 0.027, "step": 1400 }, { "epoch": 0.5632115038945477, "grad_norm": 4.025057315826416, "learning_rate": 8.703650884198518e-06, "loss": 0.0349, "step": 1410 }, { "epoch": 0.567205911723587, "grad_norm": 1.6939404010772705, "learning_rate": 8.689389617798061e-06, "loss": 0.0164, "step": 1420 }, { "epoch": 0.5712003195526263, "grad_norm": 2.2709405422210693, "learning_rate": 8.675128351397604e-06, "loss": 0.0239, "step": 1430 }, { "epoch": 0.5751947273816657, "grad_norm": 1.9596635103225708, "learning_rate": 8.66086708499715e-06, "loss": 0.0223, "step": 1440 }, { "epoch": 0.579189135210705, "grad_norm": 1.6292943954467773, "learning_rate": 8.646605818596692e-06, "loss": 0.0211, "step": 1450 }, { "epoch": 0.5831835430397444, "grad_norm": 2.063108205795288, "learning_rate": 8.632344552196236e-06, "loss": 0.0187, "step": 1460 }, { "epoch": 0.5871779508687837, "grad_norm": 2.243898868560791, "learning_rate": 8.618083285795779e-06, "loss": 0.0201, "step": 1470 }, { "epoch": 0.591172358697823, "grad_norm": 2.1789815425872803, "learning_rate": 8.603822019395322e-06, "loss": 0.0275, "step": 1480 }, { "epoch": 0.5951667665268624, "grad_norm": 2.9697136878967285, "learning_rate": 8.589560752994867e-06, "loss": 0.0265, "step": 1490 }, { "epoch": 0.5991611743559018, "grad_norm": 2.9582135677337646, "learning_rate": 8.57529948659441e-06, "loss": 0.0222, "step": 1500 }, { "epoch": 0.603155582184941, "grad_norm": 3.5698182582855225, "learning_rate": 8.561038220193954e-06, "loss": 0.0356, "step": 1510 }, { "epoch": 0.6071499900139804, "grad_norm": 1.6024861335754395, "learning_rate": 8.546776953793497e-06, "loss": 0.0262, "step": 1520 }, { "epoch": 0.6111443978430198, "grad_norm": 3.0098986625671387, "learning_rate": 8.532515687393042e-06, "loss": 0.0193, "step": 1530 }, { "epoch": 0.6151388056720591, "grad_norm": 2.4644837379455566, "learning_rate": 8.518254420992585e-06, "loss": 0.0228, "step": 1540 }, { "epoch": 0.6191332135010985, "grad_norm": 1.8468703031539917, "learning_rate": 8.503993154592128e-06, "loss": 0.0205, "step": 1550 }, { "epoch": 0.6231276213301378, "grad_norm": 1.8541923761367798, "learning_rate": 8.489731888191672e-06, "loss": 0.0288, "step": 1560 }, { "epoch": 0.6271220291591771, "grad_norm": 2.5148401260375977, "learning_rate": 8.475470621791215e-06, "loss": 0.0236, "step": 1570 }, { "epoch": 0.6311164369882165, "grad_norm": 1.897991418838501, "learning_rate": 8.46120935539076e-06, "loss": 0.0223, "step": 1580 }, { "epoch": 0.6351108448172559, "grad_norm": 2.4308152198791504, "learning_rate": 8.446948088990303e-06, "loss": 0.0212, "step": 1590 }, { "epoch": 0.6391052526462951, "grad_norm": 2.6001315116882324, "learning_rate": 8.432686822589846e-06, "loss": 0.0305, "step": 1600 }, { "epoch": 0.6430996604753345, "grad_norm": 2.6471099853515625, "learning_rate": 8.41842555618939e-06, "loss": 0.0217, "step": 1610 }, { "epoch": 0.6470940683043739, "grad_norm": 1.336553692817688, "learning_rate": 8.404164289788934e-06, "loss": 0.0219, "step": 1620 }, { "epoch": 0.6510884761334133, "grad_norm": 2.2210710048675537, "learning_rate": 8.389903023388478e-06, "loss": 0.0229, "step": 1630 }, { "epoch": 0.6550828839624525, "grad_norm": 1.4983543157577515, "learning_rate": 8.375641756988021e-06, "loss": 0.0229, "step": 1640 }, { "epoch": 0.6590772917914919, "grad_norm": 4.39826774597168, "learning_rate": 8.361380490587566e-06, "loss": 0.0258, "step": 1650 }, { "epoch": 0.6630716996205313, "grad_norm": 1.8835957050323486, "learning_rate": 8.347119224187107e-06, "loss": 0.0252, "step": 1660 }, { "epoch": 0.6670661074495706, "grad_norm": 2.090482473373413, "learning_rate": 8.332857957786652e-06, "loss": 0.0249, "step": 1670 }, { "epoch": 0.67106051527861, "grad_norm": 4.0251336097717285, "learning_rate": 8.318596691386196e-06, "loss": 0.0358, "step": 1680 }, { "epoch": 0.6750549231076493, "grad_norm": 2.3550221920013428, "learning_rate": 8.304335424985739e-06, "loss": 0.0257, "step": 1690 }, { "epoch": 0.6790493309366886, "grad_norm": 3.605212926864624, "learning_rate": 8.290074158585282e-06, "loss": 0.0243, "step": 1700 }, { "epoch": 0.683043738765728, "grad_norm": 3.9534785747528076, "learning_rate": 8.275812892184827e-06, "loss": 0.021, "step": 1710 }, { "epoch": 0.6870381465947674, "grad_norm": 2.5969812870025635, "learning_rate": 8.26155162578437e-06, "loss": 0.0202, "step": 1720 }, { "epoch": 0.6910325544238066, "grad_norm": 2.922255277633667, "learning_rate": 8.247290359383914e-06, "loss": 0.0183, "step": 1730 }, { "epoch": 0.695026962252846, "grad_norm": 2.6963930130004883, "learning_rate": 8.233029092983458e-06, "loss": 0.0288, "step": 1740 }, { "epoch": 0.6990213700818854, "grad_norm": 1.4426485300064087, "learning_rate": 8.218767826583002e-06, "loss": 0.0267, "step": 1750 }, { "epoch": 0.7030157779109247, "grad_norm": 1.6854075193405151, "learning_rate": 8.204506560182545e-06, "loss": 0.0189, "step": 1760 }, { "epoch": 0.7070101857399641, "grad_norm": 2.4257256984710693, "learning_rate": 8.190245293782088e-06, "loss": 0.0256, "step": 1770 }, { "epoch": 0.7110045935690034, "grad_norm": 2.0894641876220703, "learning_rate": 8.175984027381633e-06, "loss": 0.02, "step": 1780 }, { "epoch": 0.7149990013980427, "grad_norm": 2.5761284828186035, "learning_rate": 8.161722760981175e-06, "loss": 0.0205, "step": 1790 }, { "epoch": 0.7189934092270821, "grad_norm": 4.715319633483887, "learning_rate": 8.14746149458072e-06, "loss": 0.0219, "step": 1800 }, { "epoch": 0.7229878170561215, "grad_norm": 2.933713674545288, "learning_rate": 8.133200228180263e-06, "loss": 0.0215, "step": 1810 }, { "epoch": 0.7269822248851607, "grad_norm": 2.0981826782226562, "learning_rate": 8.118938961779806e-06, "loss": 0.0193, "step": 1820 }, { "epoch": 0.7309766327142001, "grad_norm": 3.5366568565368652, "learning_rate": 8.104677695379351e-06, "loss": 0.0287, "step": 1830 }, { "epoch": 0.7349710405432395, "grad_norm": 2.165713310241699, "learning_rate": 8.090416428978894e-06, "loss": 0.0183, "step": 1840 }, { "epoch": 0.7389654483722788, "grad_norm": 3.0615482330322266, "learning_rate": 8.076155162578438e-06, "loss": 0.0233, "step": 1850 }, { "epoch": 0.7429598562013181, "grad_norm": 2.1526572704315186, "learning_rate": 8.06189389617798e-06, "loss": 0.0316, "step": 1860 }, { "epoch": 0.7469542640303575, "grad_norm": 1.9894167184829712, "learning_rate": 8.047632629777526e-06, "loss": 0.0214, "step": 1870 }, { "epoch": 0.7509486718593968, "grad_norm": 1.5561473369598389, "learning_rate": 8.033371363377069e-06, "loss": 0.0246, "step": 1880 }, { "epoch": 0.7549430796884362, "grad_norm": 2.030789375305176, "learning_rate": 8.019110096976612e-06, "loss": 0.028, "step": 1890 }, { "epoch": 0.7589374875174756, "grad_norm": 2.139389991760254, "learning_rate": 8.004848830576156e-06, "loss": 0.0192, "step": 1900 }, { "epoch": 0.7629318953465148, "grad_norm": 1.4104957580566406, "learning_rate": 7.990587564175699e-06, "loss": 0.0214, "step": 1910 }, { "epoch": 0.7669263031755542, "grad_norm": 3.027559280395508, "learning_rate": 7.976326297775244e-06, "loss": 0.0225, "step": 1920 }, { "epoch": 0.7709207110045936, "grad_norm": 2.3780980110168457, "learning_rate": 7.962065031374787e-06, "loss": 0.0231, "step": 1930 }, { "epoch": 0.774915118833633, "grad_norm": 2.3442487716674805, "learning_rate": 7.94780376497433e-06, "loss": 0.0223, "step": 1940 }, { "epoch": 0.7789095266626722, "grad_norm": 2.0246195793151855, "learning_rate": 7.933542498573873e-06, "loss": 0.0164, "step": 1950 }, { "epoch": 0.7829039344917116, "grad_norm": 1.8316524028778076, "learning_rate": 7.919281232173418e-06, "loss": 0.019, "step": 1960 }, { "epoch": 0.786898342320751, "grad_norm": 1.7914390563964844, "learning_rate": 7.905019965772962e-06, "loss": 0.0178, "step": 1970 }, { "epoch": 0.7908927501497903, "grad_norm": 1.9735348224639893, "learning_rate": 7.890758699372505e-06, "loss": 0.0144, "step": 1980 }, { "epoch": 0.7948871579788296, "grad_norm": 1.8981409072875977, "learning_rate": 7.876497432972048e-06, "loss": 0.0244, "step": 1990 }, { "epoch": 0.798881565807869, "grad_norm": 1.4974066019058228, "learning_rate": 7.862236166571591e-06, "loss": 0.0239, "step": 2000 }, { "epoch": 0.798881565807869, "eval_loss": 0.014320253394544125, "eval_runtime": 7268.4159, "eval_samples_per_second": 2.755, "eval_steps_per_second": 0.345, "eval_wer": 1.2956099257632903, "step": 2000 }, { "epoch": 0.8028759736369083, "grad_norm": 3.2215046882629395, "learning_rate": 7.847974900171136e-06, "loss": 0.0215, "step": 2010 }, { "epoch": 0.8068703814659477, "grad_norm": 2.5289711952209473, "learning_rate": 7.83371363377068e-06, "loss": 0.0243, "step": 2020 }, { "epoch": 0.8108647892949871, "grad_norm": 3.7780346870422363, "learning_rate": 7.819452367370223e-06, "loss": 0.0328, "step": 2030 }, { "epoch": 0.8148591971240263, "grad_norm": 1.0975768566131592, "learning_rate": 7.805191100969766e-06, "loss": 0.0213, "step": 2040 }, { "epoch": 0.8188536049530657, "grad_norm": 2.9976754188537598, "learning_rate": 