add .pt and more

README.md

@@ -219,7 +219,7 @@ Then you can load this model and run inference.
 from sentence_transformers import SentenceTransformer
 
 # Download from the 🤗 Hub
-model = SentenceTransformer("legalvn/paraphrase-multilingual-MiniLM-L12-v2-vn-169500")
+model = SentenceTransformer("sentence_transformers_model_id")
 # Run inference
 sentences = [
     'Thời điểm đánh giá và xếp loại chất lượng hằng năm của công chức, viên chức thuộc Bộ Tài chính được diễn ra trong thời gian nào?',

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "tmp_trainer/checkpoint-169500",
+  "_name_or_path": "sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
   "architectures": [
     "BertModel"
   ],

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:90e5ca1dbfc300353fcb0bbdf78277e3c53fccc31f146ce0109cf20c2e9339ee
+size 940241530

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60fb4b92e575f89b58a2242d3aa9c76a9029e8bd13f18beffa2dc9b430b95943
+size 14244

scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:15215f4aaf333a1c2688c7f9643e09826c5b2865375050007a20130e0ced6c19
+size 1064

trainer_state.json

@@ -0,0 +1,2406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0806225910195666,
+  "eval_steps": 500,
+  "global_step": 169500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00613752976701937,
+      "grad_norm": 0.7345873117446899,
+      "learning_rate": 4.9897707837216344e-05,
+      "loss": 1.0473,
+      "step": 500
+    },
+    {
+      "epoch": 0.01227505953403874,
+      "grad_norm": 1.5779420137405396,
+      "learning_rate": 4.9795415674432687e-05,
+      "loss": 1.0447,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01841258930105811,
+      "grad_norm": 1.1132322549819946,
+      "learning_rate": 4.969312351164903e-05,
+      "loss": 1.0383,
+      "step": 1500
+    },
+    {
+      "epoch": 0.02455011906807748,
+      "grad_norm": 1.4636728763580322,
+      "learning_rate": 4.959083134886538e-05,
+      "loss": 1.0395,
+      "step": 2000
+    },
+    {
+      "epoch": 0.03068764883509685,
+      "grad_norm": 2.259823799133301,
+      "learning_rate": 4.948853918608172e-05,
+      "loss": 1.0436,
+      "step": 2500
+    },
+    {
+      "epoch": 0.03682517860211622,
+      "grad_norm": 3.276669979095459,
+      "learning_rate": 4.938624702329807e-05,
+      "loss": 1.0375,
+      "step": 3000
+    },
+    {
+      "epoch": 0.04296270836913559,
+      "grad_norm": 2.432560443878174,
+      "learning_rate": 4.928395486051441e-05,
+      "loss": 1.0189,
+      "step": 3500
+    },
+    {
+      "epoch": 0.04910023813615496,
+      "grad_norm": 2.106665849685669,
+      "learning_rate": 4.918166269773075e-05,
+      "loss": 1.0282,
+      "step": 4000
+    },
+    {
+      "epoch": 0.05523776790317433,
+      "grad_norm": 10.379436492919922,
+      "learning_rate": 4.9079370534947094e-05,
+      "loss": 1.0355,
+      "step": 4500
+    },
+    {
+      "epoch": 0.0613752976701937,
+      "grad_norm": 1.7805677652359009,
+      "learning_rate": 4.897707837216344e-05,
+      "loss": 1.0286,
+      "step": 5000
+    },
+    {
+      "epoch": 0.06751282743721307,
+      "grad_norm": 4.699483394622803,
+      "learning_rate": 4.8874786209379785e-05,
+      "loss": 1.0264,
+      "step": 5500
+    },
+    {
+      "epoch": 0.07365035720423244,
+      "grad_norm": 3.2066714763641357,
+      "learning_rate": 4.877249404659613e-05,
+      "loss": 1.0174,
+      "step": 6000
+    },
+    {
+      "epoch": 0.07978788697125182,
+      "grad_norm": 4.304272651672363,
+      "learning_rate": 4.867020188381247e-05,
+      "loss": 1.0238,
+      "step": 6500
+    },
+    {
+      "epoch": 0.08592541673827118,
+      "grad_norm": 1.9387307167053223,
+      "learning_rate": 4.856790972102881e-05,
+      "loss": 1.0217,
+      "step": 7000
+    },
+    {
+      "epoch": 0.09206294650529055,
+      "grad_norm": 7.0195488929748535,
+      "learning_rate": 4.846561755824516e-05,
+      "loss": 1.0203,
+      "step": 7500
+    },
+    {
+      "epoch": 0.09820047627230992,
+      "grad_norm": 5.841785907745361,
+      "learning_rate": 4.836332539546151e-05,
+      "loss": 1.0201,
+      "step": 8000
+    },
+    {
+      "epoch": 0.10433800603932929,
+      "grad_norm": 4.427718162536621,
+      "learning_rate": 4.826103323267785e-05,
+      "loss": 1.0266,
+      "step": 8500
+    },
+    {
+      "epoch": 0.11047553580634865,
+      "grad_norm": 2.067218780517578,
+      "learning_rate": 4.815874106989419e-05,
+      "loss": 1.0379,
+      "step": 9000
+    },
+    {
+      "epoch": 0.11661306557336804,
+      "grad_norm": 4.303891181945801,
+      "learning_rate": 4.8056448907110535e-05,
+      "loss": 1.0367,
+      "step": 9500
+    },
+    {
+      "epoch": 0.1227505953403874,
+      "grad_norm": 5.363577842712402,
+      "learning_rate": 4.795415674432688e-05,
+      "loss": 1.0384,
+      "step": 10000
+    },
+    {
+      "epoch": 0.12888812510740677,
+      "grad_norm": 7.8037028312683105,
+      "learning_rate": 4.785186458154322e-05,
+      "loss": 1.0291,
+      "step": 10500
+    },
+    {
+      "epoch": 0.13502565487442614,
+      "grad_norm": 2.8838307857513428,
+      "learning_rate": 4.774957241875956e-05,
+      "loss": 1.0362,
+      "step": 11000
+    },
+    {
+      "epoch": 0.1411631846414455,
+      "grad_norm": 18.838226318359375,
+      "learning_rate": 4.764728025597591e-05,
+      "loss": 1.0354,
+      "step": 11500
+    },
+    {
+      "epoch": 0.14730071440846487,
+      "grad_norm": 1.8900924921035767,
+      "learning_rate": 4.754498809319225e-05,
+      "loss": 1.0204,
+      "step": 12000
+    },
+    {
+      "epoch": 0.15343824417548424,
+      "grad_norm": 8.636752128601074,
+      "learning_rate": 4.74426959304086e-05,
+      "loss": 1.0401,
+      "step": 12500
+    },
+    {
+      "epoch": 0.15957577394250363,
+      "grad_norm": 3.1361780166625977,
+      "learning_rate": 4.734040376762494e-05,
+      "loss": 1.0237,
+      "step": 13000
+    },
+    {
+      "epoch": 0.165713303709523,
+      "grad_norm": 6.990635871887207,
+      "learning_rate": 4.7238111604841285e-05,
+      "loss": 1.0271,
+      "step": 13500
+    },
+    {
+      "epoch": 0.17185083347654237,
+      "grad_norm": 2.6320559978485107,
+      "learning_rate": 4.713581944205763e-05,
+      "loss": 1.0235,
+      "step": 14000
+    },
+    {
+      "epoch": 0.17798836324356174,
+      "grad_norm": 5.751638889312744,
+      "learning_rate": 4.7033527279273976e-05,
+      "loss": 1.0329,
+      "step": 14500
+    },
+    {
+      "epoch": 0.1841258930105811,
+      "grad_norm": 10.333562850952148,
+      "learning_rate": 4.693123511649032e-05,
+      "loss": 1.0474,
+      "step": 15000
+    },
+    {
+      "epoch": 0.19026342277760047,
+      "grad_norm": 9.57418155670166,
+      "learning_rate": 4.682894295370666e-05,
+      "loss": 1.0547,
+      "step": 15500
+    },
+    {
+      "epoch": 0.19640095254461984,
+      "grad_norm": 6.002443313598633,
+      "learning_rate": 4.6726650790923e-05,
+      "loss": 1.0557,
+      "step": 16000
+    },
+    {
+      "epoch": 0.2025384823116392,
+      "grad_norm": 17.77951431274414,
+      "learning_rate": 4.6624358628139344e-05,
+      "loss": 1.0626,
+      "step": 16500
+    },
+    {
+      "epoch": 0.20867601207865857,
+      "grad_norm": 10.126110076904297,
+      "learning_rate": 4.652206646535569e-05,
+      "loss": 1.0551,
+      "step": 17000
+    },
+    {
+      "epoch": 0.21481354184567794,
+      "grad_norm": 3.7376019954681396,
+      "learning_rate": 4.641977430257204e-05,
+      "loss": 1.0526,
+      "step": 17500
+    },
+    {
+      "epoch": 0.2209510716126973,
+      "grad_norm": 19.0529842376709,
+      "learning_rate": 4.6317482139788384e-05,
+      "loss": 1.125,
+      "step": 18000
+    },
+    {
+      "epoch": 0.2270886013797167,
