diff --git "a/checkpoint-6000/trainer_state.json" "b/checkpoint-6000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-6000/trainer_state.json"
@@ -0,0 +1,4261 @@
+{
+  "best_global_step": 6000,
+  "best_metric": 0.3684662456659022,
+  "best_model_checkpoint": "./SALAMA_NEW6/checkpoint-6000",
+  "epoch": 2.396245256640703,
+  "eval_steps": 2000,
+  "global_step": 6000,
+  "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
+    },
+    {
+      "epoch": 1.6015578190533253,
+      "grad_norm": 1.0202847719192505,
+      "learning_rate": 4.995721620079864e-06,
+      "loss": 0.005,
+      "step": 4010
+    },
+    {
+      "epoch": 1.6055522268823648,
+      "grad_norm": 0.7995850443840027,
+      "learning_rate": 4.981460353679407e-06,
+      "loss": 0.0059,
+      "step": 4020
+    },
+    {
+      "epoch": 1.609546634711404,
+      "grad_norm": 0.9624062180519104,
+      "learning_rate": 4.96719908727895e-06,
+      "loss": 0.0055,
+      "step": 4030
+    },
+    {
+      "epoch": 1.6135410425404433,
+      "grad_norm": 1.1462153196334839,
+      "learning_rate": 4.952937820878494e-06,
+      "loss": 0.0073,
+      "step": 4040
+    },
+    {
+      "epoch": 1.6175354503694828,
+      "grad_norm": 0.4766294062137604,
+      "learning_rate": 4.9386765544780376e-06,
+      "loss": 0.0056,
+      "step": 4050
+    },
+    {
+      "epoch": 1.621529858198522,
+      "grad_norm": 0.4788503646850586,
+      "learning_rate": 4.924415288077582e-06,
+      "loss": 0.0067,
+      "step": 4060
+    },
+    {
+      "epoch": 1.6255242660275613,
+      "grad_norm": 2.1229958534240723,
+      "learning_rate": 4.910154021677126e-06,
+      "loss": 0.006,
+      "step": 4070
+    },
+    {
+      "epoch": 1.6295186738566008,
+      "grad_norm": 0.9424548745155334,
+      "learning_rate": 4.895892755276669e-06,
+      "loss": 0.0077,
+      "step": 4080
+    },
+    {
+      "epoch": 1.6335130816856402,
+      "grad_norm": 1.8259445428848267,
+      "learning_rate": 4.881631488876213e-06,
+      "loss": 0.0057,
+      "step": 4090
+    },
+    {
+      "epoch": 1.6375074895146795,
+      "grad_norm": 1.358744502067566,
+      "learning_rate": 4.867370222475756e-06,
+      "loss": 0.0051,
+      "step": 4100
+    },
+    {
+      "epoch": 1.6415018973437188,
+      "grad_norm": 1.1335296630859375,
+      "learning_rate": 4.8531089560753e-06,
+      "loss": 0.0072,
+      "step": 4110
+    },
+    {
+      "epoch": 1.6454963051727582,
+      "grad_norm": 0.8320634365081787,
+      "learning_rate": 4.838847689674844e-06,
+      "loss": 0.0057,
+      "step": 4120
+    },
+    {
+      "epoch": 1.6494907130017975,
+      "grad_norm": 0.5541989207267761,
+      "learning_rate": 4.824586423274387e-06,
+      "loss": 0.0065,
+      "step": 4130
+    },
+    {
+      "epoch": 1.6534851208308368,
+      "grad_norm": 0.695227324962616,
+      "learning_rate": 4.810325156873931e-06,
+      "loss": 0.0065,
+      "step": 4140
+    },
+    {
+      "epoch": 1.6574795286598762,
+      "grad_norm": 1.260872483253479,
+      "learning_rate": 4.796063890473474e-06,
+      "loss": 0.0052,
+      "step": 4150
+    },
+    {
+      "epoch": 1.6614739364889155,
+      "grad_norm": 1.926358938217163,
+      "learning_rate": 4.781802624073018e-06,
+      "loss": 0.0047,
+      "step": 4160
+    },
+    {
+      "epoch": 1.6654683443179548,
+      "grad_norm": 2.476741075515747,
+      "learning_rate": 4.767541357672562e-06,
+      "loss": 0.0088,
+      "step": 4170
+    },
+    {
+      "epoch": 1.6694627521469942,
+      "grad_norm": 1.842189073562622,
+      "learning_rate": 4.753280091272106e-06,
+      "loss": 0.0063,
+      "step": 4180
+    },
+    {
+      "epoch": 1.6734571599760335,
+      "grad_norm": 0.9236751198768616,
+      "learning_rate": 4.739018824871649e-06,
+      "loss": 0.0061,
+      "step": 4190
+    },
+    {
+      "epoch": 1.6774515678050728,
+      "grad_norm": 0.2939533591270447,
+      "learning_rate": 4.724757558471192e-06,
+      "loss": 0.0041,
+      "step": 4200
+    },
+    {
+      "epoch": 1.6814459756341122,
+      "grad_norm": 1.7504230737686157,
+      "learning_rate": 4.710496292070736e-06,
+      "loss": 0.0072,
+      "step": 4210
+    },
+    {
+      "epoch": 1.6854403834631517,
+      "grad_norm": 0.4307522773742676,
+      "learning_rate": 4.6962350256702795e-06,
+      "loss": 0.0058,
+      "step": 4220
+    },
+    {
+      "epoch": 1.689434791292191,
+      "grad_norm": 1.4100841283798218,
+      "learning_rate": 4.681973759269824e-06,
+      "loss": 0.0088,
+      "step": 4230
+    },
+    {
+      "epoch": 1.6934291991212302,
