Upload folder using huggingface_hub

adapter/adapter_config.json

@@ -30,13 +30,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q_proj",
-    "up_proj",
-    "down_proj",
-    "v_proj",
     "gate_proj",
+    "up_proj",
     "k_proj",
-    "o_proj"
+    "q_proj",
+    "o_proj",
+    "down_proj",
+    "v_proj"
   ],
   "target_parameters": null,
   "task_type": "CAUSAL_LM",

adapter/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:682782976f5785f7c44acfb14220d76a1060a76654071117b7bac475d730828a
+oid sha256:265782c72aacf300145a6d734bbd94bc3046b5cb36990af055ab465db2f0ac26
 size 97307544

adapter/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9d4dd8fc367f395f33dce87624317a1d212493208257316e62e4d479a5df06ba
+oid sha256:704fe3bbe7daa27f92208067c08a7241e2a2d72436db7e16a6696b3df8210ffe
 size 194840426

adapter/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:71b8bfabdb652b608519c60f89a1b975863991f97c2154c3b9baad6e994fce65
+oid sha256:6f407eb5d94c48068ca433912c3c5abe9e0ce32994d0566eb748c467f25b7b2a
 size 14512

adapter/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b67b03ffe2f956a33a617f6a065ac0163e84842bbc176bf61cf1abe12cb3932d
+oid sha256:47e740f1493821c61eb2b8c06984dc0557ad040b70e27d8e26b31abefddd714c
 size 14512

adapter/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5eedaadfeb99ff8879563a70a5798226f64e15e6d536404ff905a34b36e914ef
+oid sha256:91681a23856743c3790ae2af065cde1392aa877097b576966b85db7daa758191
 size 1064

adapter/trainer_state.json

@@ -1,1352 +1,1546 @@
 {
-  "best_global_step": 900,
-  "best_metric": 7.772979259490967,
-  "best_model_checkpoint": "out/checkpoint-900",
-  "epoch": 15.794701986754967,
-  "eval_steps": 100,
-  "global_step": 900,
+  "best_global_step": 1050,
+  "best_metric": 7.780308246612549,
+  "best_model_checkpoint": "out/checkpoint-1050",
+  "epoch": 18.423841059602648,
+  "eval_steps": 150,
+  "global_step": 1050,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.08830022075055188,
-      "grad_norm": 8.491263389587402,
-      "learning_rate": 9.985964912280701e-05,
-      "loss": 13.351483154296876,
+      "grad_norm": 8.688450813293457,
+      "learning_rate": 9.976608187134502e-05,
+      "loss": 13.340193176269532,
       "step": 5
     },
     {
       "epoch": 0.17660044150110377,
-      "grad_norm": 4.82792329788208,
-      "learning_rate": 9.968421052631578e-05,
-      "loss": 10.233726501464844,
+      "grad_norm": 4.834760665893555,
+      "learning_rate": 9.947368421052632e-05,
+      "loss": 10.211357879638673,
       "step": 10
     },
     {
       "epoch": 0.26490066225165565,
-      "grad_norm": 1.351132869720459,
-      "learning_rate": 9.950877192982457e-05,
-      "loss": 8.952561950683593,
+      "grad_norm": 1.3143389225006104,
+      "learning_rate": 9.918128654970761e-05,
+      "loss": 8.949986267089844,
       "step": 15
     },
     {
       "epoch": 0.35320088300220753,
-      "grad_norm": 0.3755686283111572,
-      "learning_rate": 9.933333333333334e-05,
-      "loss": 8.484365081787109,
+      "grad_norm": 0.37686967849731445,
+      "learning_rate": 9.888888888888889e-05,
+      "loss": 8.484906005859376,
       "step": 20
     },
     {
       "epoch": 0.44150110375275936,
-      "grad_norm": 0.2917013466358185,
-      "learning_rate": 9.915789473684211e-05,
-      "loss": 8.363894653320312,
+      "grad_norm": 0.2962671220302582,
+      "learning_rate": 9.859649122807017e-05,
+      "loss": 8.364928436279296,
       "step": 25
     },
     {
       "epoch": 0.5298013245033113,
-      "grad_norm": 0.2885504364967346,
-      "learning_rate": 9.898245614035088e-05,
-      "loss": 8.33376693725586,
+      "grad_norm": 0.2862487733364105,
+      "learning_rate": 9.830409356725147e-05,
+      "loss": 8.333544921875,
       "step": 30
     },
     {
       "epoch": 0.6181015452538632,
-      "grad_norm": 0.2070295810699463,
-      "learning_rate": 9.880701754385965e-05,
-      "loss": 8.32794418334961,
+      "grad_norm": 0.20227763056755066,
+      "learning_rate": 9.801169590643276e-05,
+      "loss": 8.327781677246094,
       "step": 35
     },
     {
       "epoch": 0.7064017660044151,
-      "grad_norm": 0.4069546163082123,
-      "learning_rate": 9.863157894736842e-05,
-      "loss": 8.322712707519532,
+      "grad_norm": 0.39236980676651,
+      "learning_rate": 9.771929824561404e-05,
+      "loss": 8.322337341308593,
       "step": 40
     },
     {
       "epoch": 0.7947019867549668,
-      "grad_norm": 0.2271132916212082,
-      "learning_rate": 9.845614035087721e-05,
-      "loss": 8.31605224609375,
+      "grad_norm": 0.21722739934921265,
+      "learning_rate": 9.742690058479532e-05,
+      "loss": 8.316014099121094,
       "step": 45
     },
     {
       "epoch": 0.8830022075055187,
-      "grad_norm": 0.29225534200668335,
-      "learning_rate": 9.828070175438598e-05,
-      "loss": 8.311253356933594,
+      "grad_norm": 0.2812463641166687,
+      "learning_rate": 9.713450292397662e-05,
+      "loss": 8.31099853515625,
       "step": 50
     },
     {
       "epoch": 0.9713024282560706,
-      "grad_norm": 0.2632504105567932,
-      "learning_rate": 9.810526315789475e-05,
-      "loss": 8.309702301025391,
+      "grad_norm": 0.2608415186405182,
+      "learning_rate": 9.68421052631579e-05,
+      "loss": 8.309403991699218,
       "step": 55
     },
     {
       "epoch": 1.0529801324503312,
-      "grad_norm": 0.2510385513305664,
-      "learning_rate": 9.792982456140352e-05,
-      "loss": 8.305320739746094,
+      "grad_norm": 0.25310298800468445,
+      "learning_rate": 9.654970760233919e-05,
+      "loss": 8.305143737792969,
       "step": 60
     },
     {
       "epoch": 1.141280353200883,
-      "grad_norm": 0.2117137461900711,
-      "learning_rate": 9.775438596491229e-05,
-      "loss": 8.303007507324219,
+      "grad_norm": 0.20996594429016113,
+      "learning_rate": 9.625730994152047e-05,
+      "loss": 8.303257751464844,
       "step": 65
     },
     {
       "epoch": 1.229580573951435,
-      "grad_norm": 0.22979161143302917,
-      "learning_rate": 9.757894736842106e-05,
-      "loss": 8.299153137207032,
+      "grad_norm": 0.22537994384765625,
+      "learning_rate": 9.596491228070175e-05,
+      "loss": 8.29981689453125,
       "step": 70
     },
     {
       "epoch": 1.3178807947019868,
-      "grad_norm": 0.17818668484687805,
-      "learning_rate": 9.740350877192983e-05,
-      "loss": 8.295228576660156,
+      "grad_norm": 0.17918814718723297,
+      "learning_rate": 9.567251461988305e-05,
+      "loss": 8.295522308349609,
       "step": 75
     },
     {
       "epoch": 1.4061810154525387,
-      "grad_norm": 0.22605223953723907,
-      "learning_rate": 9.72280701754386e-05,
-      "loss": 8.29014892578125,
+      "grad_norm": 0.207343190908432,
+      "learning_rate": 9.538011695906434e-05,
+      "loss": 8.290687561035156,
       "step": 80
     },
     {
       "epoch": 1.4944812362030906,
-      "grad_norm": 0.34358182549476624,
-      "learning_rate": 9.705263157894738e-05,
-      "loss": 8.281324005126953,
+      "grad_norm": 0.32762715220451355,
+      "learning_rate": 9.508771929824562e-05,
+      "loss": 8.281892395019531,
       "step": 85
     },
     {
       "epoch": 1.5827814569536423,
-      "grad_norm": 0.2882179319858551,
-      "learning_rate": 9.687719298245615e-05,
-      "loss": 8.285108947753907,
+      "grad_norm": 0.2938079535961151,
+      "learning_rate": 9.47953216374269e-05,
+      "loss": 8.285610961914063,
       "step": 90
     },
     {
       "epoch": 1.6710816777041941,
-      "grad_norm": 0.25494980812072754,
-      "learning_rate": 9.670175438596492e-05,
-      "loss": 8.28400115966797,
+      "grad_norm": 0.2513900399208069,
+      "learning_rate": 9.45029239766082e-05,
+      "loss": 8.284735107421875,
       "step": 95
     },
     {
       "epoch": 1.759381898454746,
-      "grad_norm": 0.2691737711429596,
-      "learning_rate": 9.652631578947369e-05,
-      "loss": 8.282723236083985,
-      "step": 100
-    },
-    {
-      "epoch": 1.759381898454746,
-      "eval_loss": 8.279256820678711,
-      "eval_runtime": 6.4238,
-      "eval_samples_per_second": 15.567,
-      "eval_steps_per_second": 7.784,
+      "grad_norm": 0.24608662724494934,
+      "learning_rate": 9.421052631578949e-05,
+      "loss": 8.28292236328125,
       "step": 100
     },
     {
       "epoch": 1.847682119205298,