+      "grad_norm": 1.9759446382522583,
+      "learning_rate": 4.6215189977004726e-05,
+      "loss": 1.2996,
+      "step": 18500
+    },
+    {
+      "epoch": 0.23322613114673607,
+      "grad_norm": 2.7964680194854736,
+      "learning_rate": 4.611289781422107e-05,
+      "loss": 1.0703,
+      "step": 19000
+    },
+    {
+      "epoch": 0.23936366091375544,
+      "grad_norm": 18.77906036376953,
+      "learning_rate": 4.601060565143741e-05,
+      "loss": 1.0601,
+      "step": 19500
+    },
+    {
+      "epoch": 0.2455011906807748,
+      "grad_norm": 3.129147529602051,
+      "learning_rate": 4.590831348865375e-05,
+      "loss": 1.0835,
+      "step": 20000
+    },
+    {
+      "epoch": 0.2516387204477942,
+      "grad_norm": 3.018688917160034,
+      "learning_rate": 4.58060213258701e-05,
+      "loss": 1.0583,
+      "step": 20500
+    },
+    {
+      "epoch": 0.25777625021481354,
+      "grad_norm": 32.16483688354492,
+      "learning_rate": 4.570372916308644e-05,
+      "loss": 1.141,
+      "step": 21000
+    },
+    {
+      "epoch": 0.2639137799818329,
+      "grad_norm": 8.431964874267578,
+      "learning_rate": 4.5601437000302785e-05,
+      "loss": 1.0802,
+      "step": 21500
+    },
+    {
+      "epoch": 0.2700513097488523,
+      "grad_norm": 14.990534782409668,
+      "learning_rate": 4.5499144837519134e-05,
+      "loss": 1.0589,
+      "step": 22000
+    },
+    {
+      "epoch": 0.27618883951587164,
+      "grad_norm": 6.134013652801514,
+      "learning_rate": 4.5396852674735476e-05,
+      "loss": 1.086,
+      "step": 22500
+    },
+    {
+      "epoch": 0.282326369282891,
+      "grad_norm": 15.587114334106445,
+      "learning_rate": 4.529456051195182e-05,
+      "loss": 1.0743,
+      "step": 23000
+    },
+    {
+      "epoch": 0.2884638990499104,
+      "grad_norm": 20.36590003967285,
+      "learning_rate": 4.519226834916816e-05,
+      "loss": 1.0605,
+      "step": 23500
+    },
+    {
+      "epoch": 0.29460142881692974,
+      "grad_norm": 6.840169906616211,
+      "learning_rate": 4.508997618638451e-05,
+      "loss": 1.0602,
+      "step": 24000
+    },
+    {
+      "epoch": 0.3007389585839491,
+      "grad_norm": 22.10498046875,
+      "learning_rate": 4.498768402360085e-05,
+      "loss": 1.0732,
+      "step": 24500
+    },
+    {
+      "epoch": 0.3068764883509685,
+      "grad_norm": 2.3044190406799316,
+      "learning_rate": 4.488539186081719e-05,
+      "loss": 1.0614,
+      "step": 25000
+    },
+    {
+      "epoch": 0.31301401811798785,
+      "grad_norm": 19.358158111572266,
+      "learning_rate": 4.4783099698033535e-05,
+      "loss": 1.0666,
+      "step": 25500
+    },
+    {
+      "epoch": 0.31915154788500727,
+      "grad_norm": 47.77769088745117,
+      "learning_rate": 4.468080753524988e-05,
+      "loss": 1.0669,
+      "step": 26000
+    },
+    {
+      "epoch": 0.32528907765202664,
+      "grad_norm": 31.26322364807129,
+      "learning_rate": 4.4578515372466226e-05,
+      "loss": 1.0627,
+      "step": 26500
+    },
+    {
+      "epoch": 0.331426607419046,
+      "grad_norm": 31.237651824951172,
+      "learning_rate": 4.4476223209682575e-05,
+      "loss": 1.0659,
+      "step": 27000
+    },
+    {
+      "epoch": 0.33756413718606537,
+      "grad_norm": 10.651726722717285,
+      "learning_rate": 4.437393104689892e-05,
+      "loss": 1.07,
+      "step": 27500
+    },
+    {
+      "epoch": 0.34370166695308474,
+      "grad_norm": 10.767593383789062,
+      "learning_rate": 4.427163888411526e-05,
+      "loss": 1.0783,
+      "step": 28000
+    },
+    {
+      "epoch": 0.3498391967201041,
+      "grad_norm": 13.177048683166504,
+      "learning_rate": 4.41693467213316e-05,
+      "loss": 1.078,
+      "step": 28500
+    },
+    {
+      "epoch": 0.3559767264871235,
+      "grad_norm": 41.34538269042969,
+      "learning_rate": 4.406705455854794e-05,
+      "loss": 1.0832,
+      "step": 29000
+    },
+    {
+      "epoch": 0.36211425625414284,
+      "grad_norm": 2.813678026199341,
+      "learning_rate": 4.3964762395764285e-05,
+      "loss": 1.0695,
+      "step": 29500
+    },
+    {
+      "epoch": 0.3682517860211622,
+      "grad_norm": 32.63833236694336,
+      "learning_rate": 4.3862470232980634e-05,
+      "loss": 1.0714,
+      "step": 30000
+    },
+    {
+      "epoch": 0.3743893157881816,
+      "grad_norm": 6.144928932189941,
+      "learning_rate": 4.3760178070196976e-05,
+      "loss": 1.3794,
+      "step": 30500
+    },
+    {
+      "epoch": 0.38052684555520094,
+      "grad_norm": 21.691205978393555,
+      "learning_rate": 4.365788590741332e-05,
+      "loss": 1.0838,
+      "step": 31000
+    },
+    {
+      "epoch": 0.3866643753222203,
+      "grad_norm": 5.830935955047607,
+      "learning_rate": 4.355559374462967e-05,
+      "loss": 1.0541,
+      "step": 31500
+    },
+    {
+      "epoch": 0.3928019050892397,
+      "grad_norm": 5.59116792678833,
+      "learning_rate": 4.345330158184601e-05,
+      "loss": 1.0799,
+      "step": 32000
+    },
+    {
+      "epoch": 0.39893943485625905,
+      "grad_norm": 5.144099712371826,
+      "learning_rate": 4.335100941906235e-05,
+      "loss": 1.0622,
+      "step": 32500
+    },
+    {
+      "epoch": 0.4050769646232784,
+      "grad_norm": 30.603145599365234,
+      "learning_rate": 4.324871725627869e-05,
+      "loss": 1.0597,
+      "step": 33000
+    },
+    {
+      "epoch": 0.4112144943902978,
+      "grad_norm": 14.602336883544922,
+      "learning_rate": 4.314642509349504e-05,
+      "loss": 1.0731,
+      "step": 33500
+    },
+    {
+      "epoch": 0.41735202415731715,
+      "grad_norm": 11.165698051452637,
+      "learning_rate": 4.3044132930711384e-05,
+      "loss": 1.0871,
+      "step": 34000
+    },
+    {
+      "epoch": 0.4234895539243365,
+      "grad_norm": 7.897899627685547,
+      "learning_rate": 4.2941840767927726e-05,
+      "loss": 1.0535,
+      "step": 34500
+    },
+    {
+      "epoch": 0.4296270836913559,
+      "grad_norm": 34.061134338378906,
+      "learning_rate": 4.283954860514407e-05,
+      "loss": 1.3215,
+      "step": 35000
+    },
+    {
+      "epoch": 0.43576461345837525,
+      "grad_norm": 5.050889015197754,
+      "learning_rate": 4.273725644236041e-05,
+      "loss": 1.1501,
+      "step": 35500
+    },
+    {
+      "epoch": 0.4419021432253946,
+      "grad_norm": 11.23651123046875,
+      "learning_rate": 4.263496427957676e-05,
+      "loss": 1.1088,
+      "step": 36000
+    },
+    {
+      "epoch": 0.44803967299241404,
+      "grad_norm": 11.963942527770996,
+      "learning_rate": 4.25326721167931e-05,
+      "loss": 1.0844,
+      "step": 36500
+    },
+    {
+      "epoch": 0.4541772027594334,
+      "grad_norm": 19.694747924804688,
+      "learning_rate": 4.243037995400945e-05,
+      "loss": 1.0981,
+      "step": 37000
+    },
+    {
+      "epoch": 0.4603147325264528,
+      "grad_norm": 32.396446228027344,
+      "learning_rate": 4.232808779122579e-05,
+      "loss": 1.0856,
+      "step": 37500
+    },
+    {
+      "epoch": 0.46645226229347214,
+      "grad_norm": 30.794092178344727,
+      "learning_rate": 4.2225795628442134e-05,
+      "loss": 1.0956,
+      "step": 38000
+    },
+    {
+      "epoch": 0.4725897920604915,
+      "grad_norm": 25.20745086669922,
+      "learning_rate": 4.2123503465658476e-05,
+      "loss": 1.0813,
+      "step": 38500
+    },
+    {
+      "epoch": 0.4787273218275109,
+      "grad_norm": 13.46246337890625,
+      "learning_rate": 4.202121130287482e-05,
+      "loss": 1.0843,
+      "step": 39000
+    },
+    {
+      "epoch": 0.48486485159453024,
+      "grad_norm": 9.27540397644043,
+      "learning_rate": 4.191891914009117e-05,
+      "loss": 1.1053,
+      "step": 39500
+    },
+    {
+      "epoch": 0.4910023813615496,
+      "grad_norm": 3.5075111389160156,
+      "learning_rate": 4.181662697730751e-05,