+      "grad_norm": 1.3432307243347168,
+      "learning_rate": 4.667712492869368e-06,
+      "loss": 0.0042,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6974236069502697,
+      "grad_norm": 2.323749303817749,
+      "learning_rate": 4.653451226468911e-06,
+      "loss": 0.0064,
+      "step": 4250
+    },
+    {
+      "epoch": 1.701418014779309,
+      "grad_norm": 0.4890572726726532,
+      "learning_rate": 4.639189960068454e-06,
+      "loss": 0.0067,
+      "step": 4260
+    },
+    {
+      "epoch": 1.7054124226083482,
+      "grad_norm": 1.9534852504730225,
+      "learning_rate": 4.624928693667998e-06,
+      "loss": 0.0071,
+      "step": 4270
+    },
+    {
+      "epoch": 1.7094068304373877,
+      "grad_norm": 1.172839641571045,
+      "learning_rate": 4.6106674272675416e-06,
+      "loss": 0.0043,
+      "step": 4280
+    },
+    {
+      "epoch": 1.713401238266427,
+      "grad_norm": 1.0569114685058594,
+      "learning_rate": 4.596406160867085e-06,
+      "loss": 0.005,
+      "step": 4290
+    },
+    {
+      "epoch": 1.7173956460954662,
+      "grad_norm": 1.1828703880310059,
+      "learning_rate": 4.582144894466629e-06,
+      "loss": 0.0057,
+      "step": 4300
+    },
+    {
+      "epoch": 1.7213900539245057,
+      "grad_norm": 1.4898252487182617,
+      "learning_rate": 4.567883628066173e-06,
+      "loss": 0.0051,
+      "step": 4310
+    },
+    {
+      "epoch": 1.725384461753545,
+      "grad_norm": 1.2435832023620605,
+      "learning_rate": 4.553622361665716e-06,
+      "loss": 0.0068,
+      "step": 4320
+    },
+    {
+      "epoch": 1.7293788695825842,
+      "grad_norm": 0.9595640301704407,
+      "learning_rate": 4.53936109526526e-06,
+      "loss": 0.0078,
+      "step": 4330
+    },
+    {
+      "epoch": 1.7333732774116237,
+      "grad_norm": 0.5236431360244751,
+      "learning_rate": 4.525099828864804e-06,
+      "loss": 0.0047,
+      "step": 4340
+    },
+    {
+      "epoch": 1.7373676852406632,
+      "grad_norm": 0.7530146837234497,
+      "learning_rate": 4.510838562464347e-06,
+      "loss": 0.0059,
+      "step": 4350
+    },
+    {
+      "epoch": 1.7413620930697025,
+      "grad_norm": 0.488861620426178,
+      "learning_rate": 4.496577296063891e-06,
+      "loss": 0.0054,
+      "step": 4360
+    },
+    {
+      "epoch": 1.7453565008987417,
+      "grad_norm": 2.9665844440460205,
+      "learning_rate": 4.482316029663434e-06,
+      "loss": 0.0063,
+      "step": 4370
+    },
+    {
+      "epoch": 1.7493509087277812,
+      "grad_norm": 1.4653515815734863,
+      "learning_rate": 4.468054763262978e-06,
+      "loss": 0.0049,
+      "step": 4380
+    },
+    {
+      "epoch": 1.7533453165568205,
+      "grad_norm": 1.2187210321426392,
+      "learning_rate": 4.4537934968625215e-06,
+      "loss": 0.0057,
+      "step": 4390
+    },
+    {
+      "epoch": 1.7573397243858597,
+      "grad_norm": 0.3649451732635498,
+      "learning_rate": 4.439532230462066e-06,
+      "loss": 0.0051,
+      "step": 4400
+    },
+    {
+      "epoch": 1.7613341322148992,
+      "grad_norm": 2.18152117729187,
+      "learning_rate": 4.425270964061609e-06,
+      "loss": 0.0076,
+      "step": 4410
+    },
+    {
+      "epoch": 1.7653285400439385,
+      "grad_norm": 1.4268304109573364,
+      "learning_rate": 4.411009697661153e-06,
+      "loss": 0.0045,
+      "step": 4420
+    },
+    {
+      "epoch": 1.7693229478729777,
+      "grad_norm": 1.4249440431594849,
+      "learning_rate": 4.396748431260696e-06,
+      "loss": 0.0052,
+      "step": 4430
+    },
+    {
+      "epoch": 1.7733173557020172,
+      "grad_norm": 1.672342300415039,
+      "learning_rate": 4.38248716486024e-06,
+      "loss": 0.0108,
+      "step": 4440
+    },
+    {
+      "epoch": 1.7773117635310565,
+      "grad_norm": 0.37666410207748413,
+      "learning_rate": 4.3682258984597835e-06,
+      "loss": 0.0041,
+      "step": 4450
+    },
+    {
+      "epoch": 1.7813061713600957,
+      "grad_norm": 4.191842079162598,
+      "learning_rate": 4.353964632059327e-06,
+      "loss": 0.0045,
+      "step": 4460
+    },
+    {
+      "epoch": 1.7853005791891352,
+      "grad_norm": 2.393310785293579,
+      "learning_rate": 4.339703365658871e-06,
+      "loss": 0.0069,
+      "step": 4470
+    },
+    {
+      "epoch": 1.7892949870181747,
+      "grad_norm": 0.8967064023017883,
+      "learning_rate": 4.325442099258414e-06,
+      "loss": 0.0074,
+      "step": 4480
+    },
+    {
+      "epoch": 1.793289394847214,
+      "grad_norm": 0.6009141802787781,
+      "learning_rate": 4.311180832857958e-06,
+      "loss": 0.003,