-      "grad_norm": 0.3390202522277832,
-      "learning_rate": 9.635087719298246e-05,
-      "loss": 8.274485778808593,
+      "grad_norm": 0.3305456340312958,
+      "learning_rate": 9.391812865497076e-05,
+      "loss": 8.274761962890626,
       "step": 105
     },
     {
       "epoch": 1.9359823399558498,
-      "grad_norm": 0.32316988706588745,
-      "learning_rate": 9.617543859649123e-05,
-      "loss": 8.268724060058593,
+      "grad_norm": 0.2837885022163391,
+      "learning_rate": 9.362573099415205e-05,
+      "loss": 8.268502807617187,
       "step": 110
     },
     {
       "epoch": 2.0176600441501105,
-      "grad_norm": 0.38053184747695923,
-      "learning_rate": 9.6e-05,
-      "loss": 8.268507385253907,
+      "grad_norm": 0.37587037682533264,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 8.26836929321289,
       "step": 115
     },
     {
       "epoch": 2.1059602649006623,
-      "grad_norm": 0.42441317439079285,
-      "learning_rate": 9.582456140350877e-05,
-      "loss": 8.240388488769531,
+      "grad_norm": 0.42066144943237305,
+      "learning_rate": 9.304093567251462e-05,
+      "loss": 8.240390014648437,
       "step": 120
     },
     {
       "epoch": 2.1942604856512142,
-      "grad_norm": 0.39643266797065735,
-      "learning_rate": 9.564912280701754e-05,
-      "loss": 8.257127380371093,
+      "grad_norm": 0.4040590822696686,
+      "learning_rate": 9.27485380116959e-05,
+      "loss": 8.256266021728516,
       "step": 125
     },
     {
       "epoch": 2.282560706401766,
-      "grad_norm": 0.37942928075790405,
-      "learning_rate": 9.547368421052632e-05,
-      "loss": 8.236712646484374,
+      "grad_norm": 0.3695836365222931,
+      "learning_rate": 9.24561403508772e-05,
+      "loss": 8.237049102783203,
       "step": 130
     },
     {
       "epoch": 2.370860927152318,
-      "grad_norm": 0.4616714417934418,
-      "learning_rate": 9.52982456140351e-05,
-      "loss": 8.242158508300781,
+      "grad_norm": 0.4006362557411194,
+      "learning_rate": 9.216374269005849e-05,
+      "loss": 8.241561889648438,
       "step": 135
     },
     {
       "epoch": 2.45916114790287,
-      "grad_norm": 0.4159388840198517,
-      "learning_rate": 9.512280701754387e-05,
-      "loss": 8.228749084472657,
+      "grad_norm": 0.38214218616485596,
+      "learning_rate": 9.187134502923977e-05,
+      "loss": 8.228585815429687,
       "step": 140
     },
     {
       "epoch": 2.5474613686534218,
-      "grad_norm": 0.4015291929244995,
-      "learning_rate": 9.494736842105264e-05,
-      "loss": 8.214886474609376,
+      "grad_norm": 0.40786632895469666,
+      "learning_rate": 9.157894736842105e-05,
+      "loss": 8.215204620361328,
       "step": 145
     },
     {
       "epoch": 2.6357615894039736,
-      "grad_norm": 0.37219980359077454,
-      "learning_rate": 9.477192982456141e-05,
-      "loss": 8.210240173339844,
+      "grad_norm": 0.349801242351532,
+      "learning_rate": 9.128654970760235e-05,
+      "loss": 8.210234069824219,
+      "step": 150
+    },
+    {
+      "epoch": 2.6357615894039736,
+      "eval_loss": 8.208977699279785,
+      "eval_runtime": 6.4293,
+      "eval_samples_per_second": 15.554,
+      "eval_steps_per_second": 7.777,
       "step": 150
     },
     {
       "epoch": 2.7240618101545255,
-      "grad_norm": 0.3414648175239563,
-      "learning_rate": 9.459649122807018e-05,
-      "loss": 8.203079223632812,
+      "grad_norm": 0.37537169456481934,
+      "learning_rate": 9.099415204678363e-05,
+      "loss": 8.202523803710937,
       "step": 155
     },
     {
       "epoch": 2.8123620309050774,
-      "grad_norm": 0.3885168135166168,
-      "learning_rate": 9.442105263157895e-05,
-      "loss": 8.185888671875,
+      "grad_norm": 0.3929419219493866,
+      "learning_rate": 9.070175438596492e-05,
+      "loss": 8.18493423461914,
       "step": 160
     },
     {
       "epoch": 2.9006622516556293,
-      "grad_norm": 0.355350524187088,
-      "learning_rate": 9.424561403508773e-05,
-      "loss": 8.181796264648437,
+      "grad_norm": 0.3615664541721344,
+      "learning_rate": 9.04093567251462e-05,
+      "loss": 8.17913818359375,
       "step": 165
     },
     {
       "epoch": 2.988962472406181,
-      "grad_norm": 0.3024243712425232,
-      "learning_rate": 9.40701754385965e-05,
-      "loss": 8.189637756347656,
+      "grad_norm": 0.28156760334968567,
+      "learning_rate": 9.011695906432748e-05,
+      "loss": 8.187295532226562,
       "step": 170
     },
     {
       "epoch": 3.0706401766004414,
-      "grad_norm": 0.43162932991981506,
-      "learning_rate": 9.389473684210527e-05,
-      "loss": 8.143230438232422,
+      "grad_norm": 0.4112038016319275,
+      "learning_rate": 8.982456140350878e-05,
+      "loss": 8.141925048828124,
       "step": 175
     },
     {
       "epoch": 3.1589403973509933,
-      "grad_norm": 0.47192949056625366,
-      "learning_rate": 9.371929824561404e-05,
-      "loss": 8.154377746582032,
+      "grad_norm": 0.46913009881973267,
+      "learning_rate": 8.953216374269007e-05,
+      "loss": 8.151603698730469,
       "step": 180
     },
     {
       "epoch": 3.247240618101545,
-      "grad_norm": 0.7041413187980652,
-      "learning_rate": 9.354385964912281e-05,
-      "loss": 8.170367431640624,
+      "grad_norm": 0.7114327549934387,
+      "learning_rate": 8.923976608187135e-05,
+      "loss": 8.168936157226563,
       "step": 185
     },
     {
       "epoch": 3.335540838852097,
-      "grad_norm": 0.7336998581886292,
-      "learning_rate": 9.336842105263158e-05,
-      "loss": 8.124923706054688,
+      "grad_norm": 0.6770473122596741,
+      "learning_rate": 8.894736842105263e-05,
+      "loss": 8.124656677246094,
       "step": 190
     },
     {
       "epoch": 3.423841059602649,
-      "grad_norm": 0.5626300573348999,
-      "learning_rate": 9.319298245614035e-05,
-      "loss": 8.115943908691406,
+      "grad_norm": 0.4436897933483124,
+      "learning_rate": 8.865497076023393e-05,
+      "loss": 8.115914154052735,
       "step": 195
     },
     {
       "epoch": 3.512141280353201,
-      "grad_norm": 0.4975186586380005,
-      "learning_rate": 9.301754385964912e-05,
-      "loss": 8.125883483886719,
-      "step": 200
-    },
-    {
-      "epoch": 3.512141280353201,
-      "eval_loss": 8.111800193786621,
-      "eval_runtime": 6.4318,
-      "eval_samples_per_second": 15.548,
-      "eval_steps_per_second": 7.774,
+      "grad_norm": 0.5368486642837524,
+      "learning_rate": 8.836257309941522e-05,
+      "loss": 8.126858520507813,
       "step": 200
     },
     {
       "epoch": 3.6004415011037527,
-      "grad_norm": 0.410458505153656,
-      "learning_rate": 9.28421052631579e-05,
-      "loss": 8.097926330566406,
+      "grad_norm": 0.4201599955558777,
+      "learning_rate": 8.807017543859649e-05,
+      "loss": 8.098291015625,
       "step": 205
     },
     {
       "epoch": 3.6887417218543046,
-      "grad_norm": 0.6508727669715881,
-      "learning_rate": 9.266666666666666e-05,
-      "loss": 8.09316177368164,
+      "grad_norm": 0.478750616312027,
+      "learning_rate": 8.777777777777778e-05,
+      "loss": 8.094570922851563,
       "step": 210
     },
     {
       "epoch": 3.7770419426048565,
-      "grad_norm": 0.4533739984035492,
-      "learning_rate": 9.249122807017544e-05,
-      "loss": 8.101840209960937,
+      "grad_norm": 0.3878437280654907,
+      "learning_rate": 8.748538011695907e-05,
+      "loss": 8.102629852294921,
       "step": 215
     },
     {
       "epoch": 3.8653421633554084,
-      "grad_norm": 0.5813506841659546,
-      "learning_rate": 9.231578947368421e-05,
-      "loss": 8.057862091064454,
+      "grad_norm": 0.64486163854599,
+      "learning_rate": 8.719298245614036e-05,
+      "loss": 8.064454650878906,
       "step": 220
     },
     {
       "epoch": 3.9536423841059603,
-      "grad_norm": 0.6099780797958374,
-      "learning_rate": 9.214035087719298e-05,
-      "loss": 8.063623046875,
+      "grad_norm": 0.6539655327796936,
+      "learning_rate": 8.690058479532164e-05,
+      "loss": 8.068508911132813,
       "step": 225
     },
     {
       "epoch": 4.035320088300221,
-      "grad_norm": 0.5916600823402405,
-      "learning_rate": 9.196491228070175e-05,
-      "loss": 8.066812133789062,
+      "grad_norm": 0.5649951100349426,
+      "learning_rate": 8.660818713450293e-05,
+      "loss": 8.070999908447266,