7.790929834569311e-06, "loss": 0.0225, "step": 2050 }, { "epoch": 0.8228480127821051, "grad_norm": 2.6388089656829834, "learning_rate": 7.776668568168854e-06, "loss": 0.0246, "step": 2060 }, { "epoch": 0.8268424206111444, "grad_norm": 2.533771276473999, "learning_rate": 7.762407301768397e-06, "loss": 0.0183, "step": 2070 }, { "epoch": 0.8308368284401837, "grad_norm": 2.3957924842834473, "learning_rate": 7.748146035367942e-06, "loss": 0.0204, "step": 2080 }, { "epoch": 0.8348312362692231, "grad_norm": 2.402613401412964, "learning_rate": 7.733884768967484e-06, "loss": 0.0251, "step": 2090 }, { "epoch": 0.8388256440982624, "grad_norm": 1.1952186822891235, "learning_rate": 7.719623502567029e-06, "loss": 0.0153, "step": 2100 }, { "epoch": 0.8428200519273018, "grad_norm": 2.101468086242676, "learning_rate": 7.705362236166572e-06, "loss": 0.023, "step": 2110 }, { "epoch": 0.8468144597563412, "grad_norm": 2.7876439094543457, "learning_rate": 7.691100969766115e-06, "loss": 0.0262, "step": 2120 }, { "epoch": 0.8508088675853804, "grad_norm": 2.597797155380249, "learning_rate": 7.676839703365659e-06, "loss": 0.0198, "step": 2130 }, { "epoch": 0.8548032754144198, "grad_norm": 2.360288619995117, "learning_rate": 7.662578436965204e-06, "loss": 0.0212, "step": 2140 }, { "epoch": 0.8587976832434592, "grad_norm": 2.3129231929779053, "learning_rate": 7.648317170564747e-06, "loss": 0.0218, "step": 2150 }, { "epoch": 0.8627920910724985, "grad_norm": 2.0236942768096924, "learning_rate": 7.63405590416429e-06, "loss": 0.0225, "step": 2160 }, { "epoch": 0.8667864989015378, "grad_norm": 2.5512468814849854, "learning_rate": 7.619794637763834e-06, "loss": 0.0236, "step": 2170 }, { "epoch": 0.8707809067305772, "grad_norm": 1.3719795942306519, "learning_rate": 7.6055333713633774e-06, "loss": 0.0252, "step": 2180 }, { "epoch": 0.8747753145596165, "grad_norm": 1.4252673387527466, "learning_rate": 7.5912721049629215e-06, "loss": 0.0244, "step": 2190 }, { "epoch": 0.8787697223886559, "grad_norm": 3.229334592819214, "learning_rate": 7.577010838562466e-06, "loss": 0.0229, "step": 2200 }, { "epoch": 0.8827641302176952, "grad_norm": 3.052751064300537, "learning_rate": 7.562749572162008e-06, "loss": 0.0261, "step": 2210 }, { "epoch": 0.8867585380467345, "grad_norm": 3.38997745513916, "learning_rate": 7.548488305761552e-06, "loss": 0.02, "step": 2220 }, { "epoch": 0.8907529458757739, "grad_norm": 2.800586223602295, "learning_rate": 7.534227039361096e-06, "loss": 0.0348, "step": 2230 }, { "epoch": 0.8947473537048133, "grad_norm": 2.3889737129211426, "learning_rate": 7.519965772960639e-06, "loss": 0.0265, "step": 2240 }, { "epoch": 0.8987417615338527, "grad_norm": 2.066007614135742, "learning_rate": 7.505704506560183e-06, "loss": 0.0269, "step": 2250 }, { "epoch": 0.9027361693628919, "grad_norm": 2.4386041164398193, "learning_rate": 7.491443240159727e-06, "loss": 0.0194, "step": 2260 }, { "epoch": 0.9067305771919313, "grad_norm": 2.8002700805664062, "learning_rate": 7.47718197375927e-06, "loss": 0.0217, "step": 2270 }, { "epoch": 0.9107249850209707, "grad_norm": 1.103632926940918, "learning_rate": 7.462920707358814e-06, "loss": 0.0255, "step": 2280 }, { "epoch": 0.91471939285001, "grad_norm": 2.646819829940796, "learning_rate": 7.448659440958358e-06, "loss": 0.0199, "step": 2290 }, { "epoch": 0.9187138006790493, "grad_norm": 4.851853370666504, "learning_rate": 7.434398174557901e-06, "loss": 0.0271, "step": 2300 }, { "epoch": 0.9227082085080887, "grad_norm": 1.8411966562271118, "learning_rate": 7.420136908157445e-06, "loss": 0.019, "step": 2310 }, { "epoch": 0.926702616337128, "grad_norm": 1.5107672214508057, "learning_rate": 7.405875641756989e-06, "loss": 0.0259, "step": 2320 }, { "epoch": 0.9306970241661674, "grad_norm": 2.267580986022949, "learning_rate": 7.391614375356533e-06, "loss": 0.0219, "step": 2330 }, { "epoch": 0.9346914319952067, "grad_norm": 3.101135730743408, "learning_rate": 7.377353108956075e-06, "loss": 0.0234, "step": 2340 }, { "epoch": 0.938685839824246, "grad_norm": 1.4610391855239868, "learning_rate": 7.3630918425556194e-06, "loss": 0.0201, "step": 2350 }, { "epoch": 0.9426802476532854, "grad_norm": 5.495433807373047, "learning_rate": 7.3488305761551635e-06, "loss": 0.0357, "step": 2360 }, { "epoch": 0.9466746554823248, "grad_norm": 2.515683650970459, "learning_rate": 7.334569309754707e-06, "loss": 0.017, "step": 2370 }, { "epoch": 0.9506690633113641, "grad_norm": 2.1554369926452637, "learning_rate": 7.320308043354251e-06, "loss": 