+      "step": 4490
+    },
+    {
+      "epoch": 1.7972838026762532,
+      "grad_norm": 2.5212817192077637,
+      "learning_rate": 4.296919566457502e-06,
+      "loss": 0.0054,
+      "step": 4500
+    },
+    {
+      "epoch": 1.8012782105052927,
+      "grad_norm": 1.5466055870056152,
+      "learning_rate": 4.2826583000570456e-06,
+      "loss": 0.0063,
+      "step": 4510
+    },
+    {
+      "epoch": 1.805272618334332,
+      "grad_norm": 0.9209232926368713,
+      "learning_rate": 4.268397033656589e-06,
+      "loss": 0.0062,
+      "step": 4520
+    },
+    {
+      "epoch": 1.8092670261633712,
+      "grad_norm": 0.7686867713928223,
+      "learning_rate": 4.254135767256133e-06,
+      "loss": 0.0073,
+      "step": 4530
+    },
+    {
+      "epoch": 1.8132614339924107,
+      "grad_norm": 1.8461114168167114,
+      "learning_rate": 4.239874500855676e-06,
+      "loss": 0.0072,
+      "step": 4540
+    },
+    {
+      "epoch": 1.81725584182145,
+      "grad_norm": 0.325752854347229,
+      "learning_rate": 4.225613234455219e-06,
+      "loss": 0.005,
+      "step": 4550
+    },
+    {
+      "epoch": 1.8212502496504892,
+      "grad_norm": 0.5899893641471863,
+      "learning_rate": 4.2113519680547635e-06,
+      "loss": 0.0103,
+      "step": 4560
+    },
+    {
+      "epoch": 1.8252446574795287,
+      "grad_norm": 1.2786051034927368,
+      "learning_rate": 4.197090701654308e-06,
+      "loss": 0.0081,
+      "step": 4570
+    },
+    {
+      "epoch": 1.8292390653085682,
+      "grad_norm": 0.5355713963508606,
+      "learning_rate": 4.182829435253851e-06,
+      "loss": 0.0065,
+      "step": 4580
+    },
+    {
+      "epoch": 1.8332334731376072,
+      "grad_norm": 0.6819378137588501,
+      "learning_rate": 4.168568168853395e-06,
+      "loss": 0.0056,
+      "step": 4590
+    },
+    {
+      "epoch": 1.8372278809666467,
+      "grad_norm": 1.0736881494522095,
+      "learning_rate": 4.154306902452938e-06,
+      "loss": 0.0067,
+      "step": 4600
+    },
+    {
+      "epoch": 1.8412222887956862,
+      "grad_norm": 1.210741639137268,
+      "learning_rate": 4.1400456360524814e-06,
+      "loss": 0.0051,
+      "step": 4610
+    },
+    {
+      "epoch": 1.8452166966247254,
+      "grad_norm": 1.29851496219635,
+      "learning_rate": 4.1257843696520255e-06,
+      "loss": 0.0051,
+      "step": 4620
+    },
+    {
+      "epoch": 1.8492111044537647,
+      "grad_norm": 2.1423351764678955,
+      "learning_rate": 4.111523103251569e-06,
+      "loss": 0.0052,
+      "step": 4630
+    },
+    {
+      "epoch": 1.8532055122828042,
+      "grad_norm": 1.8427046537399292,
+      "learning_rate": 4.097261836851113e-06,
+      "loss": 0.0053,
+      "step": 4640
+    },
+    {
+      "epoch": 1.8571999201118434,
+      "grad_norm": 1.975304126739502,
+      "learning_rate": 4.083000570450656e-06,
+      "loss": 0.0096,
+      "step": 4650
+    },
+    {
+      "epoch": 1.8611943279408827,
+      "grad_norm": 1.3507777452468872,
+      "learning_rate": 4.0687393040502e-06,
+      "loss": 0.0063,
+      "step": 4660
+    },
+    {
+      "epoch": 1.8651887357699222,
+      "grad_norm": 1.3759689331054688,
+      "learning_rate": 4.054478037649744e-06,
+      "loss": 0.0036,
+      "step": 4670
+    },
+    {
+      "epoch": 1.8691831435989614,
+      "grad_norm": 2.0569324493408203,
+      "learning_rate": 4.0402167712492876e-06,
+      "loss": 0.0065,
+      "step": 4680
+    },
+    {
+      "epoch": 1.8731775514280007,
+      "grad_norm": 0.8964648246765137,
+      "learning_rate": 4.025955504848831e-06,
+      "loss": 0.0067,
+      "step": 4690
+    },
+    {
+      "epoch": 1.8771719592570402,
+      "grad_norm": 1.7848093509674072,
+      "learning_rate": 4.011694238448375e-06,
+      "loss": 0.0053,
+      "step": 4700
+    },
+    {
+      "epoch": 1.8811663670860796,
+      "grad_norm": 1.3376437425613403,
+      "learning_rate": 3.997432972047918e-06,
+      "loss": 0.0045,
+      "step": 4710
+    },
+    {
+      "epoch": 1.8851607749151187,
+      "grad_norm": 1.692192792892456,
+      "learning_rate": 3.983171705647461e-06,
+      "loss": 0.006,
+      "step": 4720
+    },
+    {
+      "epoch": 1.8891551827441582,
+      "grad_norm": 2.065828800201416,
+      "learning_rate": 3.9689104392470055e-06,
+      "loss": 0.0071,
+      "step": 4730
+    },
+    {
+      "epoch": 1.8931495905731976,
+      "grad_norm": 2.0423130989074707,
+      "learning_rate": 3.954649172846549e-06,
+      "loss": 0.0048,
+      "step": 4740
+    },
+    {
+      "epoch": 1.897143998402237,
+      "grad_norm": 2.064666509628296,