       "step": 230
     },
     {
       "epoch": 4.123620309050772,
-      "grad_norm": 0.561387836933136,
-      "learning_rate": 9.178947368421052e-05,
-      "loss": 8.047982788085937,
+      "grad_norm": 0.506738543510437,
+      "learning_rate": 8.631578947368421e-05,
+      "loss": 8.052703857421875,
       "step": 235
     },
     {
       "epoch": 4.211920529801325,
-      "grad_norm": 0.6326033473014832,
-      "learning_rate": 9.16140350877193e-05,
-      "loss": 8.045574951171876,
+      "grad_norm": 0.6467848420143127,
+      "learning_rate": 8.60233918128655e-05,
+      "loss": 8.04999237060547,
       "step": 240
     },
     {
       "epoch": 4.300220750551876,
-      "grad_norm": 0.6416261196136475,
-      "learning_rate": 9.143859649122808e-05,
-      "loss": 8.064215087890625,
+      "grad_norm": 0.6093457937240601,
+      "learning_rate": 8.573099415204678e-05,
+      "loss": 8.068696594238281,
       "step": 245
     },
     {
       "epoch": 4.3885209713024285,
-      "grad_norm": 0.6526978015899658,
-      "learning_rate": 9.126315789473685e-05,
-      "loss": 8.000985717773437,
+      "grad_norm": 0.6376983523368835,
+      "learning_rate": 8.543859649122808e-05,
+      "loss": 8.007734680175782,
       "step": 250
     },
     {
       "epoch": 4.47682119205298,
-      "grad_norm": 0.5645702481269836,
-      "learning_rate": 9.108771929824562e-05,
-      "loss": 8.016084289550781,
+      "grad_norm": 0.549198567867279,
+      "learning_rate": 8.514619883040936e-05,
+      "loss": 8.022071838378906,
       "step": 255
     },
     {
       "epoch": 4.565121412803532,
-      "grad_norm": 0.8505818247795105,
-      "learning_rate": 9.091228070175439e-05,
-      "loss": 8.033029937744141,
+      "grad_norm": 0.6188969612121582,
+      "learning_rate": 8.485380116959065e-05,
+      "loss": 8.03642578125,
       "step": 260
     },
     {
       "epoch": 4.653421633554084,
-      "grad_norm": 0.6926712989807129,
-      "learning_rate": 9.073684210526316e-05,
-      "loss": 7.961753845214844,
+      "grad_norm": 0.8166013360023499,
+      "learning_rate": 8.456140350877193e-05,
+      "loss": 7.964942169189453,
       "step": 265
     },
     {
       "epoch": 4.741721854304636,
-      "grad_norm": 0.9868171215057373,
-      "learning_rate": 9.056140350877193e-05,
-      "loss": 8.012611389160156,
+      "grad_norm": 0.6450521349906921,
+      "learning_rate": 8.426900584795321e-05,
+      "loss": 8.015605163574218,
       "step": 270
     },
     {
       "epoch": 4.830022075055187,
-      "grad_norm": 0.6124998927116394,
-      "learning_rate": 9.03859649122807e-05,
-      "loss": 8.01495361328125,
+      "grad_norm": 0.5845851302146912,
+      "learning_rate": 8.397660818713451e-05,
+      "loss": 8.016954803466797,
       "step": 275
     },
     {
       "epoch": 4.91832229580574,
-      "grad_norm": 0.7939390540122986,
-      "learning_rate": 9.021052631578947e-05,
-      "loss": 7.967988586425781,
+      "grad_norm": 0.6162020564079285,
+      "learning_rate": 8.36842105263158e-05,
+      "loss": 7.971800231933594,
       "step": 280
     },
     {
       "epoch": 5.0,
-      "grad_norm": 0.6462355852127075,
-      "learning_rate": 9.003508771929824e-05,
-      "loss": 7.989793395996093,
+      "grad_norm": 0.6822549104690552,
+      "learning_rate": 8.339181286549708e-05,
+      "loss": 7.994195556640625,
       "step": 285
     },
     {
       "epoch": 5.088300220750552,
-      "grad_norm": 0.7130389213562012,
-      "learning_rate": 8.985964912280703e-05,
-      "loss": 7.976549530029297,
+      "grad_norm": 0.8076459169387817,
+      "learning_rate": 8.309941520467836e-05,
+      "loss": 7.982322692871094,
       "step": 290
     },
     {
       "epoch": 5.176600441501104,
-      "grad_norm": 0.7381460666656494,
-      "learning_rate": 8.96842105263158e-05,
-      "loss": 7.947661590576172,
+      "grad_norm": 0.7110167741775513,
+      "learning_rate": 8.280701754385966e-05,
+      "loss": 7.952334594726563,
       "step": 295
     },
     {
       "epoch": 5.264900662251655,
-      "grad_norm": 0.6813192963600159,
-      "learning_rate": 8.950877192982457e-05,
-      "loss": 7.9572906494140625,
+      "grad_norm": 0.7477928996086121,
+      "learning_rate": 8.251461988304095e-05,
+      "loss": 7.962551879882812,
       "step": 300
     },
     {
       "epoch": 5.264900662251655,
-      "eval_loss": 7.974903583526611,
-      "eval_runtime": 6.4143,
-      "eval_samples_per_second": 15.59,
-      "eval_steps_per_second": 7.795,
+      "eval_loss": 7.9789137840271,
+      "eval_runtime": 6.4308,
+      "eval_samples_per_second": 15.55,
+      "eval_steps_per_second": 7.775,
       "step": 300
     },
     {
       "epoch": 5.3532008830022075,
-      "grad_norm": 0.7090719938278198,
-      "learning_rate": 8.933333333333334e-05,
-      "loss": 7.977705383300782,
+      "grad_norm": 0.6691218614578247,
+      "learning_rate": 8.222222222222222e-05,
+      "loss": 7.981855773925782,
       "step": 305
     },
     {
       "epoch": 5.441501103752759,
-      "grad_norm": 0.6882940530776978,
-      "learning_rate": 8.915789473684211e-05,
-      "loss": 7.9357421875,
+      "grad_norm": 0.590973973274231,
+      "learning_rate": 8.192982456140351e-05,
+      "loss": 7.941262817382812,
       "step": 310
     },
     {
       "epoch": 5.529801324503311,
-      "grad_norm": 0.535625696182251,
-      "learning_rate": 8.898245614035088e-05,
-      "loss": 7.946101379394531,
+      "grad_norm": 0.6093570590019226,
+      "learning_rate": 8.16374269005848e-05,
+      "loss": 7.952169036865234,
       "step": 315
     },
     {
       "epoch": 5.618101545253863,
-      "grad_norm": 0.5747976303100586,
-      "learning_rate": 8.880701754385965e-05,
-      "loss": 7.960299682617188,
+      "grad_norm": 0.44569653272628784,
+      "learning_rate": 8.134502923976609e-05,
+      "loss": 7.9652587890625,
       "step": 320
     },
     {
       "epoch": 5.706401766004415,
-      "grad_norm": 0.8299369215965271,
-      "learning_rate": 8.863157894736842e-05,
-      "loss": 7.949764251708984,
+      "grad_norm": 0.7363209128379822,
+      "learning_rate": 8.105263157894737e-05,
+      "loss": 7.954380798339844,
       "step": 325
     },
     {
       "epoch": 5.7947019867549665,
-      "grad_norm": 0.8384077548980713,
-      "learning_rate": 8.84561403508772e-05,
-      "loss": 7.9373046875,
+      "grad_norm": 0.8245615363121033,
+      "learning_rate": 8.076023391812866e-05,
+      "loss": 7.940582275390625,
       "step": 330
     },
     {
       "epoch": 5.883002207505519,
-      "grad_norm": 0.9269998669624329,
-      "learning_rate": 8.828070175438597e-05,
-      "loss": 7.914773559570312,
+      "grad_norm": 0.9126182198524475,
+      "learning_rate": 8.046783625730994e-05,
+      "loss": 7.918997192382813,
       "step": 335
     },
     {
       "epoch": 5.97130242825607,
-      "grad_norm": 0.7561098337173462,
-      "learning_rate": 8.810526315789475e-05,
-      "loss": 7.882290649414062,
+      "grad_norm": 0.7115766406059265,
+      "learning_rate": 8.017543859649124e-05,
+      "loss": 7.887815856933594,
       "step": 340
     },
     {
       "epoch": 6.052980132450331,
-      "grad_norm": 0.6034086346626282,
-      "learning_rate": 8.792982456140352e-05,
-      "loss": 7.8886253356933596,
+      "grad_norm": 0.5615514516830444,
+      "learning_rate": 7.988304093567252e-05,
+      "loss": 7.894483184814453,
       "step": 345
     },
     {
       "epoch": 6.141280353200883,
-      "grad_norm": 0.8741679787635803,
-      "learning_rate": 8.77543859649123e-05,
-      "loss": 7.885476684570312,
+      "grad_norm": 0.764122486114502,
+      "learning_rate": 7.959064327485381e-05,
+      "loss": 7.8913932800292965,
       "step": 350
     },
     {
       "epoch": 6.229580573951435,
-      "grad_norm": 1.0234349966049194,
-      "learning_rate": 8.757894736842106e-05,
-      "loss": 7.884663391113281,
+      "grad_norm": 0.9715237617492676,
+      "learning_rate": 7.929824561403509e-05,
+      "loss": 7.890602111816406,
       "step": 355
     },
     {
       "epoch": 6.317880794701987,
-      "grad_norm": 0.8799805045127869,
-      "learning_rate": 8.740350877192984e-05,
-      "loss": 7.982850646972656,
+      "grad_norm": 0.8448056578636169,
+      "learning_rate": 7.900584795321638e-05,
+      "loss": 7.986685180664063,
       "step": 360