+      "loss": 1.092,
+      "step": 40000
+    },
+    {
+      "epoch": 0.497139911128569,
+      "grad_norm": 6.219689846038818,
+      "learning_rate": 4.171433481452385e-05,
+      "loss": 1.081,
+      "step": 40500
+    },
+    {
+      "epoch": 0.5032774408955883,
+      "grad_norm": 16.255443572998047,
+      "learning_rate": 4.161204265174019e-05,
+      "loss": 1.0919,
+      "step": 41000
+    },
+    {
+      "epoch": 0.5094149706626078,
+      "grad_norm": 32.27680206298828,
+      "learning_rate": 4.150975048895654e-05,
+      "loss": 1.0681,
+      "step": 41500
+    },
+    {
+      "epoch": 0.5155525004296271,
+      "grad_norm": 107.78765869140625,
+      "learning_rate": 4.1407458326172884e-05,
+      "loss": 1.0826,
+      "step": 42000
+    },
+    {
+      "epoch": 0.5216900301966465,
+      "grad_norm": 8.029512405395508,
+      "learning_rate": 4.1305166163389226e-05,
+      "loss": 1.0809,
+      "step": 42500
+    },
+    {
+      "epoch": 0.5278275599636658,
+      "grad_norm": 44.91604232788086,
+      "learning_rate": 4.1202874000605575e-05,
+      "loss": 1.093,
+      "step": 43000
+    },
+    {
+      "epoch": 0.5339650897306852,
+      "grad_norm": 11.860628128051758,
+      "learning_rate": 4.110058183782192e-05,
+      "loss": 1.0709,
+      "step": 43500
+    },
+    {
+      "epoch": 0.5401026194977046,
+      "grad_norm": 30.72344398498535,
+      "learning_rate": 4.099828967503826e-05,
+      "loss": 1.0623,
+      "step": 44000
+    },
+    {
+      "epoch": 0.546240149264724,
+      "grad_norm": 19.160911560058594,
+      "learning_rate": 4.08959975122546e-05,
+      "loss": 1.0801,
+      "step": 44500
+    },
+    {
+      "epoch": 0.5523776790317433,
+      "grad_norm": 12.648110389709473,
+      "learning_rate": 4.079370534947094e-05,
+      "loss": 1.0833,
+      "step": 45000
+    },
+    {
+      "epoch": 0.5585152087987627,
+      "grad_norm": 25.488170623779297,
+      "learning_rate": 4.0691413186687285e-05,
+      "loss": 1.0816,
+      "step": 45500
+    },
+    {
+      "epoch": 0.564652738565782,
+      "grad_norm": 8.982975006103516,
+      "learning_rate": 4.0589121023903634e-05,
+      "loss": 1.0697,
+      "step": 46000
+    },
+    {
+      "epoch": 0.5707902683328014,
+      "grad_norm": 12.776642799377441,
+      "learning_rate": 4.048682886111998e-05,
+      "loss": 1.0864,
+      "step": 46500
+    },
+    {
+      "epoch": 0.5769277980998208,
+      "grad_norm": 70.14471435546875,
+      "learning_rate": 4.0384536698336325e-05,
+      "loss": 1.0744,
+      "step": 47000
+    },
+    {
+      "epoch": 0.5830653278668402,
+      "grad_norm": 26.72732162475586,
+      "learning_rate": 4.028224453555267e-05,
+      "loss": 1.0897,
+      "step": 47500
+    },
+    {
+      "epoch": 0.5892028576338595,
+      "grad_norm": 22.69961166381836,
+      "learning_rate": 4.017995237276901e-05,
+      "loss": 1.0727,
+      "step": 48000
+    },
+    {
+      "epoch": 0.5953403874008789,
+      "grad_norm": 6.76914119720459,
+      "learning_rate": 4.007766020998535e-05,
+      "loss": 1.0621,
+      "step": 48500
+    },
+    {
+      "epoch": 0.6014779171678982,
+      "grad_norm": 41.32302474975586,
+      "learning_rate": 3.99753680472017e-05,
+      "loss": 1.0582,
+      "step": 49000
+    },
+    {
+      "epoch": 0.6076154469349176,
+      "grad_norm": 47.626644134521484,
+      "learning_rate": 3.987307588441804e-05,
+      "loss": 1.0681,
+      "step": 49500
+    },
+    {
+      "epoch": 0.613752976701937,
+      "grad_norm": 27.24641227722168,
+      "learning_rate": 3.9770783721634384e-05,
+      "loss": 1.083,
+      "step": 50000
+    },
+    {
+      "epoch": 0.6198905064689564,
+      "grad_norm": 22.27688217163086,
+      "learning_rate": 3.9668491558850726e-05,
+      "loss": 1.0632,
+      "step": 50500
+    },
+    {
+      "epoch": 0.6260280362359757,
+      "grad_norm": 43.488407135009766,
+      "learning_rate": 3.9566199396067075e-05,
+      "loss": 1.0809,
+      "step": 51000
+    },
+    {
+      "epoch": 0.6321655660029951,
+      "grad_norm": 8.742218017578125,
+      "learning_rate": 3.946390723328342e-05,
+      "loss": 1.0525,
+      "step": 51500
+    },
+    {
+      "epoch": 0.6383030957700145,
+      "grad_norm": 35.785621643066406,
+      "learning_rate": 3.936161507049976e-05,
+      "loss": 1.6649,
+      "step": 52000
+    },
+    {
+      "epoch": 0.6444406255370339,
+      "grad_norm": 17.245643615722656,
+      "learning_rate": 3.925932290771611e-05,
+      "loss": 1.0873,
+      "step": 52500
+    },
+    {
+      "epoch": 0.6505781553040533,
+      "grad_norm": 11.474211692810059,
+      "learning_rate": 3.915703074493245e-05,
+      "loss": 1.0649,
+      "step": 53000
+    },
+    {
+      "epoch": 0.6567156850710726,
+      "grad_norm": 57.05070114135742,
+      "learning_rate": 3.905473858214879e-05,
+      "loss": 1.0591,
+      "step": 53500
+    },
+    {
+      "epoch": 0.662853214838092,
+      "grad_norm": 28.446794509887695,
+      "learning_rate": 3.8952446419365134e-05,
+      "loss": 1.061,
+      "step": 54000
+    },
+    {
+      "epoch": 0.6689907446051113,
+      "grad_norm": 34.050048828125,
+      "learning_rate": 3.8850154256581476e-05,
+      "loss": 1.0682,
+      "step": 54500
+    },
+    {
+      "epoch": 0.6751282743721307,
+      "grad_norm": 40.37173843383789,
+      "learning_rate": 3.874786209379782e-05,
+      "loss": 1.0616,
+      "step": 55000
+    },
+    {
+      "epoch": 0.68126580413915,
+      "grad_norm": 17.34925079345703,
+      "learning_rate": 3.864556993101417e-05,
+      "loss": 1.0827,
+      "step": 55500
+    },
+    {
+      "epoch": 0.6874033339061695,
+      "grad_norm": 57.942806243896484,
+      "learning_rate": 3.8543277768230516e-05,
+      "loss": 1.0799,
+      "step": 56000
+    },
+    {
+      "epoch": 0.6935408636731888,
+      "grad_norm": 27.28282928466797,
+      "learning_rate": 3.844098560544686e-05,
+      "loss": 1.0705,
+      "step": 56500
+    },
+    {
+      "epoch": 0.6996783934402082,
+      "grad_norm": 34.14970016479492,
+      "learning_rate": 3.83386934426632e-05,
+      "loss": 1.0821,
+      "step": 57000
+    },
+    {
+      "epoch": 0.7058159232072275,
+      "grad_norm": 24.639619827270508,
+      "learning_rate": 3.823640127987954e-05,
+      "loss": 1.0763,
+      "step": 57500
+    },
+    {
+      "epoch": 0.711953452974247,
+      "grad_norm": 45.23350524902344,
+      "learning_rate": 3.8134109117095884e-05,
+      "loss": 1.0842,
+      "step": 58000
+    },
+    {
+      "epoch": 0.7180909827412663,
+      "grad_norm": 10.940398216247559,
+      "learning_rate": 3.803181695431223e-05,
+      "loss": 1.0813,
+      "step": 58500
+    },
+    {
+      "epoch": 0.7242285125082857,
+      "grad_norm": 95.62271881103516,
+      "learning_rate": 3.7929524791528575e-05,
+      "loss": 1.0678,
+      "step": 59000
+    },
+    {
+      "epoch": 0.730366042275305,
+      "grad_norm": 39.37797164916992,
+      "learning_rate": 3.782723262874492e-05,
+      "loss": 1.0894,
+      "step": 59500
+    },
+    {
+      "epoch": 0.7365035720423244,
+      "grad_norm": 14.900856971740723,
+      "learning_rate": 3.772494046596126e-05,
+      "loss": 1.0733,
+      "step": 60000
+    },
+    {
+      "epoch": 0.7426411018093437,
+      "grad_norm": 10.382057189941406,
+      "learning_rate": 3.762264830317761e-05,
+      "loss": 1.0688,
+      "step": 60500
+    },
+    {
+      "epoch": 0.7487786315763632,
+      "grad_norm": 7.631415367126465,
+      "learning_rate": 3.752035614039395e-05,
+      "loss": 1.0665,
+      "step": 61000
+    },
+    {
+      "epoch": 0.7549161613433825,
+      "grad_norm": 38.29065704345703,
+      "learning_rate": 3.74180639776103e-05,
+      "loss": 1.0681,
+      "step": 61500
+    },
+    {
+      "epoch": 0.7610536911104019,