+      "learning_rate": 3.940387906446093e-06,
+      "loss": 0.0043,
+      "step": 4750
+    },
+    {
+      "epoch": 1.9011384062312762,
+      "grad_norm": 1.5965344905853271,
+      "learning_rate": 3.926126640045637e-06,
+      "loss": 0.0081,
+      "step": 4760
+    },
+    {
+      "epoch": 1.9051328140603156,
+      "grad_norm": 1.4221477508544922,
+      "learning_rate": 3.91186537364518e-06,
+      "loss": 0.0068,
+      "step": 4770
+    },
+    {
+      "epoch": 1.909127221889355,
+      "grad_norm": 1.6829274892807007,
+      "learning_rate": 3.897604107244723e-06,
+      "loss": 0.0048,
+      "step": 4780
+    },
+    {
+      "epoch": 1.9131216297183942,
+      "grad_norm": 2.7087581157684326,
+      "learning_rate": 3.8833428408442675e-06,
+      "loss": 0.0062,
+      "step": 4790
+    },
+    {
+      "epoch": 1.9171160375474336,
+      "grad_norm": 0.8025818467140198,
+      "learning_rate": 3.869081574443811e-06,
+      "loss": 0.0071,
+      "step": 4800
+    },
+    {
+      "epoch": 1.921110445376473,
+      "grad_norm": 0.2554706931114197,
+      "learning_rate": 3.854820308043354e-06,
+      "loss": 0.0066,
+      "step": 4810
+    },
+    {
+      "epoch": 1.9251048532055122,
+      "grad_norm": 1.1639817953109741,
+      "learning_rate": 3.840559041642898e-06,
+      "loss": 0.0053,
+      "step": 4820
+    },
+    {
+      "epoch": 1.9290992610345516,
+      "grad_norm": 1.7791696786880493,
+      "learning_rate": 3.826297775242442e-06,
+      "loss": 0.0053,
+      "step": 4830
+    },
+    {
+      "epoch": 1.9330936688635911,
+      "grad_norm": 0.6992547512054443,
+      "learning_rate": 3.8120365088419854e-06,
+      "loss": 0.0046,
+      "step": 4840
+    },
+    {
+      "epoch": 1.9370880766926302,
+      "grad_norm": 0.6868494153022766,
+      "learning_rate": 3.797775242441529e-06,
+      "loss": 0.0055,
+      "step": 4850
+    },
+    {
+      "epoch": 1.9410824845216696,
+      "grad_norm": 2.2475435733795166,
+      "learning_rate": 3.7835139760410728e-06,
+      "loss": 0.0054,
+      "step": 4860
+    },
+    {
+      "epoch": 1.9450768923507091,
+      "grad_norm": 0.551287829875946,
+      "learning_rate": 3.769252709640616e-06,
+      "loss": 0.0099,
+      "step": 4870
+    },
+    {
+      "epoch": 1.9490713001797484,
+      "grad_norm": 1.1871082782745361,
+      "learning_rate": 3.75499144324016e-06,
+      "loss": 0.003,
+      "step": 4880
+    },
+    {
+      "epoch": 1.9530657080087876,
+      "grad_norm": 0.4049108326435089,
+      "learning_rate": 3.740730176839704e-06,
+      "loss": 0.0071,
+      "step": 4890
+    },
+    {
+      "epoch": 1.9570601158378271,
+      "grad_norm": 3.0224263668060303,
+      "learning_rate": 3.726468910439247e-06,
+      "loss": 0.007,
+      "step": 4900
+    },
+    {
+      "epoch": 1.9610545236668664,
+      "grad_norm": 0.9450222253799438,
+      "learning_rate": 3.712207644038791e-06,
+      "loss": 0.0056,
+      "step": 4910
+    },
+    {
+      "epoch": 1.9650489314959056,
+      "grad_norm": 0.9193373918533325,
+      "learning_rate": 3.6979463776383344e-06,
+      "loss": 0.0056,
+      "step": 4920
+    },
+    {
+      "epoch": 1.9690433393249451,
+      "grad_norm": 0.4554131031036377,
+      "learning_rate": 3.683685111237878e-06,
+      "loss": 0.0053,
+      "step": 4930
+    },
+    {
+      "epoch": 1.9730377471539844,
+      "grad_norm": 0.42416155338287354,
+      "learning_rate": 3.669423844837422e-06,
+      "loss": 0.006,
+      "step": 4940
+    },
+    {
+      "epoch": 1.9770321549830236,
+      "grad_norm": 1.3764687776565552,
+      "learning_rate": 3.6551625784369654e-06,
+      "loss": 0.0039,
+      "step": 4950
+    },
+    {
+      "epoch": 1.9810265628120631,
+      "grad_norm": 0.7149096131324768,
+      "learning_rate": 3.640901312036509e-06,
+      "loss": 0.0055,
+      "step": 4960
+    },
+    {
+      "epoch": 1.9850209706411026,
+      "grad_norm": 0.6155685186386108,
+      "learning_rate": 3.6266400456360527e-06,
+      "loss": 0.0034,
+      "step": 4970
+    },
+    {
+      "epoch": 1.9890153784701416,
+      "grad_norm": 1.2946895360946655,
+      "learning_rate": 3.6123787792355964e-06,
+      "loss": 0.0035,
+      "step": 4980
+    },
+    {
+      "epoch": 1.9930097862991811,
+      "grad_norm": 0.5564864873886108,
+      "learning_rate": 3.5981175128351405e-06,
+      "loss": 0.0067,
+      "step": 4990
+    },
+    {
+      "epoch": 1.9970041941282206,
+      "grad_norm": 2.823580026626587,
+      "learning_rate": 3.5838562464346838e-06,
+      "loss": 0.007,
+      "step": 5000
+    },
+    {