     },
     {
       "epoch": 6.406181015452539,
-      "grad_norm": 0.730731725692749,
-      "learning_rate": 8.72280701754386e-05,
-      "loss": 7.9165794372558596,
+      "grad_norm": 0.9147307872772217,
+      "learning_rate": 7.871345029239767e-05,
+      "loss": 7.92308349609375,
       "step": 365
     },
     {
       "epoch": 6.49448123620309,
-      "grad_norm": 1.0876460075378418,
-      "learning_rate": 8.705263157894738e-05,
-      "loss": 7.914888000488281,
+      "grad_norm": 1.0160162448883057,
+      "learning_rate": 7.842105263157895e-05,
+      "loss": 7.9209846496582035,
       "step": 370
     },
     {
       "epoch": 6.582781456953643,
-      "grad_norm": 0.6480677127838135,
-      "learning_rate": 8.687719298245615e-05,
-      "loss": 7.892662048339844,
+      "grad_norm": 0.6162037253379822,
+      "learning_rate": 7.812865497076024e-05,
+      "loss": 7.8999687194824215,
       "step": 375
     },
     {
       "epoch": 6.671081677704194,
-      "grad_norm": 0.7012283205986023,
-      "learning_rate": 8.670175438596492e-05,
-      "loss": 7.865422058105469,
+      "grad_norm": 0.7081504464149475,
+      "learning_rate": 7.783625730994153e-05,
+      "loss": 7.872428894042969,
       "step": 380
     },
     {
       "epoch": 6.7593818984547465,
-      "grad_norm": 0.6792997121810913,
-      "learning_rate": 8.652631578947369e-05,
-      "loss": 7.868894958496094,
+      "grad_norm": 0.6675496697425842,
+      "learning_rate": 7.754385964912281e-05,
+      "loss": 7.877265930175781,
       "step": 385
     },
     {
       "epoch": 6.847682119205298,
-      "grad_norm": 0.8157274723052979,
-      "learning_rate": 8.635087719298246e-05,
-      "loss": 7.902152252197266,
+      "grad_norm": 0.8263344764709473,
+      "learning_rate": 7.72514619883041e-05,
+      "loss": 7.908570861816406,
       "step": 390
     },
     {
       "epoch": 6.93598233995585,
-      "grad_norm": 0.8674540519714355,
-      "learning_rate": 8.617543859649123e-05,
-      "loss": 7.812020874023437,
+      "grad_norm": 0.7107247710227966,
+      "learning_rate": 7.695906432748539e-05,
+      "loss": 7.816087341308593,
       "step": 395
     },
     {
       "epoch": 7.0176600441501105,
-      "grad_norm": 0.8302394151687622,
-      "learning_rate": 8.6e-05,
-      "loss": 7.89735107421875,
-      "step": 400
-    },
-    {
-      "epoch": 7.0176600441501105,
-      "eval_loss": 7.898021697998047,
-      "eval_runtime": 6.4239,
-      "eval_samples_per_second": 15.567,
-      "eval_steps_per_second": 7.783,
+      "grad_norm": 0.8045026063919067,
+      "learning_rate": 7.666666666666667e-05,
+      "loss": 7.903431701660156,
       "step": 400
     },
     {
       "epoch": 7.105960264900662,
-      "grad_norm": 1.1196335554122925,
-      "learning_rate": 8.582456140350877e-05,
-      "loss": 7.833503723144531,
+      "grad_norm": 0.860241174697876,
+      "learning_rate": 7.637426900584795e-05,
+      "loss": 7.839199829101562,
       "step": 405
     },
     {
       "epoch": 7.194260485651214,
-      "grad_norm": 0.7547705769538879,
-      "learning_rate": 8.564912280701755e-05,
-      "loss": 7.870079040527344,
+      "grad_norm": 0.9684602618217468,
+      "learning_rate": 7.608187134502924e-05,
+      "loss": 7.877064514160156,
       "step": 410
     },
     {
       "epoch": 7.282560706401766,
-      "grad_norm": 0.7977498769760132,
-      "learning_rate": 8.547368421052632e-05,
-      "loss": 7.873524475097656,
+      "grad_norm": 0.904358983039856,
+      "learning_rate": 7.578947368421054e-05,
+      "loss": 7.8798271179199215,
       "step": 415
     },
     {
       "epoch": 7.370860927152318,
-      "grad_norm": 0.7627144455909729,
-      "learning_rate": 8.529824561403509e-05,
-      "loss": 7.815260314941407,
+      "grad_norm": 0.8752867579460144,
+      "learning_rate": 7.549707602339182e-05,
+      "loss": 7.820896911621094,
       "step": 420
     },
     {
       "epoch": 7.459161147902869,
-      "grad_norm": 0.7699235677719116,
-      "learning_rate": 8.512280701754386e-05,
-      "loss": 7.884523010253906,
+      "grad_norm": 0.8255132436752319,
+      "learning_rate": 7.52046783625731e-05,
+      "loss": 7.889141845703125,
       "step": 425
     },
     {
       "epoch": 7.547461368653422,
-      "grad_norm": 0.9240535497665405,
-      "learning_rate": 8.494736842105263e-05,
-      "loss": 7.79596176147461,
+      "grad_norm": 0.7994998097419739,
+      "learning_rate": 7.491228070175439e-05,
+      "loss": 7.799693298339844,
       "step": 430
     },
     {
       "epoch": 7.635761589403973,
-      "grad_norm": 1.2076364755630493,
-      "learning_rate": 8.47719298245614e-05,
-      "loss": 7.870022583007812,
+      "grad_norm": 1.000003695487976,
+      "learning_rate": 7.461988304093567e-05,
+      "loss": 7.874913024902344,
       "step": 435
     },
     {
       "epoch": 7.7240618101545255,
-      "grad_norm": 0.7228724956512451,
-      "learning_rate": 8.459649122807018e-05,
-      "loss": 7.8937835693359375,
+      "grad_norm": 0.6996250152587891,
+      "learning_rate": 7.432748538011697e-05,
+      "loss": 7.898649597167969,
       "step": 440
     },
     {
       "epoch": 7.812362030905077,
-      "grad_norm": 0.6597837209701538,
-      "learning_rate": 8.442105263157896e-05,
-      "loss": 7.868130493164062,
+      "grad_norm": 0.7075074911117554,
+      "learning_rate": 7.403508771929825e-05,
+      "loss": 7.874346923828125,
       "step": 445
     },
     {
       "epoch": 7.900662251655629,
-      "grad_norm": 0.7094933390617371,
-      "learning_rate": 8.424561403508773e-05,
-      "loss": 7.8882080078125,
+      "grad_norm": 0.855798065662384,
+      "learning_rate": 7.374269005847954e-05,
+      "loss": 7.894575500488282,
+      "step": 450
+    },
+    {
+      "epoch": 7.900662251655629,
+      "eval_loss": 7.881262302398682,
+      "eval_runtime": 6.4538,
+      "eval_samples_per_second": 15.495,
+      "eval_steps_per_second": 7.747,
       "step": 450
     },
     {
       "epoch": 7.988962472406181,
-      "grad_norm": 1.1244697570800781,
-      "learning_rate": 8.40701754385965e-05,
-      "loss": 7.857438659667968,
+      "grad_norm": 1.0835974216461182,
+      "learning_rate": 7.345029239766082e-05,
+      "loss": 7.863267517089843,
       "step": 455
     },
     {
       "epoch": 8.070640176600442,
-      "grad_norm": 0.8842155337333679,
-      "learning_rate": 8.389473684210527e-05,
-      "loss": 7.775878143310547,
+      "grad_norm": 0.7638969421386719,
+      "learning_rate": 7.315789473684212e-05,
+      "loss": 7.783184814453125,
       "step": 460
     },
     {
       "epoch": 8.158940397350994,
-      "grad_norm": 1.178807020187378,
-      "learning_rate": 8.371929824561404e-05,
-      "loss": 7.8323616027832035,
+      "grad_norm": 0.9937992095947266,
+      "learning_rate": 7.28654970760234e-05,
+      "loss": 7.839315795898438,
       "step": 465
     },
     {
       "epoch": 8.247240618101545,
-      "grad_norm": 0.8146442770957947,
-      "learning_rate": 8.354385964912281e-05,
-      "loss": 7.838046264648438,
+      "grad_norm": 0.8237105011940002,
+      "learning_rate": 7.257309941520468e-05,
+      "loss": 7.841758728027344,
       "step": 470
     },
     {
       "epoch": 8.335540838852097,
-      "grad_norm": 0.8231097459793091,
-      "learning_rate": 8.336842105263158e-05,
-      "loss": 7.8477630615234375,
+      "grad_norm": 0.7182467579841614,
+      "learning_rate": 7.228070175438597e-05,
+      "loss": 7.85023193359375,
       "step": 475
     },
     {
       "epoch": 8.42384105960265,
-      "grad_norm": 0.8259224891662598,
-      "learning_rate": 8.319298245614035e-05,
-      "loss": 7.861925506591797,
+      "grad_norm": 0.8129780888557434,
+      "learning_rate": 7.198830409356726e-05,
+      "loss": 7.865559387207031,
       "step": 480
     },
     {
       "epoch": 8.512141280353202,
-      "grad_norm": 0.6924996972084045,
-      "learning_rate": 8.301754385964912e-05,
-      "loss": 7.827101135253907,
+      "grad_norm": 0.6443418264389038,
+      "learning_rate": 7.169590643274853e-05,
+      "loss": 7.831302642822266,
       "step": 485
     },
     {
       "epoch": 8.600441501103752,
-      "grad_norm": 0.7627933025360107,
-      "learning_rate": 8.28421052631579e-05,
-      "loss": 7.876766967773437,
+      "grad_norm": 0.7900895476341248,
+      "learning_rate": 7.140350877192983e-05,