0.0214, "step": 2380 }, { "epoch": 0.9546634711404034, "grad_norm": 3.1038148403167725, "learning_rate": 7.306046776953795e-06, "loss": 0.024, "step": 2390 }, { "epoch": 0.9586578789694428, "grad_norm": 4.446855068206787, "learning_rate": 7.291785510553337e-06, "loss": 0.0295, "step": 2400 }, { "epoch": 0.9626522867984821, "grad_norm": 2.9995310306549072, "learning_rate": 7.2775242441528815e-06, "loss": 0.0241, "step": 2410 }, { "epoch": 0.9666466946275215, "grad_norm": 2.6743619441986084, "learning_rate": 7.2632629777524256e-06, "loss": 0.0205, "step": 2420 }, { "epoch": 0.9706411024565608, "grad_norm": 2.39265513420105, "learning_rate": 7.249001711351969e-06, "loss": 0.0298, "step": 2430 }, { "epoch": 0.9746355102856001, "grad_norm": 1.3505173921585083, "learning_rate": 7.234740444951512e-06, "loss": 0.0195, "step": 2440 }, { "epoch": 0.9786299181146395, "grad_norm": 3.3783137798309326, "learning_rate": 7.220479178551056e-06, "loss": 0.0223, "step": 2450 }, { "epoch": 0.9826243259436789, "grad_norm": 1.7907217741012573, "learning_rate": 7.206217912150599e-06, "loss": 0.016, "step": 2460 }, { "epoch": 0.9866187337727182, "grad_norm": 1.9002418518066406, "learning_rate": 7.1919566457501435e-06, "loss": 0.0184, "step": 2470 }, { "epoch": 0.9906131416017575, "grad_norm": 1.7792202234268188, "learning_rate": 7.1776953793496876e-06, "loss": 0.0241, "step": 2480 }, { "epoch": 0.9946075494307969, "grad_norm": 2.2970614433288574, "learning_rate": 7.16343411294923e-06, "loss": 0.0202, "step": 2490 }, { "epoch": 0.9986019572598362, "grad_norm": 3.3695755004882812, "learning_rate": 7.149172846548774e-06, "loss": 0.0202, "step": 2500 }, { "epoch": 1.0023966446974235, "grad_norm": 2.2537529468536377, "learning_rate": 7.134911580148318e-06, "loss": 0.0107, "step": 2510 }, { "epoch": 1.006391052526463, "grad_norm": 1.0136545896530151, "learning_rate": 7.120650313747861e-06, "loss": 0.0059, "step": 2520 }, { "epoch": 1.0103854603555023, "grad_norm": 1.9403961896896362, "learning_rate": 7.1063890473474055e-06, "loss": 0.006, "step": 2530 }, { "epoch": 1.0143798681845417, "grad_norm": 2.3550891876220703, "learning_rate": 7.092127780946949e-06, "loss": 0.0091, "step": 2540 }, { "epoch": 1.018374276013581, "grad_norm": 1.9145513772964478, "learning_rate": 7.077866514546492e-06, "loss": 0.0093, "step": 2550 }, { "epoch": 1.0223686838426203, "grad_norm": 0.6784165501594543, "learning_rate": 7.063605248146036e-06, "loss": 0.0061, "step": 2560 }, { "epoch": 1.0263630916716597, "grad_norm": 2.786740779876709, "learning_rate": 7.04934398174558e-06, "loss": 0.008, "step": 2570 }, { "epoch": 1.030357499500699, "grad_norm": 1.3731578588485718, "learning_rate": 7.035082715345123e-06, "loss": 0.0065, "step": 2580 }, { "epoch": 1.0343519073297385, "grad_norm": 1.6200900077819824, "learning_rate": 7.020821448944667e-06, "loss": 0.0065, "step": 2590 }, { "epoch": 1.0383463151587777, "grad_norm": 0.493135929107666, "learning_rate": 7.006560182544211e-06, "loss": 0.0065, "step": 2600 }, { "epoch": 1.042340722987817, "grad_norm": 1.1139098405838013, "learning_rate": 6.992298916143754e-06, "loss": 0.0062, "step": 2610 }, { "epoch": 1.0463351308168565, "grad_norm": 1.2517800331115723, "learning_rate": 6.978037649743298e-06, "loss": 0.0059, "step": 2620 }, { "epoch": 1.0503295386458957, "grad_norm": 1.8630436658859253, "learning_rate": 6.963776383342841e-06, "loss": 0.0081, "step": 2630 }, { "epoch": 1.054323946474935, "grad_norm": 1.4949543476104736, "learning_rate": 6.949515116942385e-06, "loss": 0.0056, "step": 2640 }, { "epoch": 1.0583183543039745, "grad_norm": 1.327675700187683, "learning_rate": 6.935253850541929e-06, "loss": 0.0104, "step": 2650 }, { "epoch": 1.0623127621330137, "grad_norm": 2.0877177715301514, "learning_rate": 6.920992584141473e-06, "loss": 0.0077, "step": 2660 }, { "epoch": 1.0663071699620532, "grad_norm": 3.7482008934020996, "learning_rate": 6.906731317741015e-06, "loss": 0.0075, "step": 2670 }, { "epoch": 1.0703015777910925, "grad_norm": 1.980429768562317, "learning_rate": 6.892470051340559e-06, "loss": 0.0079, "step": 2680 }, { "epoch": 1.0742959856201317, "grad_norm": 0.3283044397830963, "learning_rate": 6.878208784940103e-06, "loss": 0.0091, "step": 2690 }, { "epoch": 1.0782903934491712, "grad_norm": 0.40156498551368713, "learning_rate": 6.863947518539647e-06, "loss": 0.0077, "step": 2700 }, { "epoch": 1.0822848012782105, "grad_norm": 0.9203258752822876, "learning_rate": 6.849686252139191e-06, "loss": 