+      "epoch": 2.000798881565808,
+      "grad_norm": 1.805331826210022,
+      "learning_rate": 3.5695949800342274e-06,
+      "loss": 0.0049,
+      "step": 5010
+    },
+    {
+      "epoch": 2.004793289394847,
+      "grad_norm": 0.3195802569389343,
+      "learning_rate": 3.555333713633771e-06,
+      "loss": 0.003,
+      "step": 5020
+    },
+    {
+      "epoch": 2.0087876972238865,
+      "grad_norm": 0.35623663663864136,
+      "learning_rate": 3.5410724472333148e-06,
+      "loss": 0.0014,
+      "step": 5030
+    },
+    {
+      "epoch": 2.012782105052926,
+      "grad_norm": 0.4197460114955902,
+      "learning_rate": 3.526811180832858e-06,
+      "loss": 0.0017,
+      "step": 5040
+    },
+    {
+      "epoch": 2.0167765128819655,
+      "grad_norm": 0.7055420279502869,
+      "learning_rate": 3.512549914432402e-06,
+      "loss": 0.0019,
+      "step": 5050
+    },
+    {
+      "epoch": 2.0207709207110045,
+      "grad_norm": 0.18588046729564667,
+      "learning_rate": 3.4982886480319454e-06,
+      "loss": 0.0021,
+      "step": 5060
+    },
+    {
+      "epoch": 2.024765328540044,
+      "grad_norm": 0.2396942675113678,
+      "learning_rate": 3.484027381631489e-06,
+      "loss": 0.0012,
+      "step": 5070
+    },
+    {
+      "epoch": 2.0287597363690835,
+      "grad_norm": 0.2188696563243866,
+      "learning_rate": 3.469766115231033e-06,
+      "loss": 0.0016,
+      "step": 5080
+    },
+    {
+      "epoch": 2.0327541441981225,
+      "grad_norm": 0.301581472158432,
+      "learning_rate": 3.4555048488305764e-06,
+      "loss": 0.0014,
+      "step": 5090
+    },
+    {
+      "epoch": 2.036748552027162,
+      "grad_norm": 0.7913476228713989,
+      "learning_rate": 3.44124358243012e-06,
+      "loss": 0.0019,
+      "step": 5100
+    },
+    {
+      "epoch": 2.0407429598562015,
+      "grad_norm": 0.28169021010398865,
+      "learning_rate": 3.4269823160296637e-06,
+      "loss": 0.0014,
+      "step": 5110
+    },
+    {
+      "epoch": 2.0447373676852405,
+      "grad_norm": 0.14495527744293213,
+      "learning_rate": 3.4127210496292074e-06,
+      "loss": 0.0029,
+      "step": 5120
+    },
+    {
+      "epoch": 2.04873177551428,
+      "grad_norm": 0.2131481021642685,
+      "learning_rate": 3.3984597832287506e-06,
+      "loss": 0.0009,
+      "step": 5130
+    },
+    {
+      "epoch": 2.0527261833433195,
+      "grad_norm": 0.12827646732330322,
+      "learning_rate": 3.3841985168282947e-06,
+      "loss": 0.0036,
+      "step": 5140
+    },
+    {
+      "epoch": 2.0567205911723585,
+      "grad_norm": 0.12576265633106232,
+      "learning_rate": 3.3699372504278384e-06,
+      "loss": 0.0013,
+      "step": 5150
+    },
+    {
+      "epoch": 2.060714999001398,
+      "grad_norm": 1.2176573276519775,
+      "learning_rate": 3.3556759840273816e-06,
+      "loss": 0.0014,
+      "step": 5160
+    },
+    {
+      "epoch": 2.0647094068304375,
+      "grad_norm": 0.11473061144351959,
+      "learning_rate": 3.3414147176269257e-06,
+      "loss": 0.0011,
+      "step": 5170
+    },
+    {
+      "epoch": 2.068703814659477,
+      "grad_norm": 0.25290143489837646,
+      "learning_rate": 3.327153451226469e-06,
+      "loss": 0.0015,
+      "step": 5180
+    },
+    {
+      "epoch": 2.072698222488516,
+      "grad_norm": 1.1201770305633545,
+      "learning_rate": 3.3128921848260127e-06,
+      "loss": 0.0025,
+      "step": 5190
+    },
+    {
+      "epoch": 2.0766926303175555,
+      "grad_norm": 0.3146374523639679,
+      "learning_rate": 3.2986309184255568e-06,
+      "loss": 0.0021,
+      "step": 5200
+    },
+    {
+      "epoch": 2.080687038146595,
+      "grad_norm": 0.29306668043136597,
+      "learning_rate": 3.2843696520251e-06,
+      "loss": 0.0016,
+      "step": 5210
+    },
+    {
+      "epoch": 2.084681445975634,
+      "grad_norm": 0.09235348552465439,
+      "learning_rate": 3.2701083856246437e-06,
+      "loss": 0.0025,
+      "step": 5220
+    },
+    {
+      "epoch": 2.0886758538046735,
+      "grad_norm": 0.2355571836233139,
+      "learning_rate": 3.2558471192241873e-06,
+      "loss": 0.0018,
+      "step": 5230
+    },
+    {
+      "epoch": 2.092670261633713,
+      "grad_norm": 0.26922300457954407,
+      "learning_rate": 3.241585852823731e-06,
+      "loss": 0.0014,
+      "step": 5240
+    },
+    {
+      "epoch": 2.096664669462752,
+      "grad_norm": 0.375124990940094,
+      "learning_rate": 3.2273245864232743e-06,
+      "loss": 0.0017,
+      "step": 5250
+    },
+    {
+      "epoch": 2.1006590772917915,
+      "grad_norm": 0.2880096137523651,