+      "loss": 7.88115234375,
       "step": 490
     },
     {
       "epoch": 8.688741721854305,
-      "grad_norm": 0.9244223237037659,
-      "learning_rate": 8.266666666666667e-05,
-      "loss": 7.841154479980469,
+      "grad_norm": 0.6896668076515198,
+      "learning_rate": 7.111111111111112e-05,
+      "loss": 7.8469383239746096,
       "step": 495
     },
     {
       "epoch": 8.777041942604857,
-      "grad_norm": 0.9384596347808838,
-      "learning_rate": 8.249122807017544e-05,
-      "loss": 7.777418518066407,
-      "step": 500
-    },
-    {
-      "epoch": 8.777041942604857,
-      "eval_loss": 7.860381603240967,
-      "eval_runtime": 6.4376,
-      "eval_samples_per_second": 15.534,
-      "eval_steps_per_second": 7.767,
+      "grad_norm": 0.8964319229125977,
+      "learning_rate": 7.08187134502924e-05,
+      "loss": 7.783286285400391,
       "step": 500
     },
     {
       "epoch": 8.865342163355407,
-      "grad_norm": 1.2778297662734985,
-      "learning_rate": 8.231578947368421e-05,
-      "loss": 7.801692199707031,
+      "grad_norm": 1.1212209463119507,
+      "learning_rate": 7.052631578947368e-05,
+      "loss": 7.807972717285156,
       "step": 505
     },
     {
       "epoch": 8.95364238410596,
-      "grad_norm": 0.9670100808143616,
-      "learning_rate": 8.214035087719298e-05,
-      "loss": 7.84215087890625,
+      "grad_norm": 1.048243522644043,
+      "learning_rate": 7.023391812865497e-05,
+      "loss": 7.847767639160156,
       "step": 510
     },
     {
       "epoch": 9.035320088300221,
-      "grad_norm": 1.118955135345459,
-      "learning_rate": 8.196491228070175e-05,
-      "loss": 7.8455451965332035,
+      "grad_norm": 1.1085704565048218,
+      "learning_rate": 6.994152046783627e-05,
+      "loss": 7.852699279785156,
       "step": 515
     },
     {
       "epoch": 9.123620309050773,
-      "grad_norm": 0.9509307146072388,
-      "learning_rate": 8.178947368421052e-05,
-      "loss": 7.744534301757812,
+      "grad_norm": 0.9384048581123352,
+      "learning_rate": 6.964912280701754e-05,
+      "loss": 7.753205871582031,
       "step": 520
     },
     {
       "epoch": 9.211920529801324,
-      "grad_norm": 0.9677459001541138,
-      "learning_rate": 8.161403508771929e-05,
-      "loss": 7.826947021484375,
+      "grad_norm": 0.9842119216918945,
+      "learning_rate": 6.935672514619883e-05,
+      "loss": 7.836062622070313,
       "step": 525
     },
     {
       "epoch": 9.300220750551876,
-      "grad_norm": 1.3572008609771729,
-      "learning_rate": 8.143859649122808e-05,
-      "loss": 7.786278533935547,
+      "grad_norm": 1.29048490524292,
+      "learning_rate": 6.906432748538012e-05,
+      "loss": 7.795281982421875,
       "step": 530
     },
     {
       "epoch": 9.388520971302428,
-      "grad_norm": 1.2385722398757935,
-      "learning_rate": 8.126315789473685e-05,
-      "loss": 7.7951904296875,
+      "grad_norm": 0.804202139377594,
+      "learning_rate": 6.87719298245614e-05,
+      "loss": 7.802140808105468,
       "step": 535
     },
     {
       "epoch": 9.47682119205298,
-      "grad_norm": 0.793428361415863,
-      "learning_rate": 8.108771929824562e-05,
-      "loss": 7.8350341796875,
+      "grad_norm": 0.8415213823318481,
+      "learning_rate": 6.847953216374268e-05,
+      "loss": 7.84085693359375,
       "step": 540
     },
     {
       "epoch": 9.565121412803531,
-      "grad_norm": 0.8982236981391907,
-      "learning_rate": 8.091228070175439e-05,
-      "loss": 7.8044586181640625,
+      "grad_norm": 0.5611467957496643,
+      "learning_rate": 6.818713450292398e-05,
+      "loss": 7.810952758789062,
       "step": 545
     },
     {
       "epoch": 9.653421633554084,
-      "grad_norm": 0.8177990317344666,
-      "learning_rate": 8.073684210526316e-05,
-      "loss": 7.854301452636719,
+      "grad_norm": 0.730029284954071,
+      "learning_rate": 6.789473684210527e-05,
+      "loss": 7.858808898925782,
       "step": 550
     },
     {
       "epoch": 9.741721854304636,
-      "grad_norm": 0.824090301990509,
-      "learning_rate": 8.056140350877193e-05,
-      "loss": 7.8012336730957035,
+      "grad_norm": 0.7788743376731873,
+      "learning_rate": 6.760233918128655e-05,
+      "loss": 7.807586669921875,
       "step": 555
     },
     {
       "epoch": 9.830022075055188,
-      "grad_norm": 1.3339343070983887,
-      "learning_rate": 8.038596491228072e-05,
-      "loss": 7.822520446777344,
+      "grad_norm": 1.4132566452026367,
+      "learning_rate": 6.730994152046783e-05,
+      "loss": 7.82841796875,
       "step": 560
     },
     {
       "epoch": 9.918322295805739,
-      "grad_norm": 1.0260347127914429,
-      "learning_rate": 8.021052631578949e-05,
-      "loss": 7.828153991699219,
+      "grad_norm": 1.0049456357955933,
+      "learning_rate": 6.701754385964913e-05,
+      "loss": 7.835298156738281,
       "step": 565
     },
     {
       "epoch": 10.0,
-      "grad_norm": 0.8278086185455322,
-      "learning_rate": 8.003508771929826e-05,
-      "loss": 7.791429138183593,
+      "grad_norm": 0.8505070209503174,
+      "learning_rate": 6.672514619883041e-05,
+      "loss": 7.798960876464844,
       "step": 570
     },
     {
       "epoch": 10.088300220750552,
-      "grad_norm": 0.7400580644607544,
-      "learning_rate": 7.985964912280703e-05,
-      "loss": 7.806646728515625,
+      "grad_norm": 0.7744261622428894,
+      "learning_rate": 6.64327485380117e-05,
+      "loss": 7.813462829589843,
       "step": 575
     },
     {
       "epoch": 10.176600441501105,
-      "grad_norm": 1.216025948524475,
-      "learning_rate": 7.96842105263158e-05,
-      "loss": 7.769792938232422,
+      "grad_norm": 0.9087972640991211,
+      "learning_rate": 6.614035087719298e-05,
+      "loss": 7.7771759033203125,
       "step": 580
     },
     {
       "epoch": 10.264900662251655,
-      "grad_norm": 0.9150897264480591,
-      "learning_rate": 7.950877192982457e-05,
-      "loss": 7.78175048828125,
+      "grad_norm": 0.759032666683197,
+      "learning_rate": 6.584795321637426e-05,
+      "loss": 7.78880615234375,
       "step": 585
     },
     {
       "epoch": 10.353200883002208,
-      "grad_norm": 0.7429578900337219,
-      "learning_rate": 7.933333333333334e-05,
-      "loss": 7.780392456054687,
+      "grad_norm": 0.7629870772361755,
+      "learning_rate": 6.555555555555556e-05,
+      "loss": 7.788232421875,
       "step": 590
     },
     {
       "epoch": 10.44150110375276,
-      "grad_norm": 0.8183691501617432,
-      "learning_rate": 7.915789473684211e-05,
-      "loss": 7.78046875,
+      "grad_norm": 0.8606480956077576,
+      "learning_rate": 6.526315789473685e-05,
+      "loss": 7.791365051269532,
       "step": 595
     },
     {
       "epoch": 10.52980132450331,
-      "grad_norm": 0.9096909165382385,
-      "learning_rate": 7.898245614035088e-05,
-      "loss": 7.847821044921875,
+      "grad_norm": 0.9902874827384949,
+      "learning_rate": 6.497076023391813e-05,
+      "loss": 7.858397674560547,
       "step": 600
     },
     {
       "epoch": 10.52980132450331,
-      "eval_loss": 7.835643291473389,
-      "eval_runtime": 6.4283,
-      "eval_samples_per_second": 15.556,
-      "eval_steps_per_second": 7.778,
+      "eval_loss": 7.83868408203125,
+      "eval_runtime": 6.4428,
+      "eval_samples_per_second": 15.521,
+      "eval_steps_per_second": 7.761,
       "step": 600
     },
     {
       "epoch": 10.618101545253863,
-      "grad_norm": 1.126974105834961,
-      "learning_rate": 7.880701754385965e-05,
-      "loss": 7.7564544677734375,
+      "grad_norm": 1.0570663213729858,
+      "learning_rate": 6.467836257309941e-05,
+      "loss": 7.763392639160156,
       "step": 605
     },
     {
       "epoch": 10.706401766004415,
-      "grad_norm": 0.6179125905036926,
-      "learning_rate": 7.863157894736843e-05,
-      "loss": 7.784625244140625,
+      "grad_norm": 0.6728692054748535,
+      "learning_rate": 6.43859649122807e-05,
+      "loss": 7.792031860351562,
       "step": 610
     },
     {
       "epoch": 10.794701986754967,
-      "grad_norm": 0.7518591284751892,
-      "learning_rate": 7.84561403508772e-05,
-      "loss": 7.776773834228516,
+      "grad_norm": 0.9234580993652344,
+      "learning_rate": 6.4093567251462e-05,
+      "loss": 7.78275146484375,
       "step": 615
     },
     {
       "epoch": 10.883002207505518,