0.0071, "step": 2710 }, { "epoch": 1.08627920910725, "grad_norm": 0.5740295648574829, "learning_rate": 6.835424985738735e-06, "loss": 0.0071, "step": 2720 }, { "epoch": 1.0902736169362892, "grad_norm": 0.5572212934494019, "learning_rate": 6.821163719338277e-06, "loss": 0.0061, "step": 2730 }, { "epoch": 1.0942680247653285, "grad_norm": 1.8993756771087646, "learning_rate": 6.806902452937821e-06, "loss": 0.0078, "step": 2740 }, { "epoch": 1.098262432594368, "grad_norm": 1.3599531650543213, "learning_rate": 6.792641186537365e-06, "loss": 0.0081, "step": 2750 }, { "epoch": 1.1022568404234072, "grad_norm": 1.9560530185699463, "learning_rate": 6.778379920136909e-06, "loss": 0.0092, "step": 2760 }, { "epoch": 1.1062512482524465, "grad_norm": 1.3307870626449585, "learning_rate": 6.764118653736452e-06, "loss": 0.0091, "step": 2770 }, { "epoch": 1.110245656081486, "grad_norm": 2.2975058555603027, "learning_rate": 6.749857387335996e-06, "loss": 0.0065, "step": 2780 }, { "epoch": 1.1142400639105252, "grad_norm": 1.8662095069885254, "learning_rate": 6.735596120935539e-06, "loss": 0.0065, "step": 2790 }, { "epoch": 1.1182344717395647, "grad_norm": 1.517377495765686, "learning_rate": 6.721334854535083e-06, "loss": 0.0067, "step": 2800 }, { "epoch": 1.122228879568604, "grad_norm": 1.8435981273651123, "learning_rate": 6.7070735881346274e-06, "loss": 0.0082, "step": 2810 }, { "epoch": 1.1262232873976432, "grad_norm": 0.9961625337600708, "learning_rate": 6.69281232173417e-06, "loss": 0.0087, "step": 2820 }, { "epoch": 1.1302176952266827, "grad_norm": 1.4639012813568115, "learning_rate": 6.678551055333714e-06, "loss": 0.008, "step": 2830 }, { "epoch": 1.134212103055722, "grad_norm": 2.6728835105895996, "learning_rate": 6.664289788933258e-06, "loss": 0.01, "step": 2840 }, { "epoch": 1.1382065108847614, "grad_norm": 2.244205951690674, "learning_rate": 6.650028522532801e-06, "loss": 0.0055, "step": 2850 }, { "epoch": 1.1422009187138007, "grad_norm": 1.568909764289856, "learning_rate": 6.635767256132345e-06, "loss": 0.008, "step": 2860 }, { "epoch": 1.14619532654284, "grad_norm": 1.443452000617981, "learning_rate": 6.621505989731889e-06, "loss": 0.0079, "step": 2870 }, { "epoch": 1.1501897343718794, "grad_norm": 0.7626383900642395, "learning_rate": 6.607244723331432e-06, "loss": 0.0065, "step": 2880 }, { "epoch": 1.1541841422009187, "grad_norm": 0.9980650544166565, "learning_rate": 6.592983456930976e-06, "loss": 0.008, "step": 2890 }, { "epoch": 1.158178550029958, "grad_norm": 2.2537312507629395, "learning_rate": 6.57872219053052e-06, "loss": 0.0055, "step": 2900 }, { "epoch": 1.1621729578589974, "grad_norm": 1.9546359777450562, "learning_rate": 6.5644609241300625e-06, "loss": 0.0063, "step": 2910 }, { "epoch": 1.1661673656880367, "grad_norm": 1.7826281785964966, "learning_rate": 6.5501996577296066e-06, "loss": 0.0094, "step": 2920 }, { "epoch": 1.1701617735170762, "grad_norm": 1.5453565120697021, "learning_rate": 6.535938391329151e-06, "loss": 0.007, "step": 2930 }, { "epoch": 1.1741561813461154, "grad_norm": 2.085813045501709, "learning_rate": 6.521677124928694e-06, "loss": 0.0071, "step": 2940 }, { "epoch": 1.1781505891751547, "grad_norm": 1.3330625295639038, "learning_rate": 6.507415858528238e-06, "loss": 0.0076, "step": 2950 }, { "epoch": 1.1821449970041942, "grad_norm": 1.4080004692077637, "learning_rate": 6.493154592127781e-06, "loss": 0.0097, "step": 2960 }, { "epoch": 1.1861394048332334, "grad_norm": 1.0290985107421875, "learning_rate": 6.478893325727325e-06, "loss": 0.0113, "step": 2970 }, { "epoch": 1.190133812662273, "grad_norm": 1.6050559282302856, "learning_rate": 6.464632059326869e-06, "loss": 0.0091, "step": 2980 }, { "epoch": 1.1941282204913122, "grad_norm": 2.549539566040039, "learning_rate": 6.450370792926413e-06, "loss": 0.0089, "step": 2990 }, { "epoch": 1.1981226283203514, "grad_norm": 1.9209243059158325, "learning_rate": 6.436109526525957e-06, "loss": 0.0063, "step": 3000 }, { "epoch": 1.202117036149391, "grad_norm": 1.004465103149414, "learning_rate": 6.421848260125499e-06, "loss": 0.0071, "step": 3010 }, { "epoch": 1.2061114439784302, "grad_norm": 0.8008792996406555, "learning_rate": 6.407586993725043e-06, "loss": 0.0058, "step": 3020 }, { "epoch": 1.2101058518074694, "grad_norm": 0.7581857442855835, "learning_rate": 6.393325727324587e-06, "loss": 0.0106, "step": 3030 }, { "epoch": 1.214100259636509, "grad_norm": 0.4473499059677124, "learning_rate": 6.379064460924131e-06, "loss": 0.0065, "step": 3040 }, { "epoch": 1.2180946674655482, "grad_norm": 0.9399948120117188, "learning_rate": 6.364803194523675e-06, "loss": 0.006, "step": 3050 }, { "epoch": 1.2220890752945877, "grad_norm": 1.572412133216858, "learning_rate": 6.350541928123218e-06, "loss": 0.0093, "step": 3060 }, { "epoch": 1.226083483123627, "grad_norm": 2.0445048809051514, "learning_rate": 6.336280661722761e-06, "loss": 0.0061, "step": 3070 }, { "epoch": 1.2300778909526662, "grad_norm": 1.4518450498580933, "learning_rate": 6.322019395322305e-06, "loss": 0.0063, "step": 3080 }, { "epoch": 1.2340722987817057, "grad_norm": 1.0120337009429932, "learning_rate": 6.307758128921849e-06, "loss": 0.0116, "step": 3090 }, { "epoch": 1.238066706610745, "grad_norm": 0.40242263674736023, "learning_rate": 6.293496862521392e-06, "loss": 0.009, "step": 3100 }, { "epoch": 1.2420611144397844, "grad_norm": 1.763005256652832, "learning_rate": 6.279235596120936e-06, "loss": 0.0107, "step": 3110 }, { "epoch": 1.2460555222688237, "grad_norm": 1.8232554197311401, "learning_rate": 6.26497432972048e-06, "loss": 0.0071, "step": 3120 }, { "epoch": 1.250049930097863, "grad_norm": 2.130310297012329, "learning_rate": 6.250713063320023e-06, "loss": 0.0049, "step": 3130 }, { "epoch": 1.2540443379269024, "grad_norm": 1.1470189094543457, "learning_rate": 6.236451796919567e-06, "loss": 0.0082, "step": 3140 }, { "epoch": 1.2580387457559417, "grad_norm": 0.8706045746803284, "learning_rate": 6.222190530519111e-06, "loss": 0.008, "step": 3150 }, { "epoch": 1.262033153584981, "grad_norm": 2.534973382949829, "learning_rate": 6.207929264118654e-06, "loss": 0.0081, "step": 3160 }, { "epoch": 1.2660275614140204, "grad_norm": 1.874556064605713, "learning_rate": 6.193667997718198e-06, "loss": 0.0048, "step": 3170 }, { "epoch": 1.2700219692430597, "grad_norm": 2.186999559402466, "learning_rate": 6.179406731317742e-06, "loss": 0.0076, "step": 3180 }, { "epoch": 1.2740163770720991, "grad_norm": 2.7856943607330322, "learning_rate": 6.165145464917285e-06, "loss": 0.0077, "step": 3190 }, { "epoch": 1.2780107849011384, "grad_norm": 1.1540160179138184, "learning_rate": 6.1508841985168285e-06, "loss": 0.0111, "step": 3200 }, { "epoch": 1.2820051927301779, "grad_norm": 1.7507003545761108, "learning_rate": 6.136622932116373e-06, "loss": 0.0063, "step": 3210 }, { "epoch": 1.2859996005592171, "grad_norm": 0.28601911664009094, "learning_rate": 6.122361665715916e-06, "loss": 0.0071, "step": 3220 }, { "epoch": 1.2899940083882564, "grad_norm": 2.5565030574798584, "learning_rate": 6.10810039931546e-06, "loss": 0.0095, "step": 3230 }, { "epoch": 1.2939884162172959, "grad_norm": 1.0041239261627197, "learning_rate": 6.093839132915004e-06, "loss": 0.0066, "step": 3240 }, { "epoch": 1.2979828240463351, "grad_norm": 0.6574738025665283, "learning_rate": 6.0795778665145464e-06, "loss": 0.0076, "step": 3250 }, { "epoch": 1.3019772318753744, "grad_norm": 2.0627009868621826, "learning_rate": 6.0653166001140905e-06, "loss": 0.007, "step": 3260 }, { "epoch": 1.3059716397044139, "grad_norm": 1.0173168182373047, "learning_rate": 6.051055333713635e-06, "loss": 0.0066, "step": 3270 }, { "epoch": 1.3099660475334531, "grad_norm": 0.653717577457428, "learning_rate": 6.036794067313178e-06, "loss": 0.0091, "step": 3280 }, { "epoch": 1.3139604553624924, "grad_norm": 2.0671188831329346, "learning_rate": 6.022532800912721e-06, "loss": 0.007, "step": 3290 }, { "epoch": 1.3179548631915319, "grad_norm": 1.411763072013855, "learning_rate": 6.008271534512265e-06, "loss": 0.0068, "step": 3300 }, { "epoch": 1.3219492710205711, "grad_norm": 0.860016942024231, "learning_rate": 5.9940102681118085e-06, "loss": 0.0058, "step": 3310 }, { "epoch": 1.3259436788496106, "grad_norm": 0.9128891229629517, "learning_rate": 5.9797490017113525e-06, "loss": 0.0091, "step": 3320 }, { "epoch": 1.3299380866786499, "grad_norm": 1.0276684761047363, "learning_rate": 5.965487735310897e-06, "loss": 0.0107, "step": 3330 }, { "epoch": 1.3339324945076894, "grad_norm": 1.8556329011917114, "learning_rate": 5.951226468910439e-06, "loss": 0.0057, "step": 3340 }, { "epoch": 1.3379269023367286, "grad_norm": 0.9202171564102173, "learning_rate": 5.936965202509983e-06, "loss": 0.0079, "step": 3350 }, { "epoch": 1.3419213101657679, "grad_norm": 0.5609256029129028, "learning_rate": 5.922703936109527e-06, "loss": 0.009, "step": 3360 }, { "epoch": 1.3459157179948074, "grad_norm": 1.7083600759506226, "learning_rate": 5.9084426697090705e-06, "loss": 0.0078, "step": 3370 }, { "epoch": 