+      "learning_rate": 3.2130633200228184e-06,
+      "loss": 0.0011,
+      "step": 5260
+    },
+    {
+      "epoch": 2.104653485120831,
+      "grad_norm": 0.1137225404381752,
+      "learning_rate": 3.198802053622362e-06,
+      "loss": 0.0014,
+      "step": 5270
+    },
+    {
+      "epoch": 2.10864789294987,
+      "grad_norm": 0.10410904139280319,
+      "learning_rate": 3.1845407872219053e-06,
+      "loss": 0.0016,
+      "step": 5280
+    },
+    {
+      "epoch": 2.1126423007789095,
+      "grad_norm": 0.32652702927589417,
+      "learning_rate": 3.1702795208214494e-06,
+      "loss": 0.0013,
+      "step": 5290
+    },
+    {
+      "epoch": 2.116636708607949,
+      "grad_norm": 0.11358338594436646,
+      "learning_rate": 3.1560182544209926e-06,
+      "loss": 0.0012,
+      "step": 5300
+    },
+    {
+      "epoch": 2.1206311164369884,
+      "grad_norm": 1.5589320659637451,
+      "learning_rate": 3.1417569880205367e-06,
+      "loss": 0.0018,
+      "step": 5310
+    },
+    {
+      "epoch": 2.1246255242660275,
+      "grad_norm": 0.2536420226097107,
+      "learning_rate": 3.1274957216200804e-06,
+      "loss": 0.0011,
+      "step": 5320
+    },
+    {
+      "epoch": 2.128619932095067,
+      "grad_norm": 0.6009411215782166,
+      "learning_rate": 3.1132344552196236e-06,
+      "loss": 0.0015,
+      "step": 5330
+    },
+    {
+      "epoch": 2.1326143399241064,
+      "grad_norm": 0.49442583322525024,
+      "learning_rate": 3.0989731888191677e-06,
+      "loss": 0.0012,
+      "step": 5340
+    },
+    {
+      "epoch": 2.1366087477531455,
+      "grad_norm": 0.10753143578767776,
+      "learning_rate": 3.084711922418711e-06,
+      "loss": 0.0012,
+      "step": 5350
+    },
+    {
+      "epoch": 2.140603155582185,
+      "grad_norm": 0.4049818217754364,
+      "learning_rate": 3.0704506560182546e-06,
+      "loss": 0.0013,
+      "step": 5360
+    },
+    {
+      "epoch": 2.1445975634112244,
+      "grad_norm": 0.1319737732410431,
+      "learning_rate": 3.0561893896177987e-06,
+      "loss": 0.0014,
+      "step": 5370
+    },
+    {
+      "epoch": 2.1485919712402635,
+      "grad_norm": 0.1411842256784439,
+      "learning_rate": 3.041928123217342e-06,
+      "loss": 0.0043,
+      "step": 5380
+    },
+    {
+      "epoch": 2.152586379069303,
+      "grad_norm": 0.8159494996070862,
+      "learning_rate": 3.0276668568168852e-06,
+      "loss": 0.0009,
+      "step": 5390
+    },
+    {
+      "epoch": 2.1565807868983424,
+      "grad_norm": 1.0257997512817383,
+      "learning_rate": 3.0134055904164293e-06,
+      "loss": 0.0013,
+      "step": 5400
+    },
+    {
+      "epoch": 2.1605751947273815,
+      "grad_norm": 0.17758966982364655,
+      "learning_rate": 2.999144324015973e-06,
+      "loss": 0.0023,
+      "step": 5410
+    },
+    {
+      "epoch": 2.164569602556421,
+      "grad_norm": 1.4170286655426025,
+      "learning_rate": 2.9848830576155162e-06,
+      "loss": 0.0032,
+      "step": 5420
+    },
+    {
+      "epoch": 2.1685640103854604,
+      "grad_norm": 0.18177098035812378,
+      "learning_rate": 2.9706217912150603e-06,
+      "loss": 0.0025,
+      "step": 5430
+    },
+    {
+      "epoch": 2.1725584182145,
+      "grad_norm": 0.3801667094230652,
+      "learning_rate": 2.9563605248146036e-06,
+      "loss": 0.0015,
+      "step": 5440
+    },
+    {
+      "epoch": 2.176552826043539,
+      "grad_norm": 0.4188409447669983,
+      "learning_rate": 2.9420992584141473e-06,
+      "loss": 0.0009,
+      "step": 5450
+    },
+    {
+      "epoch": 2.1805472338725784,
+      "grad_norm": 0.13465771079063416,
+      "learning_rate": 2.9278379920136913e-06,
+      "loss": 0.0012,
+      "step": 5460
+    },
+    {
+      "epoch": 2.184541641701618,
+      "grad_norm": 0.18084383010864258,
+      "learning_rate": 2.9135767256132346e-06,
+      "loss": 0.0033,
+      "step": 5470
+    },
+    {
+      "epoch": 2.188536049530657,
+      "grad_norm": 0.9558360576629639,
+      "learning_rate": 2.8993154592127783e-06,
+      "loss": 0.0052,
+      "step": 5480
+    },
+    {
+      "epoch": 2.1925304573596964,
+      "grad_norm": 0.07331178337335587,
+      "learning_rate": 2.885054192812322e-06,
+      "loss": 0.0019,
+      "step": 5490
+    },
+    {
+      "epoch": 2.196524865188736,
+      "grad_norm": 0.3440713584423065,
+      "learning_rate": 2.8707929264118656e-06,
+      "loss": 0.0016,
+      "step": 5500
+    },
+    {
+      "epoch": 2.200519273017775,
+      "grad_norm": 0.32798197865486145,
+      "learning_rate": 2.856531660011409e-06,
+      "loss": 0.0012,
+      "step": 5510
+    },
+    {