+      "grad_norm": 44.250144958496094,
+      "learning_rate": 3.731577181482664e-05,
+      "loss": 1.301,
+      "step": 62000
+    },
+    {
+      "epoch": 0.7671912208774213,
+      "grad_norm": 3.7524375915527344,
+      "learning_rate": 3.721347965204298e-05,
+      "loss": 1.0907,
+      "step": 62500
+    },
+    {
+      "epoch": 0.7733287506444406,
+      "grad_norm": 41.10980987548828,
+      "learning_rate": 3.7111187489259325e-05,
+      "loss": 1.3941,
+      "step": 63000
+    },
+    {
+      "epoch": 0.77946628041146,
+      "grad_norm": 70.5405044555664,
+      "learning_rate": 3.700889532647567e-05,
+      "loss": 1.1355,
+      "step": 63500
+    },
+    {
+      "epoch": 0.7856038101784794,
+      "grad_norm": 84.906005859375,
+      "learning_rate": 3.690660316369201e-05,
+      "loss": 1.2196,
+      "step": 64000
+    },
+    {
+      "epoch": 0.7917413399454988,
+      "grad_norm": 86.1518325805664,
+      "learning_rate": 3.680431100090835e-05,
+      "loss": 1.225,
+      "step": 64500
+    },
+    {
+      "epoch": 0.7978788697125181,
+      "grad_norm": 15.497425079345703,
+      "learning_rate": 3.67020188381247e-05,
+      "loss": 1.1437,
+      "step": 65000
+    },
+    {
+      "epoch": 0.8040163994795375,
+      "grad_norm": 77.67127227783203,
+      "learning_rate": 3.659972667534105e-05,
+      "loss": 1.0787,
+      "step": 65500
+    },
+    {
+      "epoch": 0.8101539292465568,
+      "grad_norm": 54.82986068725586,
+      "learning_rate": 3.649743451255739e-05,
+      "loss": 1.0686,
+      "step": 66000
+    },
+    {
+      "epoch": 0.8162914590135762,
+      "grad_norm": 16.840795516967773,
+      "learning_rate": 3.639514234977373e-05,
+      "loss": 1.1017,
+      "step": 66500
+    },
+    {
+      "epoch": 0.8224289887805956,
+      "grad_norm": 9.104973793029785,
+      "learning_rate": 3.6292850186990075e-05,
+      "loss": 1.0999,
+      "step": 67000
+    },
+    {
+      "epoch": 0.828566518547615,
+      "grad_norm": 21.70631980895996,
+      "learning_rate": 3.6190558024206417e-05,
+      "loss": 1.0771,
+      "step": 67500
+    },
+    {
+      "epoch": 0.8347040483146343,
+      "grad_norm": 26.37040901184082,
+      "learning_rate": 3.6088265861422765e-05,
+      "loss": 1.1015,
+      "step": 68000
+    },
+    {
+      "epoch": 0.8408415780816537,
+      "grad_norm": 17.264562606811523,
+      "learning_rate": 3.598597369863911e-05,
+      "loss": 1.0826,
+      "step": 68500
+    },
+    {
+      "epoch": 0.846979107848673,
+      "grad_norm": 46.9713249206543,
+      "learning_rate": 3.588368153585545e-05,
+      "loss": 1.1046,
+      "step": 69000
+    },
+    {
+      "epoch": 0.8531166376156925,
+      "grad_norm": 21.621822357177734,
+      "learning_rate": 3.578138937307179e-05,
+      "loss": 1.0735,
+      "step": 69500
+    },
+    {
+      "epoch": 0.8592541673827118,
+      "grad_norm": 23.59171485900879,
+      "learning_rate": 3.567909721028814e-05,
+      "loss": 1.1056,
+      "step": 70000
+    },
+    {
+      "epoch": 0.8653916971497312,
+      "grad_norm": 52.57392883300781,
+      "learning_rate": 3.557680504750448e-05,
+      "loss": 1.1077,
+      "step": 70500
+    },
+    {
+      "epoch": 0.8715292269167505,
+      "grad_norm": 4.959003448486328,
+      "learning_rate": 3.547451288472083e-05,
+      "loss": 1.0897,
+      "step": 71000
+    },
+    {
+      "epoch": 0.8776667566837699,
+      "grad_norm": 10.618963241577148,
+      "learning_rate": 3.537222072193717e-05,
+      "loss": 1.0775,
+      "step": 71500
+    },
+    {
+      "epoch": 0.8838042864507892,
+      "grad_norm": 34.424346923828125,
+      "learning_rate": 3.5269928559153515e-05,
+      "loss": 1.0907,
+      "step": 72000
+    },
+    {
+      "epoch": 0.8899418162178087,
+      "grad_norm": 16.29790496826172,
+      "learning_rate": 3.516763639636986e-05,
+      "loss": 1.0705,
+      "step": 72500
+    },
+    {
+      "epoch": 0.8960793459848281,
+      "grad_norm": 39.85658645629883,
+      "learning_rate": 3.50653442335862e-05,
+      "loss": 1.0776,
+      "step": 73000
+    },
+    {
+      "epoch": 0.9022168757518474,
+      "grad_norm": 42.42219161987305,
+      "learning_rate": 3.496305207080254e-05,
+      "loss": 1.0896,
+      "step": 73500
+    },
+    {
+      "epoch": 0.9083544055188668,
+      "grad_norm": 67.0084228515625,
+      "learning_rate": 3.4860759908018884e-05,
+      "loss": 1.0889,
+      "step": 74000
+    },
+    {
+      "epoch": 0.9144919352858861,
+      "grad_norm": 39.18657684326172,
+      "learning_rate": 3.475846774523523e-05,
+      "loss": 1.0804,
+      "step": 74500
+    },
+    {
+      "epoch": 0.9206294650529055,
+      "grad_norm": 76.6010513305664,
+      "learning_rate": 3.4656175582451574e-05,
+      "loss": 1.1087,
+      "step": 75000
+    },
+    {
+      "epoch": 0.9267669948199249,
+      "grad_norm": 7.220675945281982,
+      "learning_rate": 3.455388341966792e-05,
+      "loss": 1.0738,
+      "step": 75500
+    },
+    {
+      "epoch": 0.9329045245869443,
+      "grad_norm": 32.431453704833984,
+      "learning_rate": 3.4451591256884265e-05,
+      "loss": 1.0806,
+      "step": 76000
+    },
+    {
+      "epoch": 0.9390420543539636,
+      "grad_norm": 89.43970489501953,
+      "learning_rate": 3.434929909410061e-05,
+      "loss": 1.0899,
+      "step": 76500
+    },
+    {
+      "epoch": 0.945179584120983,
+      "grad_norm": 25.720674514770508,
+      "learning_rate": 3.424700693131695e-05,
+      "loss": 1.0814,
+      "step": 77000
+    },
+    {
+      "epoch": 0.9513171138880023,
+      "grad_norm": 44.011478424072266,
+      "learning_rate": 3.41447147685333e-05,
+      "loss": 1.0723,
+      "step": 77500
+    },
+    {
+      "epoch": 0.9574546436550218,
+      "grad_norm": 67.16317749023438,
+      "learning_rate": 3.404242260574964e-05,
+      "loss": 1.0923,
+      "step": 78000
+    },
+    {
+      "epoch": 0.9635921734220411,
+      "grad_norm": 116.56925201416016,
+      "learning_rate": 3.394013044296598e-05,
+      "loss": 1.0748,
+      "step": 78500
+    },
+    {
+      "epoch": 0.9697297031890605,
+      "grad_norm": 99.30931091308594,
+      "learning_rate": 3.3837838280182324e-05,
+      "loss": 1.0745,
+      "step": 79000
+    },
+    {
+      "epoch": 0.9758672329560798,
+      "grad_norm": 60.88938522338867,
+      "learning_rate": 3.3735546117398667e-05,
+      "loss": 1.081,
+      "step": 79500
+    },
+    {
+      "epoch": 0.9820047627230992,
+      "grad_norm": 38.99333572387695,
+      "learning_rate": 3.3633253954615015e-05,
+      "loss": 1.08,
+      "step": 80000
+    },
+    {
+      "epoch": 0.9881422924901185,
+      "grad_norm": 125.93304443359375,
+      "learning_rate": 3.3530961791831364e-05,
+      "loss": 1.0905,
+      "step": 80500
+    },
+    {
+      "epoch": 0.994279822257138,
+      "grad_norm": 39.0662841796875,
+      "learning_rate": 3.3428669629047706e-05,
+      "loss": 1.1064,
+      "step": 81000
+    },
+    {
+      "epoch": 1.0004173520241573,
+      "grad_norm": 105.9404296875,
+      "learning_rate": 3.332637746626405e-05,
+      "loss": 1.0929,
+      "step": 81500
+    },
+    {
+      "epoch": 1.0065548817911767,
+      "grad_norm": 74.30028533935547,
+      "learning_rate": 3.322408530348039e-05,
+      "loss": 1.0815,
+      "step": 82000
+    },
+    {
+      "epoch": 1.0126924115581961,
+      "grad_norm": 14.825672149658203,
+      "learning_rate": 3.312179314069673e-05,
+      "loss": 1.0768,
+      "step": 82500
+    },
+    {
+      "epoch": 1.0188299413252153,
+      "grad_norm": 58.09488296508789,
+      "learning_rate": 3.3019500977913074e-05,
+      "loss": 1.1004,
+      "step": 83000
+    },
+    {
+      "epoch": 1.0249674710922347,
+      "grad_norm": 33.13391876220703,
+      "learning_rate": 3.2917208815129416e-05,
+      "loss": 1.0835,