1.3499101258238466, "grad_norm": 0.5730917453765869, "learning_rate": 5.8941814033086146e-06, "loss": 0.0075, "step": 3380 }, { "epoch": 1.3539045336528859, "grad_norm": 0.7475340366363525, "learning_rate": 5.879920136908158e-06, "loss": 0.0086, "step": 3390 }, { "epoch": 1.3578989414819254, "grad_norm": 1.2629673480987549, "learning_rate": 5.865658870507701e-06, "loss": 0.0066, "step": 3400 }, { "epoch": 1.3618933493109646, "grad_norm": 1.5995068550109863, "learning_rate": 5.851397604107245e-06, "loss": 0.0077, "step": 3410 }, { "epoch": 1.3658877571400039, "grad_norm": 0.6721827983856201, "learning_rate": 5.837136337706789e-06, "loss": 0.0056, "step": 3420 }, { "epoch": 1.3698821649690434, "grad_norm": 1.5127277374267578, "learning_rate": 5.822875071306332e-06, "loss": 0.007, "step": 3430 }, { "epoch": 1.3738765727980826, "grad_norm": 0.8223251104354858, "learning_rate": 5.808613804905876e-06, "loss": 0.0055, "step": 3440 }, { "epoch": 1.377870980627122, "grad_norm": 2.7876274585723877, "learning_rate": 5.79435253850542e-06, "loss": 0.0086, "step": 3450 }, { "epoch": 1.3818653884561614, "grad_norm": 0.45988646149635315, "learning_rate": 5.780091272104963e-06, "loss": 0.0033, "step": 3460 }, { "epoch": 1.3858597962852008, "grad_norm": 0.7772718071937561, "learning_rate": 5.765830005704507e-06, "loss": 0.009, "step": 3470 }, { "epoch": 1.38985420411424, "grad_norm": 1.917909026145935, "learning_rate": 5.751568739304051e-06, "loss": 0.0063, "step": 3480 }, { "epoch": 1.3938486119432794, "grad_norm": 1.467574954032898, "learning_rate": 5.737307472903594e-06, "loss": 0.0057, "step": 3490 }, { "epoch": 1.3978430197723188, "grad_norm": 2.2506048679351807, "learning_rate": 5.723046206503138e-06, "loss": 0.0072, "step": 3500 }, { "epoch": 1.401837427601358, "grad_norm": 2.0237884521484375, "learning_rate": 5.708784940102682e-06, "loss": 0.0099, "step": 3510 }, { "epoch": 1.4058318354303974, "grad_norm": 1.5853831768035889, "learning_rate": 5.694523673702225e-06, "loss": 0.008, "step": 3520 }, { "epoch": 1.4098262432594368, "grad_norm": 0.22210288047790527, "learning_rate": 5.680262407301768e-06, "loss": 0.007, "step": 3530 }, { "epoch": 1.413820651088476, "grad_norm": 0.654780924320221, "learning_rate": 5.6660011409013125e-06, "loss": 0.0054, "step": 3540 }, { "epoch": 1.4178150589175154, "grad_norm": 1.5929639339447021, "learning_rate": 5.651739874500856e-06, "loss": 0.0077, "step": 3550 }, { "epoch": 1.4218094667465548, "grad_norm": 1.2836332321166992, "learning_rate": 5.6374786081004e-06, "loss": 0.0057, "step": 3560 }, { "epoch": 1.425803874575594, "grad_norm": 2.508904457092285, "learning_rate": 5.623217341699944e-06, "loss": 0.0064, "step": 3570 }, { "epoch": 1.4297982824046336, "grad_norm": 1.0835081338882446, "learning_rate": 5.608956075299488e-06, "loss": 0.0057, "step": 3580 }, { "epoch": 1.4337926902336728, "grad_norm": 0.40751364827156067, "learning_rate": 5.59469480889903e-06, "loss": 0.0053, "step": 3590 }, { "epoch": 1.4377870980627123, "grad_norm": 1.3708760738372803, "learning_rate": 5.5804335424985745e-06, "loss": 0.0108, "step": 3600 }, { "epoch": 1.4417815058917516, "grad_norm": 1.1257152557373047, "learning_rate": 5.566172276098119e-06, "loss": 0.007, "step": 3610 }, { "epoch": 1.4457759137207908, "grad_norm": 1.2688618898391724, "learning_rate": 5.551911009697661e-06, "loss": 0.0048, "step": 3620 }, { "epoch": 1.4497703215498303, "grad_norm": 0.5364936590194702, "learning_rate": 5.537649743297205e-06, "loss": 0.0108, "step": 3630 }, { "epoch": 1.4537647293788696, "grad_norm": 1.7911590337753296, "learning_rate": 5.523388476896749e-06, "loss": 0.0058, "step": 3640 }, { "epoch": 1.4577591372079088, "grad_norm": 2.314695119857788, "learning_rate": 5.509127210496292e-06, "loss": 0.0079, "step": 3650 }, { "epoch": 1.4617535450369483, "grad_norm": 1.7157478332519531, "learning_rate": 5.4948659440958365e-06, "loss": 0.0075, "step": 3660 }, { "epoch": 1.4657479528659876, "grad_norm": 1.470686435699463, "learning_rate": 5.480604677695381e-06, "loss": 0.008, "step": 3670 }, { "epoch": 1.4697423606950268, "grad_norm": 0.8573768734931946, "learning_rate": 5.466343411294923e-06, "loss": 0.0173, "step": 3680 }, { "epoch": 1.4737367685240663, "grad_norm": 2.28192138671875, "learning_rate": 5.452082144894467e-06, "loss": 0.0107, "step": 3690 }, { "epoch": 1.4777311763531056, "grad_norm": 1.3094779253005981, "learning_rate": 5.437820878494011e-06, "loss": 0.01, "step": 3700 }, { "epoch": 1.481725584182145, "grad_norm": 