-      "grad_norm": 0.8314757943153381,
-      "learning_rate": 7.828070175438597e-05,
-      "loss": 7.780368041992188,
+      "grad_norm": 0.8953201174736023,
+      "learning_rate": 6.380116959064327e-05,
+      "loss": 7.788031005859375,
       "step": 620
     },
     {
       "epoch": 10.97130242825607,
-      "grad_norm": 1.2431392669677734,
-      "learning_rate": 7.810526315789474e-05,
-      "loss": 7.7441253662109375,
+      "grad_norm": 1.1980770826339722,
+      "learning_rate": 6.350877192982456e-05,
+      "loss": 7.754103088378907,
       "step": 625
     },
     {
       "epoch": 11.052980132450331,
-      "grad_norm": 1.3096091747283936,
-      "learning_rate": 7.792982456140351e-05,
-      "loss": 7.799203491210937,
+      "grad_norm": 1.1482347249984741,
+      "learning_rate": 6.321637426900586e-05,
+      "loss": 7.806501770019532,
       "step": 630
     },
     {
       "epoch": 11.141280353200884,
-      "grad_norm": 0.8238083720207214,
-      "learning_rate": 7.775438596491228e-05,
-      "loss": 7.763633728027344,
+      "grad_norm": 1.0469353199005127,
+      "learning_rate": 6.292397660818714e-05,
+      "loss": 7.7762001037597654,
       "step": 635
     },
     {
       "epoch": 11.229580573951434,
-      "grad_norm": 1.0971180200576782,
-      "learning_rate": 7.757894736842105e-05,
-      "loss": 7.757233428955078,
+      "grad_norm": 0.9315469264984131,
+      "learning_rate": 6.263157894736842e-05,
+      "loss": 7.766180419921875,
       "step": 640
     },
     {
       "epoch": 11.317880794701987,
-      "grad_norm": 0.7882246971130371,
-      "learning_rate": 7.740350877192984e-05,
-      "loss": 7.761444091796875,
+      "grad_norm": 1.0782599449157715,
+      "learning_rate": 6.233918128654971e-05,
+      "loss": 7.772328948974609,
       "step": 645
     },
     {
       "epoch": 11.406181015452539,
-      "grad_norm": 0.839775800704956,
-      "learning_rate": 7.722807017543861e-05,
-      "loss": 7.744635009765625,
+      "grad_norm": 0.8344573378562927,
+      "learning_rate": 6.2046783625731e-05,
+      "loss": 7.755815124511718,
       "step": 650
     },
     {
       "epoch": 11.494481236203091,
-      "grad_norm": 0.8806105256080627,
-      "learning_rate": 7.705263157894738e-05,
-      "loss": 7.816458129882813,
+      "grad_norm": 0.8328990936279297,
+      "learning_rate": 6.175438596491228e-05,
+      "loss": 7.825865173339844,
       "step": 655
     },
     {
       "epoch": 11.582781456953642,
-      "grad_norm": 1.2837175130844116,
-      "learning_rate": 7.687719298245615e-05,
-      "loss": 7.762530517578125,
+      "grad_norm": 1.287616491317749,
+      "learning_rate": 6.146198830409357e-05,
+      "loss": 7.775987243652343,
       "step": 660
     },
     {
       "epoch": 11.671081677704194,
-      "grad_norm": 1.185728669166565,
-      "learning_rate": 7.670175438596492e-05,
-      "loss": 7.823221588134766,
+      "grad_norm": 0.8391528725624084,
+      "learning_rate": 6.116959064327486e-05,
+      "loss": 7.830711364746094,
       "step": 665
     },
     {
       "epoch": 11.759381898454746,
-      "grad_norm": 0.9250527620315552,
-      "learning_rate": 7.652631578947369e-05,
-      "loss": 7.674348449707031,
+      "grad_norm": 0.9515014290809631,
+      "learning_rate": 6.0877192982456147e-05,
+      "loss": 7.6853485107421875,
       "step": 670
     },
     {
       "epoch": 11.847682119205299,
-      "grad_norm": 1.121268630027771,
-      "learning_rate": 7.635087719298246e-05,
-      "loss": 7.706462097167969,
+      "grad_norm": 1.0046449899673462,
+      "learning_rate": 6.0584795321637434e-05,
+      "loss": 7.714703369140625,
       "step": 675
     },
     {
       "epoch": 11.93598233995585,
-      "grad_norm": 1.1073074340820312,
-      "learning_rate": 7.617543859649123e-05,
-      "loss": 7.757592010498047,
+      "grad_norm": 0.866513192653656,
+      "learning_rate": 6.0292397660818714e-05,
+      "loss": 7.766542053222656,
       "step": 680
     },
     {
       "epoch": 12.01766004415011,
-      "grad_norm": 1.1414549350738525,
-      "learning_rate": 7.6e-05,
-      "loss": 7.787910461425781,
+      "grad_norm": 0.997880756855011,
+      "learning_rate": 6e-05,
+      "loss": 7.7974853515625,
       "step": 685
     },
     {
       "epoch": 12.105960264900663,
-      "grad_norm": 0.9097177386283875,
-      "learning_rate": 7.582456140350878e-05,
-      "loss": 7.710475158691406,
+      "grad_norm": 0.8557000756263733,
+      "learning_rate": 5.970760233918129e-05,
+      "loss": 7.726460266113281,
       "step": 690
     },
     {
       "epoch": 12.194260485651213,
-      "grad_norm": 0.7728700041770935,
-      "learning_rate": 7.564912280701755e-05,
-      "loss": 7.790855407714844,
+      "grad_norm": 0.9331446886062622,
+      "learning_rate": 5.9415204678362576e-05,
+      "loss": 7.802294921875,
       "step": 695
     },
     {
       "epoch": 12.282560706401766,
-      "grad_norm": 0.8290122151374817,
-      "learning_rate": 7.547368421052632e-05,
-      "loss": 7.77270278930664,
-      "step": 700
-    },
-    {
-      "epoch": 12.282560706401766,
-      "eval_loss": 7.8054633140563965,
-      "eval_runtime": 6.4301,
-      "eval_samples_per_second": 15.552,
-      "eval_steps_per_second": 7.776,
+      "grad_norm": 0.795706570148468,
+      "learning_rate": 5.9122807017543856e-05,
+      "loss": 7.7851318359375,
       "step": 700
     },
     {
       "epoch": 12.370860927152318,
-      "grad_norm": 0.8372697830200195,
-      "learning_rate": 7.529824561403509e-05,
-      "loss": 7.703733825683594,
+      "grad_norm": 0.8647646307945251,
+      "learning_rate": 5.8830409356725144e-05,
+      "loss": 7.714936828613281,
       "step": 705
     },
     {
       "epoch": 12.45916114790287,
-      "grad_norm": 0.9445962905883789,
-      "learning_rate": 7.512280701754386e-05,
-      "loss": 7.709912109375,
+      "grad_norm": 1.1638087034225464,
+      "learning_rate": 5.853801169590644e-05,
+      "loss": 7.7236328125,
       "step": 710
     },
     {
       "epoch": 12.54746136865342,
-      "grad_norm": 1.180027961730957,
-      "learning_rate": 7.494736842105263e-05,
-      "loss": 7.7696685791015625,
+      "grad_norm": 0.8338577151298523,
+      "learning_rate": 5.8245614035087725e-05,
+      "loss": 7.780918884277344,
       "step": 715
     },
     {
       "epoch": 12.635761589403973,
-      "grad_norm": 0.9895918965339661,
-      "learning_rate": 7.47719298245614e-05,
-      "loss": 7.770545959472656,
+      "grad_norm": 0.8834207653999329,
+      "learning_rate": 5.7953216374269005e-05,
+      "loss": 7.779144287109375,
       "step": 720
     },
     {
       "epoch": 12.724061810154526,
-      "grad_norm": 0.9101853966712952,
-      "learning_rate": 7.459649122807017e-05,
-      "loss": 7.720480346679688,
+      "grad_norm": 0.6919476389884949,
+      "learning_rate": 5.766081871345029e-05,
+      "loss": 7.729824829101562,
       "step": 725
     },
     {
       "epoch": 12.812362030905078,
-      "grad_norm": 0.9100523591041565,
-      "learning_rate": 7.442105263157894e-05,
-      "loss": 7.747254943847656,
+      "grad_norm": 0.942736804485321,
+      "learning_rate": 5.736842105263158e-05,
+      "loss": 7.761282348632813,
       "step": 730
     },
     {
       "epoch": 12.900662251655628,
-      "grad_norm": 0.7405063509941101,
-      "learning_rate": 7.424561403508771e-05,
-      "loss": 7.7278076171875,
+      "grad_norm": 0.8791229128837585,
+      "learning_rate": 5.7076023391812874e-05,
+      "loss": 7.742418670654297,
       "step": 735
     },
     {
       "epoch": 12.98896247240618,
-      "grad_norm": 0.7498451471328735,
-      "learning_rate": 7.407017543859649e-05,
-      "loss": 7.661986541748047,
+      "grad_norm": 0.9587666988372803,
+      "learning_rate": 5.678362573099415e-05,
+      "loss": 7.673837280273437,
       "step": 740
     },
     {
       "epoch": 13.070640176600442,
-      "grad_norm": 0.9334259629249573,
-      "learning_rate": 7.389473684210527e-05,
-      "loss": 7.728338623046875,
+      "grad_norm": 0.9188740849494934,
+      "learning_rate": 5.649122807017544e-05,
+      "loss": 7.7466583251953125,
       "step": 745
     },
     {
       "epoch": 13.158940397350994,
-      "grad_norm": 0.9905465245246887,
-      "learning_rate": 7.371929824561404e-05,
-      "loss": 7.696038818359375,