+      "step": 83500
+    },
+    {
+      "epoch": 1.0311050008592542,
+      "grad_norm": 23.691511154174805,
+      "learning_rate": 3.2814916652345765e-05,
+      "loss": 1.0765,
+      "step": 84000
+    },
+    {
+      "epoch": 1.0372425306262736,
+      "grad_norm": 57.42161178588867,
+      "learning_rate": 3.271262448956211e-05,
+      "loss": 1.0906,
+      "step": 84500
+    },
+    {
+      "epoch": 1.043380060393293,
+      "grad_norm": 52.550350189208984,
+      "learning_rate": 3.2610332326778456e-05,
+      "loss": 1.096,
+      "step": 85000
+    },
+    {
+      "epoch": 1.0495175901603122,
+      "grad_norm": 89.49881744384766,
+      "learning_rate": 3.25080401639948e-05,
+      "loss": 1.1085,
+      "step": 85500
+    },
+    {
+      "epoch": 1.0556551199273316,
+      "grad_norm": 11.250481605529785,
+      "learning_rate": 3.240574800121114e-05,
+      "loss": 1.0913,
+      "step": 86000
+    },
+    {
+      "epoch": 1.061792649694351,
+      "grad_norm": 31.0516357421875,
+      "learning_rate": 3.230345583842748e-05,
+      "loss": 1.0974,
+      "step": 86500
+    },
+    {
+      "epoch": 1.0679301794613705,
+      "grad_norm": 94.23751068115234,
+      "learning_rate": 3.220116367564383e-05,
+      "loss": 1.0763,
+      "step": 87000
+    },
+    {
+      "epoch": 1.0740677092283897,
+      "grad_norm": 26.741689682006836,
+      "learning_rate": 3.209887151286017e-05,
+      "loss": 1.0894,
+      "step": 87500
+    },
+    {
+      "epoch": 1.080205238995409,
+      "grad_norm": 39.82439422607422,
+      "learning_rate": 3.1996579350076515e-05,
+      "loss": 1.1065,
+      "step": 88000
+    },
+    {
+      "epoch": 1.0863427687624285,
+      "grad_norm": 7.871860027313232,
+      "learning_rate": 3.189428718729286e-05,
+      "loss": 1.0898,
+      "step": 88500
+    },
+    {
+      "epoch": 1.092480298529448,
+      "grad_norm": 35.93474578857422,
+      "learning_rate": 3.17919950245092e-05,
+      "loss": 1.1036,
+      "step": 89000
+    },
+    {
+      "epoch": 1.0986178282964671,
+      "grad_norm": 146.91851806640625,
+      "learning_rate": 3.168970286172555e-05,
+      "loss": 1.0825,
+      "step": 89500
+    },
+    {
+      "epoch": 1.1047553580634866,
+      "grad_norm": 187.8535614013672,
+      "learning_rate": 3.15874106989419e-05,
+      "loss": 1.1164,
+      "step": 90000
+    },
+    {
+      "epoch": 1.110892887830506,
+      "grad_norm": 77.92440795898438,
+      "learning_rate": 3.148511853615824e-05,
+      "loss": 1.0811,
+      "step": 90500
+    },
+    {
+      "epoch": 1.1170304175975254,
+      "grad_norm": 62.233360290527344,
+      "learning_rate": 3.138282637337458e-05,
+      "loss": 1.115,
+      "step": 91000
+    },
+    {
+      "epoch": 1.1231679473645446,
+      "grad_norm": 54.17692947387695,
+      "learning_rate": 3.128053421059092e-05,
+      "loss": 1.1123,
+      "step": 91500
+    },
+    {
+      "epoch": 1.129305477131564,
+      "grad_norm": 49.207454681396484,
+      "learning_rate": 3.1178242047807265e-05,
+      "loss": 1.0846,
+      "step": 92000
+    },
+    {
+      "epoch": 1.1354430068985835,
+      "grad_norm": 49.67575454711914,
+      "learning_rate": 3.107594988502361e-05,
+      "loss": 1.0917,
+      "step": 92500
+    },
+    {
+      "epoch": 1.1415805366656029,
+      "grad_norm": 154.799072265625,
+      "learning_rate": 3.0973657722239956e-05,
+      "loss": 1.0879,
+      "step": 93000
+    },
+    {
+      "epoch": 1.1477180664326223,
+      "grad_norm": 146.50338745117188,
+      "learning_rate": 3.08713655594563e-05,
+      "loss": 1.0969,
+      "step": 93500
+    },
+    {
+      "epoch": 1.1538555961996415,
+      "grad_norm": 155.8907928466797,
+      "learning_rate": 3.076907339667264e-05,
+      "loss": 1.0849,
+      "step": 94000
+    },
+    {
+      "epoch": 1.159993125966661,
+      "grad_norm": 26.285480499267578,
+      "learning_rate": 3.066678123388899e-05,
+      "loss": 1.0852,
+      "step": 94500
+    },
+    {
+      "epoch": 1.1661306557336804,
+      "grad_norm": 31.642507553100586,
+      "learning_rate": 3.056448907110533e-05,
+      "loss": 1.0774,
+      "step": 95000
+    },
+    {
+      "epoch": 1.1722681855006996,
+      "grad_norm": 19.160383224487305,
+      "learning_rate": 3.0462196908321673e-05,
+      "loss": 1.0984,
+      "step": 95500
+    },
+    {
+      "epoch": 1.178405715267719,
+      "grad_norm": 38.4587516784668,
+      "learning_rate": 3.0359904745538015e-05,
+      "loss": 1.0936,
+      "step": 96000
+    },
+    {
+      "epoch": 1.1845432450347384,
+      "grad_norm": 54.26088333129883,
+      "learning_rate": 3.0257612582754364e-05,
+      "loss": 1.0842,
+      "step": 96500
+    },
+    {
+      "epoch": 1.1906807748017578,
+      "grad_norm": 22.72319221496582,
+      "learning_rate": 3.0155320419970706e-05,
+      "loss": 1.0895,
+      "step": 97000
+    },
+    {
+      "epoch": 1.1968183045687772,
+      "grad_norm": 88.04389190673828,
+      "learning_rate": 3.0053028257187048e-05,
+      "loss": 1.09,
+      "step": 97500
+    },
+    {
+      "epoch": 1.2029558343357964,
+      "grad_norm": 49.89582061767578,
+      "learning_rate": 2.9950736094403394e-05,
+      "loss": 1.0813,
+      "step": 98000
+    },
+    {
+      "epoch": 1.2090933641028159,
+      "grad_norm": 107.11421966552734,
+      "learning_rate": 2.9848443931619736e-05,
+      "loss": 1.0965,
+      "step": 98500
+    },
+    {
+      "epoch": 1.2152308938698353,
+      "grad_norm": 9.076912879943848,
+      "learning_rate": 2.9746151768836078e-05,
+      "loss": 1.1017,
+      "step": 99000
+    },
+    {
+      "epoch": 1.2213684236368547,
+      "grad_norm": 64.24601745605469,
+      "learning_rate": 2.9643859606052427e-05,
+      "loss": 1.1045,
+      "step": 99500
+    },
+    {
+      "epoch": 1.227505953403874,
+      "grad_norm": 38.064544677734375,
+      "learning_rate": 2.954156744326877e-05,
+      "loss": 1.093,
+      "step": 100000
+    },
+    {
+      "epoch": 1.2336434831708933,
+      "grad_norm": 98.80583190917969,
+      "learning_rate": 2.9439275280485114e-05,
+      "loss": 1.0903,
+      "step": 100500
+    },
+    {
+      "epoch": 1.2397810129379128,
+      "grad_norm": 78.89453125,
+      "learning_rate": 2.9336983117701456e-05,
+      "loss": 1.1133,
+      "step": 101000
+    },
+    {
+      "epoch": 1.2459185427049322,
+      "grad_norm": 53.44337844848633,
+      "learning_rate": 2.9234690954917798e-05,
+      "loss": 1.0883,
+      "step": 101500
+    },
+    {
+      "epoch": 1.2520560724719516,
+      "grad_norm": 46.80719757080078,
+      "learning_rate": 2.913239879213414e-05,
+      "loss": 1.1192,
+      "step": 102000
+    },
+    {
+      "epoch": 1.2581936022389708,
+      "grad_norm": 126.20846557617188,
+      "learning_rate": 2.903010662935049e-05,
+      "loss": 1.0817,
+      "step": 102500
+    },
+    {
+      "epoch": 1.2643311320059902,
+      "grad_norm": 27.287694931030273,
+      "learning_rate": 2.892781446656683e-05,
+      "loss": 1.0822,
+      "step": 103000
+    },
+    {
+      "epoch": 1.2704686617730097,
+      "grad_norm": 20.588132858276367,
+      "learning_rate": 2.8825522303783176e-05,
+      "loss": 1.0915,
+      "step": 103500
+    },
+    {
+      "epoch": 1.2766061915400289,
+      "grad_norm": 113.87678527832031,
+      "learning_rate": 2.872323014099952e-05,
+      "loss": 1.1128,
+      "step": 104000
+    },
+    {
+      "epoch": 1.2827437213070483,
+      "grad_norm": 51.455162048339844,
+      "learning_rate": 2.862093797821586e-05,
+      "loss": 1.0786,
+      "step": 104500
+    },
+    {
+      "epoch": 1.2888812510740677,
+      "grad_norm": 5.032041072845459,
+      "learning_rate": 2.8518645815432206e-05,
+      "loss": 1.1101,
+      "step": 105000
+    },
+    {
+      "epoch": 1.2950187808410871,
