End of training

README.md

@@ -4,6 +4,7 @@ license: mit
 base_model: agentica-org/DeepCoder-14B-Preview
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: SWE-BENCH-2000-enriched-reasoning-claude-localization_qwen_code_14b_2000_enriched_reasoning
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # SWE-BENCH-2000-enriched-reasoning-claude-localization_qwen_code_14b_2000_enriched_reasoning
 
-This model is a fine-tuned version of [agentica-org/DeepCoder-14B-Preview](https://huggingface.co/agentica-org/DeepCoder-14B-Preview) on an unknown dataset.
+This model is a fine-tuned version of [agentica-org/DeepCoder-14B-Preview](https://huggingface.co/agentica-org/DeepCoder-14B-Preview) on the SWE-BENCH-2000-enriched-reasoning-claude-localization dataset.
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.982264665757162,
+    "total_flos": 5.395725452134318e+17,
+    "train_loss": 0.5498430520458951,
+    "train_runtime": 28891.7819,
+    "train_samples_per_second": 0.609,
+    "train_steps_per_second": 0.013
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.982264665757162,
+    "total_flos": 5.395725452134318e+17,
+    "train_loss": 0.5498430520458951,
+    "train_runtime": 28891.7819,
+    "train_samples_per_second": 0.609,
+    "train_steps_per_second": 0.013
+}