+      "grad_norm": 0.9566793441772461,
+      "learning_rate": 5.619883040935673e-05,
+      "loss": 7.713945770263672,
+      "step": 750
+    },
+    {
+      "epoch": 13.158940397350994,
+      "eval_loss": 7.808779716491699,
+      "eval_runtime": 6.4337,
+      "eval_samples_per_second": 15.543,
+      "eval_steps_per_second": 7.772,
       "step": 750
     },
     {
       "epoch": 13.247240618101545,
-      "grad_norm": 0.8847911953926086,
-      "learning_rate": 7.354385964912281e-05,
-      "loss": 7.754905700683594,
+      "grad_norm": 0.8719848990440369,
+      "learning_rate": 5.5906432748538016e-05,
+      "loss": 7.7702789306640625,
       "step": 755
     },
     {
       "epoch": 13.335540838852097,
-      "grad_norm": 0.964031457901001,
-      "learning_rate": 7.336842105263158e-05,
-      "loss": 7.729930114746094,
+      "grad_norm": 0.7353575825691223,
+      "learning_rate": 5.56140350877193e-05,
+      "loss": 7.742958831787109,
       "step": 760
     },
     {
       "epoch": 13.42384105960265,
-      "grad_norm": 1.171689748764038,
-      "learning_rate": 7.319298245614035e-05,
-      "loss": 7.657225036621094,
+      "grad_norm": 0.9595194458961487,
+      "learning_rate": 5.5321637426900584e-05,
+      "loss": 7.6720130920410154,
       "step": 765
     },
     {
       "epoch": 13.512141280353202,
-      "grad_norm": 1.213448405265808,
-      "learning_rate": 7.301754385964912e-05,
-      "loss": 7.7045951843261715,
+      "grad_norm": 1.060484766960144,
+      "learning_rate": 5.502923976608188e-05,
+      "loss": 7.718379211425781,
       "step": 770
     },
     {
       "epoch": 13.600441501103752,
-      "grad_norm": 0.9362474083900452,
-      "learning_rate": 7.28421052631579e-05,
-      "loss": 7.736686706542969,
+      "grad_norm": 0.8258070349693298,
+      "learning_rate": 5.4736842105263165e-05,
+      "loss": 7.750814819335938,
       "step": 775
     },
     {
       "epoch": 13.688741721854305,
-      "grad_norm": 1.029884934425354,
-      "learning_rate": 7.266666666666667e-05,
-      "loss": 7.688392639160156,
+      "grad_norm": 0.818719744682312,
+      "learning_rate": 5.4444444444444446e-05,
+      "loss": 7.706378936767578,
       "step": 780
     },
     {
       "epoch": 13.777041942604857,
-      "grad_norm": 0.9884915947914124,
-      "learning_rate": 7.249122807017544e-05,
-      "loss": 7.726716613769531,
+      "grad_norm": 0.8701429963111877,
+      "learning_rate": 5.415204678362573e-05,
+      "loss": 7.7416229248046875,
       "step": 785
     },
     {
       "epoch": 13.865342163355407,
-      "grad_norm": 1.251460313796997,
-      "learning_rate": 7.231578947368421e-05,
-      "loss": 7.6784309387207035,
+      "grad_norm": 1.0176366567611694,
+      "learning_rate": 5.385964912280702e-05,
+      "loss": 7.693745422363281,
       "step": 790
     },
     {
       "epoch": 13.95364238410596,
-      "grad_norm": 1.0668705701828003,
-      "learning_rate": 7.214035087719298e-05,
-      "loss": 7.6979225158691404,
+      "grad_norm": 0.9578682780265808,
+      "learning_rate": 5.356725146198831e-05,
+      "loss": 7.712064361572265,
       "step": 795
     },
     {
       "epoch": 14.035320088300221,
-      "grad_norm": 1.088294506072998,
-      "learning_rate": 7.196491228070175e-05,
-      "loss": 7.721007537841797,
-      "step": 800
-    },
-    {
-      "epoch": 14.035320088300221,
-      "eval_loss": 7.794251918792725,
-      "eval_runtime": 6.4539,
-      "eval_samples_per_second": 15.495,
-      "eval_steps_per_second": 7.747,
+      "grad_norm": 1.0366216897964478,
+      "learning_rate": 5.327485380116959e-05,
+      "loss": 7.734315490722656,
       "step": 800
     },
     {
       "epoch": 14.123620309050773,
-      "grad_norm": 1.0041288137435913,
-      "learning_rate": 7.178947368421054e-05,
-      "loss": 7.717352294921875,
+      "grad_norm": 1.1003059148788452,
+      "learning_rate": 5.2982456140350875e-05,
+      "loss": 7.734405517578125,
       "step": 805
     },
     {
       "epoch": 14.211920529801324,
-      "grad_norm": 1.33892822265625,
-      "learning_rate": 7.16140350877193e-05,
-      "loss": 7.6124725341796875,
+      "grad_norm": 1.0265618562698364,
+      "learning_rate": 5.269005847953217e-05,
+      "loss": 7.630126953125,
       "step": 810
     },
     {
       "epoch": 14.300220750551876,
-      "grad_norm": 0.9114180207252502,
-      "learning_rate": 7.143859649122808e-05,
-      "loss": 7.643186950683594,
+      "grad_norm": 0.8349760174751282,
+      "learning_rate": 5.2397660818713456e-05,
+      "loss": 7.664075469970703,
       "step": 815
     },
     {
       "epoch": 14.388520971302428,
-      "grad_norm": 1.032524824142456,
-      "learning_rate": 7.126315789473685e-05,
-      "loss": 7.698175048828125,
+      "grad_norm": 1.0714111328125,
+      "learning_rate": 5.210526315789474e-05,
+      "loss": 7.714525604248047,
       "step": 820
     },
     {
       "epoch": 14.47682119205298,
-      "grad_norm": 0.9611299633979797,
-      "learning_rate": 7.108771929824562e-05,
-      "loss": 7.699397277832031,
+      "grad_norm": 0.882725715637207,
+      "learning_rate": 5.1812865497076024e-05,
+      "loss": 7.717776489257813,
       "step": 825
     },
     {
       "epoch": 14.565121412803531,
-      "grad_norm": 1.1177457571029663,
-      "learning_rate": 7.091228070175439e-05,
-      "loss": 7.656683349609375,
+      "grad_norm": 0.972795844078064,
+      "learning_rate": 5.152046783625731e-05,
+      "loss": 7.673316192626953,
       "step": 830
     },
     {
       "epoch": 14.653421633554084,
-      "grad_norm": 0.9402754902839661,
-      "learning_rate": 7.073684210526316e-05,
-      "loss": 7.710595703125,
+      "grad_norm": 1.125019907951355,
+      "learning_rate": 5.1228070175438605e-05,
+      "loss": 7.724612426757813,
       "step": 835
     },
     {
       "epoch": 14.741721854304636,
-      "grad_norm": 1.3685545921325684,
-      "learning_rate": 7.056140350877193e-05,
-      "loss": 7.613078308105469,
+      "grad_norm": 1.2645599842071533,
+      "learning_rate": 5.093567251461988e-05,
+      "loss": 7.629542541503906,
       "step": 840
     },
     {
       "epoch": 14.830022075055188,
-      "grad_norm": 0.9739080667495728,
-      "learning_rate": 7.03859649122807e-05,
-      "loss": 7.692152404785157,
+      "grad_norm": 0.7888765335083008,
+      "learning_rate": 5.064327485380117e-05,
+      "loss": 7.708338928222656,
       "step": 845
     },
     {
       "epoch": 14.918322295805739,
-      "grad_norm": 1.3308042287826538,
-      "learning_rate": 7.021052631578949e-05,
-      "loss": 7.760948944091797,
+      "grad_norm": 0.9637308716773987,
+      "learning_rate": 5.035087719298246e-05,
+      "loss": 7.7762908935546875,
       "step": 850
     },
     {
       "epoch": 15.0,
-      "grad_norm": 1.3810853958129883,
-      "learning_rate": 7.003508771929826e-05,
-      "loss": 7.7296295166015625,
+      "grad_norm": 1.1657062768936157,
+      "learning_rate": 5.005847953216375e-05,
+      "loss": 7.745877838134765,
       "step": 855
     },
     {
       "epoch": 15.088300220750552,
-      "grad_norm": 1.1920428276062012,
-      "learning_rate": 6.985964912280703e-05,
-      "loss": 7.647270202636719,
+      "grad_norm": 1.0518969297409058,
+      "learning_rate": 4.9766081871345035e-05,
+      "loss": 7.669528961181641,
       "step": 860
     },
     {
       "epoch": 15.176600441501105,
-      "grad_norm": 1.2685319185256958,
-      "learning_rate": 6.96842105263158e-05,
-      "loss": 7.698595428466797,
+      "grad_norm": 1.1921542882919312,
+      "learning_rate": 4.9473684210526315e-05,
+      "loss": 7.719200134277344,
       "step": 865
     },
     {
       "epoch": 15.264900662251655,
-      "grad_norm": 1.2542896270751953,
-      "learning_rate": 6.950877192982457e-05,
-      "loss": 7.63843002319336,
+      "grad_norm": 0.9973335266113281,
+      "learning_rate": 4.91812865497076e-05,
+      "loss": 7.660971069335938,
       "step": 870
     },
     {
       "epoch": 15.353200883002208,
-      "grad_norm": 1.675700068473816,
-      "learning_rate": 6.933333333333334e-05,
-      "loss": 7.586814880371094,
+      "grad_norm": 1.463598370552063,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 7.609901428222656,
       "step": 875
     },
     {
       "epoch": 15.44150110375276,
-      "grad_norm": 1.493030309677124,