+      "grad_norm": 165.85589599609375,
+      "learning_rate": 2.8416353652648548e-05,
+      "loss": 1.097,
+      "step": 105500
+    },
+    {
+      "epoch": 1.3011563106081065,
+      "grad_norm": 34.687225341796875,
+      "learning_rate": 2.8314061489864897e-05,
+      "loss": 1.095,
+      "step": 106000
+    },
+    {
+      "epoch": 1.3072938403751257,
+      "grad_norm": 38.05297088623047,
+      "learning_rate": 2.821176932708124e-05,
+      "loss": 1.0884,
+      "step": 106500
+    },
+    {
+      "epoch": 1.3134313701421452,
+      "grad_norm": 8.639577865600586,
+      "learning_rate": 2.810947716429758e-05,
+      "loss": 1.09,
+      "step": 107000
+    },
+    {
+      "epoch": 1.3195688999091646,
+      "grad_norm": 19.865270614624023,
+      "learning_rate": 2.8007185001513923e-05,
+      "loss": 1.1057,
+      "step": 107500
+    },
+    {
+      "epoch": 1.3257064296761838,
+      "grad_norm": 35.60298538208008,
+      "learning_rate": 2.790489283873027e-05,
+      "loss": 1.087,
+      "step": 108000
+    },
+    {
+      "epoch": 1.3318439594432032,
+      "grad_norm": 49.35670471191406,
+      "learning_rate": 2.780260067594661e-05,
+      "loss": 1.1009,
+      "step": 108500
+    },
+    {
+      "epoch": 1.3379814892102226,
+      "grad_norm": 14.54593276977539,
+      "learning_rate": 2.770030851316296e-05,
+      "loss": 1.0849,
+      "step": 109000
+    },
+    {
+      "epoch": 1.344119018977242,
+      "grad_norm": 187.47262573242188,
+      "learning_rate": 2.75980163503793e-05,
+      "loss": 1.0886,
+      "step": 109500
+    },
+    {
+      "epoch": 1.3502565487442615,
+      "grad_norm": 44.326168060302734,
+      "learning_rate": 2.7495724187595644e-05,
+      "loss": 1.0805,
+      "step": 110000
+    },
+    {
+      "epoch": 1.356394078511281,
+      "grad_norm": 136.01849365234375,
+      "learning_rate": 2.739343202481199e-05,
+      "loss": 1.0808,
+      "step": 110500
+    },
+    {
+      "epoch": 1.3625316082783,
+      "grad_norm": 157.3839111328125,
+      "learning_rate": 2.729113986202833e-05,
+      "loss": 1.1025,
+      "step": 111000
+    },
+    {
+      "epoch": 1.3686691380453195,
+      "grad_norm": 64.22801208496094,
+      "learning_rate": 2.7188847699244673e-05,
+      "loss": 1.0955,
+      "step": 111500
+    },
+    {
+      "epoch": 1.374806667812339,
+      "grad_norm": 19.307771682739258,
+      "learning_rate": 2.7086555536461022e-05,
+      "loss": 1.0824,
+      "step": 112000
+    },
+    {
+      "epoch": 1.3809441975793582,
+      "grad_norm": 14.489301681518555,
+      "learning_rate": 2.6984263373677364e-05,
+      "loss": 1.0835,
+      "step": 112500
+    },
+    {
+      "epoch": 1.3870817273463776,
+      "grad_norm": 199.36927795410156,
+      "learning_rate": 2.688197121089371e-05,
+      "loss": 1.1168,
+      "step": 113000
+    },
+    {
+      "epoch": 1.393219257113397,
+      "grad_norm": 148.1888427734375,
+      "learning_rate": 2.677967904811005e-05,
+      "loss": 1.0881,
+      "step": 113500
+    },
+    {
+      "epoch": 1.3993567868804164,
+      "grad_norm": 44.604610443115234,
+      "learning_rate": 2.6677386885326393e-05,
+      "loss": 1.0946,
+      "step": 114000
+    },
+    {
+      "epoch": 1.4054943166474358,
+      "grad_norm": 40.0570068359375,
+      "learning_rate": 2.6575094722542736e-05,
+      "loss": 1.0819,
+      "step": 114500
+    },
+    {
+      "epoch": 1.411631846414455,
+      "grad_norm": 54.24287033081055,
+      "learning_rate": 2.647280255975908e-05,
+      "loss": 1.1155,
+      "step": 115000
+    },
+    {
+      "epoch": 1.4177693761814745,
+      "grad_norm": 160.6737060546875,
+      "learning_rate": 2.637051039697543e-05,
+      "loss": 1.1021,
+      "step": 115500
+    },
+    {
+      "epoch": 1.423906905948494,
+      "grad_norm": 189.32138061523438,
+      "learning_rate": 2.6268218234191772e-05,
+      "loss": 1.102,
+      "step": 116000
+    },
+    {
+      "epoch": 1.430044435715513,
+      "grad_norm": 73.93120574951172,
+      "learning_rate": 2.6165926071408114e-05,
+      "loss": 1.0733,
+      "step": 116500
+    },
+    {
+      "epoch": 1.4361819654825325,
+      "grad_norm": 13.932085037231445,
+      "learning_rate": 2.6063633908624456e-05,
+      "loss": 1.0987,
+      "step": 117000
+    },
+    {
+      "epoch": 1.442319495249552,
+      "grad_norm": 171.73492431640625,
+      "learning_rate": 2.59613417458408e-05,
+      "loss": 1.1103,
+      "step": 117500
+    },
+    {
+      "epoch": 1.4484570250165714,
+      "grad_norm": 63.9893798828125,
+      "learning_rate": 2.5859049583057143e-05,
+      "loss": 1.1034,
+      "step": 118000
+    },
+    {
+      "epoch": 1.4545945547835908,
+      "grad_norm": 45.00864028930664,
+      "learning_rate": 2.5756757420273492e-05,
+      "loss": 1.0987,
+      "step": 118500
+    },
+    {
+      "epoch": 1.46073208455061,
+      "grad_norm": 21.32621192932129,
+      "learning_rate": 2.5654465257489834e-05,
+      "loss": 1.0908,
+      "step": 119000
+    },
+    {
+      "epoch": 1.4668696143176294,
+      "grad_norm": 80.76024627685547,
+      "learning_rate": 2.5552173094706176e-05,
+      "loss": 1.0986,
+      "step": 119500
+    },
+    {
+      "epoch": 1.4730071440846488,
+      "grad_norm": 34.15319061279297,
+      "learning_rate": 2.5449880931922522e-05,
+      "loss": 1.0988,
+      "step": 120000
+    },
+    {
+      "epoch": 1.4791446738516683,
+      "grad_norm": 41.0086784362793,
+      "learning_rate": 2.5347588769138864e-05,
+      "loss": 1.1023,
+      "step": 120500
+    },
+    {
+      "epoch": 1.4852822036186875,
+      "grad_norm": 33.04220962524414,
+      "learning_rate": 2.5245296606355206e-05,
+      "loss": 1.1013,
+      "step": 121000
+    },
+    {
+      "epoch": 1.4914197333857069,
+      "grad_norm": 51.40751647949219,
+      "learning_rate": 2.5143004443571555e-05,
+      "loss": 1.0896,
+      "step": 121500
+    },
+    {
+      "epoch": 1.4975572631527263,
+      "grad_norm": 10.27536392211914,
+      "learning_rate": 2.5040712280787897e-05,
+      "loss": 1.8455,
+      "step": 122000
+    },
+    {
+      "epoch": 1.5036947929197457,
+      "grad_norm": 138.4582061767578,
+      "learning_rate": 2.4938420118004242e-05,
+      "loss": 1.1155,
+      "step": 122500
+    },
+    {
+      "epoch": 1.5098323226867651,
+      "grad_norm": 222.63157653808594,
+      "learning_rate": 2.4836127955220584e-05,
+      "loss": 1.1502,
+      "step": 123000
+    },
+    {
+      "epoch": 1.5159698524537844,
+      "grad_norm": 17.325204849243164,
+      "learning_rate": 2.4733835792436926e-05,
+      "loss": 1.1183,
+      "step": 123500
+    },
+    {
+      "epoch": 1.5221073822208038,
+      "grad_norm": 33.44583511352539,
+      "learning_rate": 2.4631543629653272e-05,
+      "loss": 1.0958,
+      "step": 124000
+    },
+    {
+      "epoch": 1.5282449119878232,
+      "grad_norm": 36.93500900268555,
+      "learning_rate": 2.4529251466869614e-05,
+      "loss": 1.1098,
+      "step": 124500
+    },
+    {
+      "epoch": 1.5343824417548424,
+      "grad_norm": 24.32937240600586,
+      "learning_rate": 2.442695930408596e-05,
+      "loss": 1.1021,
+      "step": 125000
+    },
+    {
+      "epoch": 1.5405199715218618,
+      "grad_norm": 138.19097900390625,
+      "learning_rate": 2.4324667141302305e-05,
+      "loss": 1.0912,
+      "step": 125500
+    },
+    {
+      "epoch": 1.5466575012888812,
+      "grad_norm": 110.18413543701172,
+      "learning_rate": 2.4222374978518647e-05,
+      "loss": 1.0961,
+      "step": 126000
+    },
+    {
+      "epoch": 1.5527950310559007,
+      "grad_norm": 98.88742065429688,
+      "learning_rate": 2.412008281573499e-05,
+      "loss": 1.0858,
+      "step": 126500
+    },
+    {
+      "epoch": 1.55893256082292,
+      "grad_norm": 182.95082092285156,
+      "learning_rate": 2.4017790652951334e-05,