trainer_state.json

@@ -0,0 +1,2605 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.982264665757162,
+  "eval_steps": 500,
+  "global_step": 366,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008185538881309686,
+      "grad_norm": 5.047900636948469,
+      "learning_rate": 0.0,
+      "loss": 1.0078,
+      "step": 1
+    },
+    {
+      "epoch": 0.01637107776261937,
+      "grad_norm": 5.2987761219424625,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 1.0618,
+      "step": 2
+    },
+    {
+      "epoch": 0.02455661664392906,
+      "grad_norm": 5.22433195772352,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.9962,
+      "step": 3
+    },
+    {
+      "epoch": 0.03274215552523874,
+      "grad_norm": 4.925223250266536,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 1.0561,
+      "step": 4
+    },
+    {
+      "epoch": 0.040927694406548434,
+      "grad_norm": 4.885522263345388,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 1.0432,
+      "step": 5
+    },
+    {
+      "epoch": 0.04911323328785812,
+      "grad_norm": 5.009721578407571,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 1.0377,
+      "step": 6
+    },
+    {
+      "epoch": 0.0572987721691678,
+      "grad_norm": 4.079479665797117,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 1.0026,
+      "step": 7
+    },
+    {
+      "epoch": 0.06548431105047749,
+      "grad_norm": 3.7320192096023583,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.9673,
+      "step": 8
+    },
+    {
+      "epoch": 0.07366984993178717,
+      "grad_norm": 3.6443053135214356,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.98,
+      "step": 9
+    },
+    {
+      "epoch": 0.08185538881309687,
+      "grad_norm": 2.5301682637219276,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.9222,
+      "step": 10
+    },
+    {
+      "epoch": 0.09004092769440655,
+      "grad_norm": 2.3134082671081404,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.9259,
+      "step": 11
+    },
+    {
+      "epoch": 0.09822646657571624,
+      "grad_norm": 2.268917597794808,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.9505,
+      "step": 12
+    },
+    {
+      "epoch": 0.10641200545702592,
+      "grad_norm": 1.6472898181111486,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.8894,
+      "step": 13
+    },
+    {
+      "epoch": 0.1145975443383356,
+      "grad_norm": 1.5832024939026488,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.8923,
+      "step": 14
+    },
+    {
+      "epoch": 0.12278308321964529,
+      "grad_norm": 1.4463210728565452,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.8531,
+      "step": 15
+    },
+    {
+      "epoch": 0.13096862210095497,
+      "grad_norm": 1.3259843494584223,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 0.8401,
+      "step": 16
+    },
+    {
+      "epoch": 0.13915416098226466,
+      "grad_norm": 1.1897212939154778,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.8209,
+      "step": 17
+    },
+    {
+      "epoch": 0.14733969986357434,
+      "grad_norm": 1.1245723639473528,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 0.7891,
+      "step": 18
+    },
+    {
+      "epoch": 0.15552523874488403,
+      "grad_norm": 0.996785394559854,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.7682,
+      "step": 19
+    },
+    {
+      "epoch": 0.16371077762619374,
+      "grad_norm": 1.0126927854201089,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.789,
+      "step": 20
+    },
+    {
+      "epoch": 0.17189631650750342,
+      "grad_norm": 0.9375631398803144,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.7929,
+      "step": 21
+    },
+    {
+      "epoch": 0.1800818553888131,
+      "grad_norm": 0.9285154735154034,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.7703,
+      "step": 22
+    },
+    {
+      "epoch": 0.1882673942701228,
+      "grad_norm": 0.867212289749987,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.7706,
+      "step": 23
+    },
+    {
+      "epoch": 0.19645293315143247,
+      "grad_norm": 0.8599939840408395,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.7514,
+      "step": 24
+    },
+    {
+      "epoch": 0.20463847203274216,
+      "grad_norm": 0.8634713379832526,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.7819,
+      "step": 25
+    },
+    {
+      "epoch": 0.21282401091405184,
+      "grad_norm": 0.8747636881078926,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 0.7554,
+      "step": 26
+    },
+    {
+      "epoch": 0.22100954979536153,
+      "grad_norm": 0.9181726806879716,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.7558,
+      "step": 27
+    },
+    {
+      "epoch": 0.2291950886766712,
+      "grad_norm": 0.7498863543954535,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 0.6947,
+      "step": 28
+    },
+    {
+      "epoch": 0.2373806275579809,
+      "grad_norm": 0.7959019444169159,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 0.7303,
+      "step": 29
+    },
+    {
+      "epoch": 0.24556616643929058,
+      "grad_norm": 0.7877801896430789,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 0.7068,
+      "step": 30
+    },
+    {
+      "epoch": 0.25375170532060026,
+      "grad_norm": 0.772905987576781,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.7515,
+      "step": 31
+    },
+    {
+      "epoch": 0.26193724420190995,
+      "grad_norm": 0.6726982296746401,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 0.6874,