0.6152220964431763, "learning_rate": 5.4235596120935544e-06, "loss": 0.0078, "step": 3710 }, { "epoch": 1.4857199920111843, "grad_norm": 1.8408433198928833, "learning_rate": 5.409298345693098e-06, "loss": 0.0075, "step": 3720 }, { "epoch": 1.4897143998402238, "grad_norm": 2.0556797981262207, "learning_rate": 5.395037079292642e-06, "loss": 0.0094, "step": 3730 }, { "epoch": 1.493708807669263, "grad_norm": 1.437679648399353, "learning_rate": 5.380775812892185e-06, "loss": 0.0075, "step": 3740 }, { "epoch": 1.4977032154983023, "grad_norm": 1.2906135320663452, "learning_rate": 5.366514546491729e-06, "loss": 0.006, "step": 3750 }, { "epoch": 1.5016976233273418, "grad_norm": 2.008298397064209, "learning_rate": 5.352253280091273e-06, "loss": 0.0063, "step": 3760 }, { "epoch": 1.505692031156381, "grad_norm": 1.9057350158691406, "learning_rate": 5.337992013690816e-06, "loss": 0.0092, "step": 3770 }, { "epoch": 1.5096864389854203, "grad_norm": 0.8335756063461304, "learning_rate": 5.32373074729036e-06, "loss": 0.0072, "step": 3780 }, { "epoch": 1.5136808468144598, "grad_norm": 0.6777927875518799, "learning_rate": 5.309469480889904e-06, "loss": 0.0048, "step": 3790 }, { "epoch": 1.517675254643499, "grad_norm": 0.7423845529556274, "learning_rate": 5.295208214489447e-06, "loss": 0.006, "step": 3800 }, { "epoch": 1.5216696624725383, "grad_norm": 0.7639596462249756, "learning_rate": 5.280946948088991e-06, "loss": 0.0064, "step": 3810 }, { "epoch": 1.5256640703015778, "grad_norm": 0.5224437713623047, "learning_rate": 5.266685681688534e-06, "loss": 0.0051, "step": 3820 }, { "epoch": 1.5296584781306173, "grad_norm": 2.200313091278076, "learning_rate": 5.252424415288078e-06, "loss": 0.0075, "step": 3830 }, { "epoch": 1.5336528859596563, "grad_norm": 0.9855785369873047, "learning_rate": 5.238163148887622e-06, "loss": 0.0045, "step": 3840 }, { "epoch": 1.5376472937886958, "grad_norm": 0.26075270771980286, "learning_rate": 5.223901882487166e-06, "loss": 0.0074, "step": 3850 }, { "epoch": 1.5416417016177353, "grad_norm": 0.7130741477012634, "learning_rate": 5.209640616086708e-06, "loss": 0.0071, "step": 3860 }, { "epoch": 1.5456361094467745, "grad_norm": 0.7751367688179016, "learning_rate": 5.195379349686252e-06, "loss": 0.0062, "step": 3870 }, { "epoch": 1.5496305172758138, "grad_norm": 0.3975371718406677, "learning_rate": 5.1811180832857964e-06, "loss": 0.0046, "step": 3880 }, { "epoch": 1.5536249251048533, "grad_norm": 0.9797967672348022, "learning_rate": 5.16685681688534e-06, "loss": 0.0051, "step": 3890 }, { "epoch": 1.5576193329338925, "grad_norm": 1.2865179777145386, "learning_rate": 5.152595550484884e-06, "loss": 0.0063, "step": 3900 }, { "epoch": 1.5616137407629318, "grad_norm": 0.38849377632141113, "learning_rate": 5.138334284084428e-06, "loss": 0.006, "step": 3910 }, { "epoch": 1.5656081485919713, "grad_norm": 1.3372318744659424, "learning_rate": 5.12407301768397e-06, "loss": 0.0089, "step": 3920 }, { "epoch": 1.5696025564210105, "grad_norm": 2.214596748352051, "learning_rate": 5.109811751283514e-06, "loss": 0.0072, "step": 3930 }, { "epoch": 1.5735969642500498, "grad_norm": 1.0024957656860352, "learning_rate": 5.0955504848830584e-06, "loss": 0.0071, "step": 3940 }, { "epoch": 1.5775913720790893, "grad_norm": 1.6112749576568604, "learning_rate": 5.081289218482601e-06, "loss": 0.0079, "step": 3950 }, { "epoch": 1.5815857799081288, "grad_norm": 1.7288084030151367, "learning_rate": 5.067027952082145e-06, "loss": 0.0059, "step": 3960 }, { "epoch": 1.5855801877371678, "grad_norm": 1.181510329246521, "learning_rate": 5.052766685681689e-06, "loss": 0.0063, "step": 3970 }, { "epoch": 1.5895745955662073, "grad_norm": 0.6476355195045471, "learning_rate": 5.038505419281232e-06, "loss": 0.0095, "step": 3980 }, { "epoch": 1.5935690033952468, "grad_norm": 2.223113775253296, "learning_rate": 5.024244152880776e-06, "loss": 0.008, "step": 3990 }, { "epoch": 1.597563411224286, "grad_norm": 0.8112757802009583, "learning_rate": 5.0099828864803205e-06, "loss": 0.0087, "step": 4000 }, { "epoch": 1.597563411224286, "eval_loss": 0.008109057322144508, "eval_runtime": 7231.5664, "eval_samples_per_second": 2.769, "eval_steps_per_second": 0.346, "eval_wer": 0.6836733584774409, "step": 4000 } ], "logging_steps": 10, "max_steps": 7512, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 2000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 3.693085084975104e+19, "train_batch_size": 16, "trial_name": null, "trial_params": null }