+      "loss": 1.0784,
+      "step": 127000
+    },
+    {
+      "epoch": 1.5650700905899395,
+      "grad_norm": 14.683859825134277,
+      "learning_rate": 2.391549849016768e-05,
+      "loss": 1.1112,
+      "step": 127500
+    },
+    {
+      "epoch": 1.5712076203569587,
+      "grad_norm": 93.70809173583984,
+      "learning_rate": 2.3813206327384022e-05,
+      "loss": 1.1067,
+      "step": 128000
+    },
+    {
+      "epoch": 1.5773451501239781,
+      "grad_norm": 230.3051300048828,
+      "learning_rate": 2.3710914164600367e-05,
+      "loss": 1.0986,
+      "step": 128500
+    },
+    {
+      "epoch": 1.5834826798909973,
+      "grad_norm": 72.11837768554688,
+      "learning_rate": 2.360862200181671e-05,
+      "loss": 1.0824,
+      "step": 129000
+    },
+    {
+      "epoch": 1.5896202096580168,
+      "grad_norm": 130.1708221435547,
+      "learning_rate": 2.3506329839033055e-05,
+      "loss": 1.1072,
+      "step": 129500
+    },
+    {
+      "epoch": 1.5957577394250362,
+      "grad_norm": 221.41827392578125,
+      "learning_rate": 2.3404037676249397e-05,
+      "loss": 1.1098,
+      "step": 130000
+    },
+    {
+      "epoch": 1.6018952691920556,
+      "grad_norm": 101.14270782470703,
+      "learning_rate": 2.3301745513465742e-05,
+      "loss": 1.0962,
+      "step": 130500
+    },
+    {
+      "epoch": 1.608032798959075,
+      "grad_norm": 68.84784698486328,
+      "learning_rate": 2.3199453350682084e-05,
+      "loss": 1.1108,
+      "step": 131000
+    },
+    {
+      "epoch": 1.6141703287260944,
+      "grad_norm": 92.93330383300781,
+      "learning_rate": 2.3097161187898426e-05,
+      "loss": 1.1187,
+      "step": 131500
+    },
+    {
+      "epoch": 1.6203078584931137,
+      "grad_norm": 17.309310913085938,
+      "learning_rate": 2.2994869025114775e-05,
+      "loss": 1.0923,
+      "step": 132000
+    },
+    {
+      "epoch": 1.626445388260133,
+      "grad_norm": 53.46000671386719,
+      "learning_rate": 2.2892576862331117e-05,
+      "loss": 1.1003,
+      "step": 132500
+    },
+    {
+      "epoch": 1.6325829180271523,
+      "grad_norm": 125.82015991210938,
+      "learning_rate": 2.279028469954746e-05,
+      "loss": 1.0865,
+      "step": 133000
+    },
+    {
+      "epoch": 1.6387204477941717,
+      "grad_norm": 103.5467758178711,
+      "learning_rate": 2.2687992536763805e-05,
+      "loss": 1.099,
+      "step": 133500
+    },
+    {
+      "epoch": 1.6448579775611911,
+      "grad_norm": 32.24409866333008,
+      "learning_rate": 2.2585700373980147e-05,
+      "loss": 1.0838,
+      "step": 134000
+    },
+    {
+      "epoch": 1.6509955073282105,
+      "grad_norm": 16.080642700195312,
+      "learning_rate": 2.2483408211196492e-05,
+      "loss": 1.0792,
+      "step": 134500
+    },
+    {
+      "epoch": 1.65713303709523,
+      "grad_norm": 127.77115631103516,
+      "learning_rate": 2.2381116048412838e-05,
+      "loss": 1.0966,
+      "step": 135000
+    },
+    {
+      "epoch": 1.6632705668622494,
+      "grad_norm": 124.41118621826172,
+      "learning_rate": 2.227882388562918e-05,
+      "loss": 1.0782,
+      "step": 135500
+    },
+    {
+      "epoch": 1.6694080966292688,
+      "grad_norm": 50.092193603515625,
+      "learning_rate": 2.2176531722845522e-05,
+      "loss": 1.1123,
+      "step": 136000
+    },
+    {
+      "epoch": 1.675545626396288,
+      "grad_norm": 42.088958740234375,
+      "learning_rate": 2.2074239560061867e-05,
+      "loss": 1.0923,
+      "step": 136500
+    },
+    {
+      "epoch": 1.6816831561633074,
+      "grad_norm": 147.70936584472656,
+      "learning_rate": 2.1971947397278213e-05,
+      "loss": 1.0873,
+      "step": 137000
+    },
+    {
+      "epoch": 1.6878206859303266,
+      "grad_norm": 23.003494262695312,
+      "learning_rate": 2.1869655234494555e-05,
+      "loss": 1.0807,
+      "step": 137500
+    },
+    {
+      "epoch": 1.693958215697346,
+      "grad_norm": 24.13724136352539,
+      "learning_rate": 2.17673630717109e-05,
+      "loss": 1.083,
+      "step": 138000
+    },
+    {
+      "epoch": 1.7000957454643655,
+      "grad_norm": 280.0939025878906,
+      "learning_rate": 2.1665070908927242e-05,
+      "loss": 1.0864,
+      "step": 138500
+    },
+    {
+      "epoch": 1.706233275231385,
+      "grad_norm": 32.00410842895508,
+      "learning_rate": 2.1562778746143588e-05,
+      "loss": 1.0828,
+      "step": 139000
+    },
+    {
+      "epoch": 1.7123708049984043,
+      "grad_norm": 7.247683525085449,
+      "learning_rate": 2.1460486583359933e-05,
+      "loss": 1.0973,
+      "step": 139500
+    },
+    {
+      "epoch": 1.7185083347654238,
+      "grad_norm": 44.28471374511719,
+      "learning_rate": 2.1358194420576275e-05,
+      "loss": 1.1022,
+      "step": 140000
+    },
+    {
+      "epoch": 1.724645864532443,
+      "grad_norm": 9.166769027709961,
+      "learning_rate": 2.1255902257792617e-05,
+      "loss": 1.0837,
+      "step": 140500
+    },
+    {
+      "epoch": 1.7307833942994624,
+      "grad_norm": 131.9899139404297,
+      "learning_rate": 2.115361009500896e-05,
+      "loss": 1.0985,
+      "step": 141000
+    },
+    {
+      "epoch": 1.7369209240664816,
+      "grad_norm": 7.971253395080566,
+      "learning_rate": 2.1051317932225305e-05,
+      "loss": 1.1049,
+      "step": 141500
+    },
+    {
+      "epoch": 1.743058453833501,
+      "grad_norm": 40.334327697753906,
+      "learning_rate": 2.094902576944165e-05,
+      "loss": 1.079,
+      "step": 142000
+    },
+    {
+      "epoch": 1.7491959836005204,
+      "grad_norm": 100.06372833251953,
+      "learning_rate": 2.0846733606657992e-05,
+      "loss": 1.0757,
+      "step": 142500
+    },
+    {
+      "epoch": 1.7553335133675398,
+      "grad_norm": 149.7917938232422,
+      "learning_rate": 2.0744441443874338e-05,
+      "loss": 1.0808,
+      "step": 143000
+    },
+    {
+      "epoch": 1.7614710431345593,
+      "grad_norm": 13.513212203979492,
+      "learning_rate": 2.064214928109068e-05,
+      "loss": 1.0743,
+      "step": 143500
+    },
+    {
+      "epoch": 1.7676085729015787,
+      "grad_norm": 47.62020492553711,
+      "learning_rate": 2.0539857118307025e-05,
+      "loss": 1.0933,
+      "step": 144000
+    },
+    {
+      "epoch": 1.773746102668598,
+      "grad_norm": 46.22666931152344,
+      "learning_rate": 2.043756495552337e-05,
+      "loss": 1.0938,
+      "step": 144500
+    },
+    {
+      "epoch": 1.7798836324356173,
+      "grad_norm": 17.870939254760742,
+      "learning_rate": 2.0335272792739713e-05,
+      "loss": 1.1121,
+      "step": 145000
+    },
+    {
+      "epoch": 1.7860211622026367,
+      "grad_norm": 267.29339599609375,
+      "learning_rate": 2.0232980629956055e-05,
+      "loss": 1.1138,
+      "step": 145500
+    },
+    {
+      "epoch": 1.792158691969656,
+      "grad_norm": 49.400753021240234,
+      "learning_rate": 2.01306884671724e-05,
+      "loss": 1.1063,
+      "step": 146000
+    },
+    {
+      "epoch": 1.7982962217366754,
+      "grad_norm": 45.64429473876953,
+      "learning_rate": 2.0028396304388745e-05,
+      "loss": 1.097,
+      "step": 146500
+    },
+    {
+      "epoch": 1.8044337515036948,
+      "grad_norm": 57.519508361816406,
+      "learning_rate": 1.9926104141605088e-05,
+      "loss": 1.0999,
+      "step": 147000
+    },
+    {
+      "epoch": 1.8105712812707142,
+      "grad_norm": 30.997135162353516,
+      "learning_rate": 1.9823811978821433e-05,
+      "loss": 1.1035,
+      "step": 147500
+    },
+    {
+      "epoch": 1.8167088110377336,
+      "grad_norm": 183.94775390625,
+      "learning_rate": 1.9721519816037775e-05,
+      "loss": 1.0786,
+      "step": 148000
+    },
+    {
+      "epoch": 1.822846340804753,
+      "grad_norm": 62.93962860107422,
+      "learning_rate": 1.9619227653254117e-05,
+      "loss": 1.0824,