+      "epoch": 2.2045136808468144,
+      "grad_norm": 0.6357255578041077,
+      "learning_rate": 2.842270393610953e-06,
+      "loss": 0.0011,
+      "step": 5520
+    },
+    {
+      "epoch": 2.208508088675854,
+      "grad_norm": 0.5576416850090027,
+      "learning_rate": 2.8280091272104966e-06,
+      "loss": 0.0013,
+      "step": 5530
+    },
+    {
+      "epoch": 2.212502496504893,
+      "grad_norm": 0.24470700323581696,
+      "learning_rate": 2.81374786081004e-06,
+      "loss": 0.0016,
+      "step": 5540
+    },
+    {
+      "epoch": 2.2164969043339324,
+      "grad_norm": 0.13638091087341309,
+      "learning_rate": 2.799486594409584e-06,
+      "loss": 0.0024,
+      "step": 5550
+    },
+    {
+      "epoch": 2.220491312162972,
+      "grad_norm": 0.17628563940525055,
+      "learning_rate": 2.785225328009127e-06,
+      "loss": 0.0026,
+      "step": 5560
+    },
+    {
+      "epoch": 2.2244857199920114,
+      "grad_norm": 0.3423454165458679,
+      "learning_rate": 2.770964061608671e-06,
+      "loss": 0.0045,
+      "step": 5570
+    },
+    {
+      "epoch": 2.2284801278210504,
+      "grad_norm": 0.16189979016780853,
+      "learning_rate": 2.756702795208215e-06,
+      "loss": 0.001,
+      "step": 5580
+    },
+    {
+      "epoch": 2.23247453565009,
+      "grad_norm": 0.2525485157966614,
+      "learning_rate": 2.7424415288077582e-06,
+      "loss": 0.0016,
+      "step": 5590
+    },
+    {
+      "epoch": 2.2364689434791294,
+      "grad_norm": 0.23971009254455566,
+      "learning_rate": 2.728180262407302e-06,
+      "loss": 0.0013,
+      "step": 5600
+    },
+    {
+      "epoch": 2.2404633513081684,
+      "grad_norm": 0.1524650901556015,
+      "learning_rate": 2.7139189960068456e-06,
+      "loss": 0.0008,
+      "step": 5610
+    },
+    {
+      "epoch": 2.244457759137208,
+      "grad_norm": 0.3585372865200043,
+      "learning_rate": 2.6996577296063892e-06,
+      "loss": 0.0021,
+      "step": 5620
+    },
+    {
+      "epoch": 2.2484521669662474,
+      "grad_norm": 0.1550854593515396,
+      "learning_rate": 2.6853964632059333e-06,
+      "loss": 0.0015,
+      "step": 5630
+    },
+    {
+      "epoch": 2.2524465747952864,
+      "grad_norm": 0.1160367801785469,
+      "learning_rate": 2.6711351968054766e-06,
+      "loss": 0.0025,
+      "step": 5640
+    },
+    {
+      "epoch": 2.256440982624326,
+      "grad_norm": 0.8458707928657532,
+      "learning_rate": 2.6568739304050202e-06,
+      "loss": 0.0017,
+      "step": 5650
+    },
+    {
+      "epoch": 2.2604353904533654,
+      "grad_norm": 0.4168691039085388,
+      "learning_rate": 2.642612664004564e-06,
+      "loss": 0.0013,
+      "step": 5660
+    },
+    {
+      "epoch": 2.2644297982824044,
+      "grad_norm": 0.21303381025791168,
+      "learning_rate": 2.6283513976041076e-06,
+      "loss": 0.0018,
+      "step": 5670
+    },
+    {
+      "epoch": 2.268424206111444,
+      "grad_norm": 0.08199158310890198,
+      "learning_rate": 2.614090131203651e-06,
+      "loss": 0.0018,
+      "step": 5680
+    },
+    {
+      "epoch": 2.2724186139404834,
+      "grad_norm": 1.4730623960494995,
+      "learning_rate": 2.599828864803195e-06,
+      "loss": 0.0021,
+      "step": 5690
+    },
+    {
+      "epoch": 2.276413021769523,
+      "grad_norm": 0.15917807817459106,
+      "learning_rate": 2.5855675984027386e-06,
+      "loss": 0.0022,
+      "step": 5700
+    },
+    {
+      "epoch": 2.280407429598562,
+      "grad_norm": 0.3687230050563812,
+      "learning_rate": 2.571306332002282e-06,
+      "loss": 0.0013,
+      "step": 5710
+    },
+    {
+      "epoch": 2.2844018374276014,
+      "grad_norm": 0.08973414450883865,
+      "learning_rate": 2.557045065601826e-06,
+      "loss": 0.0017,
+      "step": 5720
+    },
+    {
+      "epoch": 2.288396245256641,
+      "grad_norm": 0.2926589846611023,
+      "learning_rate": 2.542783799201369e-06,
+      "loss": 0.0016,
+      "step": 5730
+    },
+    {
+      "epoch": 2.29239065308568,
+      "grad_norm": 0.26070043444633484,
+      "learning_rate": 2.528522532800913e-06,
+      "loss": 0.0023,
+      "step": 5740
+    },
+    {
+      "epoch": 2.2963850609147194,
+      "grad_norm": 0.23214001953601837,
+      "learning_rate": 2.514261266400457e-06,
+      "loss": 0.0018,
+      "step": 5750
+    },
+    {
+      "epoch": 2.300379468743759,
+      "grad_norm": 1.664347767829895,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0015,
+      "step": 5760
+    },
+    {
+      "epoch": 2.304373876572798,
+      "grad_norm": 0.2514505684375763,
+      "learning_rate": 2.485738733599544e-06,