-      "learning_rate": 6.915789473684211e-05,
-      "loss": 7.547953033447266,
+      "grad_norm": 0.9667314887046814,
+      "learning_rate": 4.859649122807018e-05,
+      "loss": 7.575991821289063,
       "step": 880
     },
     {
       "epoch": 15.52980132450331,
-      "grad_norm": 1.2002837657928467,
-      "learning_rate": 6.898245614035088e-05,
-      "loss": 7.665699768066406,
+      "grad_norm": 1.1865555047988892,
+      "learning_rate": 4.8304093567251464e-05,
+      "loss": 7.685313415527344,
       "step": 885
     },
     {
       "epoch": 15.618101545253863,
-      "grad_norm": 1.176683783531189,
-      "learning_rate": 6.880701754385966e-05,
-      "loss": 7.657095336914063,
+      "grad_norm": 1.10698401927948,
+      "learning_rate": 4.801169590643275e-05,
+      "loss": 7.676809692382813,
       "step": 890
     },
     {
       "epoch": 15.706401766004415,
-      "grad_norm": 1.0782877206802368,
-      "learning_rate": 6.863157894736843e-05,
-      "loss": 7.686164855957031,
+      "grad_norm": 0.9209511876106262,
+      "learning_rate": 4.771929824561404e-05,
+      "loss": 7.709003448486328,
       "step": 895
     },
     {
       "epoch": 15.794701986754967,
-      "grad_norm": 0.9397177696228027,
-      "learning_rate": 6.84561403508772e-05,
-      "loss": 7.728889465332031,
+      "grad_norm": 1.0440056324005127,
+      "learning_rate": 4.7426900584795326e-05,
+      "loss": 7.752204895019531,
       "step": 900
     },
     {
       "epoch": 15.794701986754967,
-      "eval_loss": 7.772979259490967,
-      "eval_runtime": 6.4336,
-      "eval_samples_per_second": 15.543,
-      "eval_steps_per_second": 7.772,
+      "eval_loss": 7.7843523025512695,
+      "eval_runtime": 6.4306,
+      "eval_samples_per_second": 15.551,
+      "eval_steps_per_second": 7.775,
       "step": 900
+    },
+    {
+      "epoch": 15.883002207505518,
+      "grad_norm": 1.0428849458694458,
+      "learning_rate": 4.7134502923976607e-05,
+      "loss": 7.657199859619141,
+      "step": 905
+    },
+    {
+      "epoch": 15.97130242825607,
+      "grad_norm": 1.2458996772766113,
+      "learning_rate": 4.68421052631579e-05,
+      "loss": 7.6968017578125,
+      "step": 910
+    },
+    {
+      "epoch": 16.05298013245033,
+      "grad_norm": 1.4557669162750244,
+      "learning_rate": 4.654970760233918e-05,
+      "loss": 7.624080657958984,
+      "step": 915
+    },
+    {
+      "epoch": 16.141280353200884,
+      "grad_norm": 1.2798360586166382,
+      "learning_rate": 4.625730994152047e-05,
+      "loss": 7.594389343261719,
+      "step": 920
+    },
+    {
+      "epoch": 16.229580573951434,
+      "grad_norm": 1.331041932106018,
+      "learning_rate": 4.5964912280701756e-05,
+      "loss": 7.655438995361328,
+      "step": 925
+    },
+    {
+      "epoch": 16.31788079470199,
+      "grad_norm": 0.9816513657569885,
+      "learning_rate": 4.567251461988304e-05,
+      "loss": 7.63082275390625,
+      "step": 930
+    },
+    {
+      "epoch": 16.40618101545254,
+      "grad_norm": 1.3507262468338013,
+      "learning_rate": 4.538011695906433e-05,
+      "loss": 7.636911010742187,
+      "step": 935
+    },
+    {
+      "epoch": 16.49448123620309,
+      "grad_norm": 1.2413368225097656,
+      "learning_rate": 4.508771929824562e-05,
+      "loss": 7.713497924804687,
+      "step": 940
+    },
+    {
+      "epoch": 16.582781456953644,
+      "grad_norm": 0.8101910352706909,
+      "learning_rate": 4.4795321637426905e-05,
+      "loss": 7.683601379394531,
+      "step": 945
+    },
+    {
+      "epoch": 16.671081677704194,
+      "grad_norm": 1.1739321947097778,
+      "learning_rate": 4.450292397660819e-05,
+      "loss": 7.682159423828125,
+      "step": 950
+    },
+    {
+      "epoch": 16.759381898454745,
+      "grad_norm": 1.3702641725540161,
+      "learning_rate": 4.421052631578947e-05,
+      "loss": 7.643263244628907,
+      "step": 955
+    },
+    {
+      "epoch": 16.8476821192053,
+      "grad_norm": 1.1454497575759888,
+      "learning_rate": 4.3918128654970766e-05,
+      "loss": 7.628999328613281,
+      "step": 960
+    },
+    {
+      "epoch": 16.93598233995585,
+      "grad_norm": 1.1330296993255615,
+      "learning_rate": 4.362573099415205e-05,
+      "loss": 7.696270751953125,
+      "step": 965
+    },
+    {
+      "epoch": 17.01766004415011,
+      "grad_norm": 1.3744090795516968,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 7.638075256347657,
+      "step": 970
+    },
+    {
+      "epoch": 17.105960264900663,
+      "grad_norm": 1.225696086883545,
+      "learning_rate": 4.304093567251462e-05,
+      "loss": 7.642805480957032,
+      "step": 975
+    },
+    {
+      "epoch": 17.194260485651213,
+      "grad_norm": 1.040528416633606,
+      "learning_rate": 4.274853801169591e-05,
+      "loss": 7.692190551757813,
+      "step": 980
+    },
+    {
+      "epoch": 17.282560706401767,
+      "grad_norm": 1.1747312545776367,
+      "learning_rate": 4.2456140350877196e-05,
+      "loss": 7.6657867431640625,
+      "step": 985
+    },
+    {
+      "epoch": 17.370860927152318,
+      "grad_norm": 1.628137230873108,
+      "learning_rate": 4.216374269005848e-05,
+      "loss": 7.625160217285156,
+      "step": 990
+    },
+    {
+      "epoch": 17.45916114790287,
+      "grad_norm": 1.4963520765304565,
+      "learning_rate": 4.1871345029239764e-05,
+      "loss": 7.607908630371094,
+      "step": 995
+    },
+    {
+      "epoch": 17.547461368653423,
+      "grad_norm": 1.3517507314682007,
+      "learning_rate": 4.157894736842106e-05,
+      "loss": 7.665956115722656,
+      "step": 1000
+    },
+    {
+      "epoch": 17.635761589403973,
+      "grad_norm": 1.0717331171035767,
+      "learning_rate": 4.128654970760234e-05,
+      "loss": 7.5733787536621096,
+      "step": 1005
+    },
+    {
+      "epoch": 17.724061810154524,
+      "grad_norm": 1.2811627388000488,
+      "learning_rate": 4.099415204678363e-05,
+      "loss": 7.589814758300781,
+      "step": 1010
+    },
+    {
+      "epoch": 17.812362030905078,
+      "grad_norm": 1.0915024280548096,
+      "learning_rate": 4.070175438596491e-05,
+      "loss": 7.614756011962891,
+      "step": 1015
+    },
+    {
+      "epoch": 17.90066225165563,
+      "grad_norm": 1.2693991661071777,
+      "learning_rate": 4.04093567251462e-05,
+      "loss": 7.637123107910156,
+      "step": 1020
+    },
+    {
+      "epoch": 17.988962472406183,
+      "grad_norm": 1.4724719524383545,
+      "learning_rate": 4.011695906432749e-05,
+      "loss": 7.569842529296875,
+      "step": 1025
+    },
+    {
+      "epoch": 18.070640176600442,
+      "grad_norm": 1.2235865592956543,
+      "learning_rate": 3.9824561403508774e-05,
+      "loss": 7.599622344970703,
+      "step": 1030
+    },
+    {
+      "epoch": 18.158940397350992,
+      "grad_norm": 1.1637712717056274,
+      "learning_rate": 3.953216374269006e-05,
+      "loss": 7.6607002258300785,
+      "step": 1035
+    },
+    {
+      "epoch": 18.247240618101547,
+      "grad_norm": 1.1987440586090088,
+      "learning_rate": 3.923976608187135e-05,
+      "loss": 7.622169494628906,
+      "step": 1040
+    },
+    {
+      "epoch": 18.335540838852097,
+      "grad_norm": 1.1098897457122803,
+      "learning_rate": 3.894736842105263e-05,
+      "loss": 7.577146911621094,
+      "step": 1045
+    },
+    {
+      "epoch": 18.423841059602648,
+      "grad_norm": 1.5642309188842773,
+      "learning_rate": 3.865497076023392e-05,
+      "loss": 7.521853637695313,
+      "step": 1050
+    },
+    {
+      "epoch": 18.423841059602648,
+      "eval_loss": 7.780308246612549,
+      "eval_runtime": 6.4382,
+      "eval_samples_per_second": 15.532,
+      "eval_steps_per_second": 7.766,
+      "step": 1050
     }
   ],
   "logging_steps": 5,
-  "max_steps": 2850,
+  "max_steps": 1710,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 50,
-  "save_steps": 100,
+  "num_train_epochs": 30,
+  "save_steps": 150,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -1359,7 +1553,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.787689275481129e+17,
+  "total_flos": 5.5847672511830426e+17,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null

adapter/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b7eea1c75bf1b71215bf3bc7f7303dc74f22c391447a62b2646402fa4883682
+oid sha256:f695763f609e3e99672af968d6bed61f1dd8e19afae2db4ff9cc0b3329fc9925
 size 4792