+      "step": 148500
+    },
+    {
+      "epoch": 1.8289838705717723,
+      "grad_norm": 26.8454532623291,
+      "learning_rate": 1.9516935490470466e-05,
+      "loss": 1.1097,
+      "step": 149000
+    },
+    {
+      "epoch": 1.8351214003387917,
+      "grad_norm": 103.21609497070312,
+      "learning_rate": 1.9414643327686808e-05,
+      "loss": 1.0744,
+      "step": 149500
+    },
+    {
+      "epoch": 1.8412589301058109,
+      "grad_norm": 12.109946250915527,
+      "learning_rate": 1.931235116490315e-05,
+      "loss": 1.0902,
+      "step": 150000
+    },
+    {
+      "epoch": 1.8473964598728303,
+      "grad_norm": 182.6466522216797,
+      "learning_rate": 1.9210059002119492e-05,
+      "loss": 1.0841,
+      "step": 150500
+    },
+    {
+      "epoch": 1.8535339896398497,
+      "grad_norm": 149.785400390625,
+      "learning_rate": 1.9107766839335837e-05,
+      "loss": 1.0961,
+      "step": 151000
+    },
+    {
+      "epoch": 1.8596715194068691,
+      "grad_norm": 76.48492431640625,
+      "learning_rate": 1.9005474676552183e-05,
+      "loss": 1.0778,
+      "step": 151500
+    },
+    {
+      "epoch": 1.8658090491738886,
+      "grad_norm": 101.30828857421875,
+      "learning_rate": 1.8903182513768525e-05,
+      "loss": 1.0784,
+      "step": 152000
+    },
+    {
+      "epoch": 1.871946578940908,
+      "grad_norm": 159.51016235351562,
+      "learning_rate": 1.880089035098487e-05,
+      "loss": 1.0741,
+      "step": 152500
+    },
+    {
+      "epoch": 1.8780841087079272,
+      "grad_norm": 116.58811950683594,
+      "learning_rate": 1.8698598188201212e-05,
+      "loss": 1.0879,
+      "step": 153000
+    },
+    {
+      "epoch": 1.8842216384749466,
+      "grad_norm": 59.34739303588867,
+      "learning_rate": 1.8596306025417558e-05,
+      "loss": 1.079,
+      "step": 153500
+    },
+    {
+      "epoch": 1.8903591682419658,
+      "grad_norm": 28.612041473388672,
+      "learning_rate": 1.8494013862633903e-05,
+      "loss": 1.0967,
+      "step": 154000
+    },
+    {
+      "epoch": 1.8964966980089852,
+      "grad_norm": 34.84067916870117,
+      "learning_rate": 1.8391721699850245e-05,
+      "loss": 1.0906,
+      "step": 154500
+    },
+    {
+      "epoch": 1.9026342277760047,
+      "grad_norm": 156.6701202392578,
+      "learning_rate": 1.8289429537066587e-05,
+      "loss": 1.0836,
+      "step": 155000
+    },
+    {
+      "epoch": 1.908771757543024,
+      "grad_norm": 11.411686897277832,
+      "learning_rate": 1.8187137374282933e-05,
+      "loss": 1.0932,
+      "step": 155500
+    },
+    {
+      "epoch": 1.9149092873100435,
+      "grad_norm": 188.55970764160156,
+      "learning_rate": 1.808484521149928e-05,
+      "loss": 1.0823,
+      "step": 156000
+    },
+    {
+      "epoch": 1.921046817077063,
+      "grad_norm": 18.585180282592773,
+      "learning_rate": 1.798255304871562e-05,
+      "loss": 1.087,
+      "step": 156500
+    },
+    {
+      "epoch": 1.9271843468440824,
+      "grad_norm": 398.6981201171875,
+      "learning_rate": 1.7880260885931966e-05,
+      "loss": 1.0892,
+      "step": 157000
+    },
+    {
+      "epoch": 1.9333218766111016,
+      "grad_norm": 6.879120826721191,
+      "learning_rate": 1.7777968723148308e-05,
+      "loss": 1.0842,
+      "step": 157500
+    },
+    {
+      "epoch": 1.939459406378121,
+      "grad_norm": 159.97511291503906,
+      "learning_rate": 1.767567656036465e-05,
+      "loss": 1.0837,
+      "step": 158000
+    },
+    {
+      "epoch": 1.9455969361451402,
+      "grad_norm": 57.745304107666016,
+      "learning_rate": 1.7573384397580995e-05,
+      "loss": 1.1001,
+      "step": 158500
+    },
+    {
+      "epoch": 1.9517344659121596,
+      "grad_norm": 80.60443115234375,
+      "learning_rate": 1.747109223479734e-05,
+      "loss": 1.0727,
+      "step": 159000
+    },
+    {
+      "epoch": 1.957871995679179,
+      "grad_norm": 164.41299438476562,
+      "learning_rate": 1.7368800072013683e-05,
+      "loss": 1.0875,
+      "step": 159500
+    },
+    {
+      "epoch": 1.9640095254461984,
+      "grad_norm": 126.25370025634766,
+      "learning_rate": 1.7266507909230028e-05,
+      "loss": 1.0845,
+      "step": 160000
+    },
+    {
+      "epoch": 1.9701470552132179,
+      "grad_norm": 46.17078399658203,
+      "learning_rate": 1.716421574644637e-05,
+      "loss": 1.0805,
+      "step": 160500
+    },
+    {
+      "epoch": 1.9762845849802373,
+      "grad_norm": 79.331787109375,
+      "learning_rate": 1.7061923583662716e-05,
+      "loss": 1.0825,
+      "step": 161000
+    },
+    {
+      "epoch": 1.9824221147472565,
+      "grad_norm": 23.304096221923828,
+      "learning_rate": 1.6959631420879058e-05,
+      "loss": 1.0886,
+      "step": 161500
+    },
+    {
+      "epoch": 1.988559644514276,
+      "grad_norm": 198.92140197753906,
+      "learning_rate": 1.6857339258095403e-05,
+      "loss": 1.0856,
+      "step": 162000
+    },
+    {
+      "epoch": 1.9946971742812951,
+      "grad_norm": 313.79461669921875,
+      "learning_rate": 1.6755047095311745e-05,
+      "loss": 1.0816,
+      "step": 162500
+    },
+    {
+      "epoch": 2.0008347040483145,
+      "grad_norm": 102.85086822509766,
+      "learning_rate": 1.6652754932528087e-05,
+      "loss": 1.1005,
+      "step": 163000
+    },
+    {
+      "epoch": 2.006972233815334,
+      "grad_norm": 89.94127655029297,
+      "learning_rate": 1.6550462769744436e-05,
+      "loss": 1.0775,
+      "step": 163500
+    },
+    {
+      "epoch": 2.0131097635823534,
+      "grad_norm": 74.4475326538086,
+      "learning_rate": 1.6448170606960778e-05,
+      "loss": 1.0875,
+      "step": 164000
+    },
+    {
+      "epoch": 2.019247293349373,
+      "grad_norm": 56.56633377075195,
+      "learning_rate": 1.634587844417712e-05,
+      "loss": 1.09,
+      "step": 164500
+    },
+    {
+      "epoch": 2.0253848231163922,
+      "grad_norm": 59.08357620239258,
+      "learning_rate": 1.6243586281393466e-05,
+      "loss": 1.086,
+      "step": 165000
+    },
+    {
+      "epoch": 2.0315223528834117,
+      "grad_norm": 126.40411376953125,
+      "learning_rate": 1.6141294118609808e-05,
+      "loss": 1.087,
+      "step": 165500
+    },
+    {
+      "epoch": 2.0376598826504306,
+      "grad_norm": 73.0719985961914,
+      "learning_rate": 1.6039001955826153e-05,
+      "loss": 1.0815,
+      "step": 166000
+    },
+    {
+      "epoch": 2.04379741241745,
+      "grad_norm": 70.06752014160156,
+      "learning_rate": 1.59367097930425e-05,
+      "loss": 1.0832,
+      "step": 166500
+    },
+    {
+      "epoch": 2.0499349421844695,
+      "grad_norm": 56.33417892456055,
+      "learning_rate": 1.583441763025884e-05,
+      "loss": 1.0801,
+      "step": 167000
+    },
+    {
+      "epoch": 2.056072471951489,
+      "grad_norm": 127.28369903564453,
+      "learning_rate": 1.5732125467475183e-05,
+      "loss": 1.0828,
+      "step": 167500
+    },
+    {
+      "epoch": 2.0622100017185083,
+      "grad_norm": 102.63399505615234,
+      "learning_rate": 1.5629833304691528e-05,
+      "loss": 1.0819,
+      "step": 168000
+    },
+    {
+      "epoch": 2.0683475314855277,
+      "grad_norm": 50.66084671020508,
+      "learning_rate": 1.5527541141907874e-05,
+      "loss": 1.0767,
+      "step": 168500
+    },
+    {
+      "epoch": 2.074485061252547,
+      "grad_norm": 162.90655517578125,
+      "learning_rate": 1.5425248979124216e-05,
+      "loss": 1.0819,
+      "step": 169000
+    },
+    {
+      "epoch": 2.0806225910195666,
+      "grad_norm": 41.8231201171875,
+      "learning_rate": 1.532295681634056e-05,
+      "loss": 1.1013,
+      "step": 169500
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 244398,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:320a99fb4b91f8525d40d71b170ac4e34508c4a26597a496c07d42c089137f1c
+size 5432