+      "loss": 0.0013,
+      "step": 5770
+    },
+    {
+      "epoch": 2.3083682844018374,
+      "grad_norm": 0.11796104162931442,
+      "learning_rate": 2.4714774671990875e-06,
+      "loss": 0.0034,
+      "step": 5780
+    },
+    {
+      "epoch": 2.312362692230877,
+      "grad_norm": 0.6030930280685425,
+      "learning_rate": 2.4572162007986312e-06,
+      "loss": 0.002,
+      "step": 5790
+    },
+    {
+      "epoch": 2.316357100059916,
+      "grad_norm": 1.2356213331222534,
+      "learning_rate": 2.442954934398175e-06,
+      "loss": 0.0019,
+      "step": 5800
+    },
+    {
+      "epoch": 2.3203515078889554,
+      "grad_norm": 0.17977607250213623,
+      "learning_rate": 2.428693667997718e-06,
+      "loss": 0.0014,
+      "step": 5810
+    },
+    {
+      "epoch": 2.324345915717995,
+      "grad_norm": 0.6751307249069214,
+      "learning_rate": 2.414432401597262e-06,
+      "loss": 0.0011,
+      "step": 5820
+    },
+    {
+      "epoch": 2.3283403235470344,
+      "grad_norm": 0.3013259172439575,
+      "learning_rate": 2.400171135196806e-06,
+      "loss": 0.0011,
+      "step": 5830
+    },
+    {
+      "epoch": 2.3323347313760734,
+      "grad_norm": 0.4793102741241455,
+      "learning_rate": 2.3859098687963496e-06,
+      "loss": 0.0022,
+      "step": 5840
+    },
+    {
+      "epoch": 2.336329139205113,
+      "grad_norm": 0.20744891464710236,
+      "learning_rate": 2.371648602395893e-06,
+      "loss": 0.0015,
+      "step": 5850
+    },
+    {
+      "epoch": 2.3403235470341524,
+      "grad_norm": 0.18747301399707794,
+      "learning_rate": 2.3573873359954365e-06,
+      "loss": 0.001,
+      "step": 5860
+    },
+    {
+      "epoch": 2.3443179548631914,
+      "grad_norm": 0.285792738199234,
+      "learning_rate": 2.34312606959498e-06,
+      "loss": 0.0012,
+      "step": 5870
+    },
+    {
+      "epoch": 2.348312362692231,
+      "grad_norm": 0.3376331031322479,
+      "learning_rate": 2.328864803194524e-06,
+      "loss": 0.0017,
+      "step": 5880
+    },
+    {
+      "epoch": 2.3523067705212704,
+      "grad_norm": 0.43353843688964844,
+      "learning_rate": 2.3146035367940675e-06,
+      "loss": 0.0016,
+      "step": 5890
+    },
+    {
+      "epoch": 2.3563011783503094,
+      "grad_norm": 0.214224711060524,
+      "learning_rate": 2.300342270393611e-06,
+      "loss": 0.001,
+      "step": 5900
+    },
+    {
+      "epoch": 2.360295586179349,
+      "grad_norm": 0.2301313877105713,
+      "learning_rate": 2.286081003993155e-06,
+      "loss": 0.0017,
+      "step": 5910
+    },
+    {
+      "epoch": 2.3642899940083884,
+      "grad_norm": 0.1547565907239914,
+      "learning_rate": 2.2718197375926985e-06,
+      "loss": 0.0027,
+      "step": 5920
+    },
+    {
+      "epoch": 2.3682844018374274,
+      "grad_norm": 0.08032135665416718,
+      "learning_rate": 2.257558471192242e-06,
+      "loss": 0.0024,
+      "step": 5930
+    },
+    {
+      "epoch": 2.372278809666467,
+      "grad_norm": 0.12723907828330994,
+      "learning_rate": 2.2432972047917854e-06,
+      "loss": 0.0017,
+      "step": 5940
+    },
+    {
+      "epoch": 2.3762732174955064,
+      "grad_norm": 0.13044141232967377,
+      "learning_rate": 2.229035938391329e-06,
+      "loss": 0.0011,
+      "step": 5950
+    },
+    {
+      "epoch": 2.380267625324546,
+      "grad_norm": 0.07979059964418411,
+      "learning_rate": 2.214774671990873e-06,
+      "loss": 0.0009,
+      "step": 5960
+    },
+    {
+      "epoch": 2.384262033153585,
+      "grad_norm": 0.4637604057788849,
+      "learning_rate": 2.2005134055904164e-06,
+      "loss": 0.0022,
+      "step": 5970
+    },
+    {
+      "epoch": 2.3882564409826244,
+      "grad_norm": 0.11707437038421631,
+      "learning_rate": 2.18625213918996e-06,
+      "loss": 0.0013,
+      "step": 5980
+    },
+    {
+      "epoch": 2.392250848811664,
+      "grad_norm": 0.1091870591044426,
+      "learning_rate": 2.171990872789504e-06,
+      "loss": 0.0025,
+      "step": 5990
+    },
+    {
+      "epoch": 2.396245256640703,
+      "grad_norm": 0.060703348368406296,
+      "learning_rate": 2.1577296063890475e-06,
+      "loss": 0.0012,
+      "step": 6000
+    },
+    {
+      "epoch": 2.396245256640703,
+      "eval_loss": 0.005135956685990095,
+      "eval_runtime": 7288.4368,
+      "eval_samples_per_second": 2.748,
+      "eval_steps_per_second": 0.344,
+      "eval_wer": 0.3684662456659022,
+      "step": 6000
+    }
+  ],
+  "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": 5.539223607902208e+19,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}