+      "step": 32
+    },
+    {
+      "epoch": 0.27012278308321963,
+      "grad_norm": 0.6981101156625776,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.6868,
+      "step": 33
+    },
+    {
+      "epoch": 0.2783083219645293,
+      "grad_norm": 0.8348069175195246,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 0.7238,
+      "step": 34
+    },
+    {
+      "epoch": 0.286493860845839,
+      "grad_norm": 0.7509783495037521,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 0.6648,
+      "step": 35
+    },
+    {
+      "epoch": 0.2946793997271487,
+      "grad_norm": 0.784785626436805,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 0.7305,
+      "step": 36
+    },
+    {
+      "epoch": 0.30286493860845837,
+      "grad_norm": 0.7784217280083267,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.72,
+      "step": 37
+    },
+    {
+      "epoch": 0.31105047748976805,
+      "grad_norm": 0.6792555897775979,
+      "learning_rate": 1e-05,
+      "loss": 0.6846,
+      "step": 38
+    },
+    {
+      "epoch": 0.31923601637107774,
+      "grad_norm": 0.7083077195323103,
+      "learning_rate": 9.999772047343259e-06,
+      "loss": 0.71,
+      "step": 39
+    },
+    {
+      "epoch": 0.3274215552523875,
+      "grad_norm": 0.7443230527609791,
+      "learning_rate": 9.999088210158001e-06,
+      "loss": 0.7021,
+      "step": 40
+    },
+    {
+      "epoch": 0.33560709413369716,
+      "grad_norm": 0.7154970338915503,
+      "learning_rate": 9.997948550797227e-06,
+      "loss": 0.686,
+      "step": 41
+    },
+    {
+      "epoch": 0.34379263301500684,
+      "grad_norm": 0.7296710924336983,
+      "learning_rate": 9.99635317317629e-06,
+      "loss": 0.7018,
+      "step": 42
+    },
+    {
+      "epoch": 0.3519781718963165,
+      "grad_norm": 0.7259044278286824,
+      "learning_rate": 9.994302222763415e-06,
+      "loss": 0.6813,
+      "step": 43
+    },
+    {
+      "epoch": 0.3601637107776262,
+      "grad_norm": 0.740222584457023,
+      "learning_rate": 9.991795886566443e-06,
+      "loss": 0.6657,
+      "step": 44
+    },
+    {
+      "epoch": 0.3683492496589359,
+      "grad_norm": 0.7155278090635278,
+      "learning_rate": 9.988834393115768e-06,
+      "loss": 0.683,
+      "step": 45
+    },
+    {
+      "epoch": 0.3765347885402456,
+      "grad_norm": 0.7096691812719568,
+      "learning_rate": 9.98541801244351e-06,
+      "loss": 0.6981,
+      "step": 46
+    },
+    {
+      "epoch": 0.38472032742155526,
+      "grad_norm": 0.7446388890832425,
+      "learning_rate": 9.981547056058893e-06,
+      "loss": 0.6972,
+      "step": 47
+    },
+    {
+      "epoch": 0.39290586630286495,
+      "grad_norm": 0.7152687230516146,
+      "learning_rate": 9.977221876919833e-06,
+      "loss": 0.677,
+      "step": 48
+    },
+    {
+      "epoch": 0.40109140518417463,
+      "grad_norm": 0.747253188787375,
+      "learning_rate": 9.97244286940076e-06,
+      "loss": 0.6812,
+      "step": 49
+    },
+    {
+      "epoch": 0.4092769440654843,
+      "grad_norm": 0.6911041309644335,
+      "learning_rate": 9.967210469256657e-06,
+      "loss": 0.6878,
+      "step": 50
+    },
+    {
+      "epoch": 0.417462482946794,
+      "grad_norm": 0.7131373784880878,
+      "learning_rate": 9.961525153583327e-06,
+      "loss": 0.673,
+      "step": 51
+    },
+    {
+      "epoch": 0.4256480218281037,
+      "grad_norm": 0.7564271286626736,
+      "learning_rate": 9.955387440773902e-06,
+      "loss": 0.661,
+      "step": 52
+    },
+    {
+      "epoch": 0.43383356070941337,
+      "grad_norm": 0.7544926363411585,
+      "learning_rate": 9.948797890471552e-06,
+      "loss": 0.7061,
+      "step": 53
+    },
+    {
+      "epoch": 0.44201909959072305,
+      "grad_norm": 0.7649555238443022,
+      "learning_rate": 9.94175710351848e-06,
+      "loss": 0.681,
+      "step": 54
+    },
+    {
+      "epoch": 0.45020463847203274,
+      "grad_norm": 0.7587260643246102,
+      "learning_rate": 9.93426572190112e-06,
+      "loss": 0.6885,
+      "step": 55
+    },
+    {
+      "epoch": 0.4583901773533424,
+      "grad_norm": 0.7789441239465993,
+      "learning_rate": 9.926324428691612e-06,
+      "loss": 0.6742,
+      "step": 56
+    },
+    {
+      "epoch": 0.4665757162346521,
+      "grad_norm": 0.774593544062445,
+      "learning_rate": 9.917933947985508e-06,
+      "loss": 0.6513,
+      "step": 57
+    },
+    {
+      "epoch": 0.4747612551159618,
+      "grad_norm": 0.7422600042223242,
+      "learning_rate": 9.909095044835755e-06,
+      "loss": 0.6637,
+      "step": 58
+    },
+    {
+      "epoch": 0.4829467939972715,
+      "grad_norm": 0.6974210308238069,
+      "learning_rate": 9.899808525182935e-06,
+      "loss": 0.6727,
+      "step": 59
+    },
+    {
+      "epoch": 0.49113233287858116,
+      "grad_norm": 0.770047615620957,
+      "learning_rate": 9.89007523578178e-06,
+      "loss": 0.6718,
+      "step": 60
+    },
+    {
+      "epoch": 0.49931787175989084,
+      "grad_norm": 0.7292536875451305,
+      "learning_rate": 9.879896064123961e-06,
+      "loss": 0.6869,
+      "step": 61
+    },
+    {
+      "epoch": 0.5075034106412005,
+      "grad_norm": 0.6447105816936861,
+      "learning_rate": 9.869271938357168e-06,
+      "loss": 0.6651,
+      "step": 62
+    },
+    {
+      "epoch": 0.5156889495225102,
+      "grad_norm": 0.6879618636331486,
+      "learning_rate": 9.858203827200477e-06,
+      "loss": 0.6684,
+      "step": 63
+    },
+    {
+      "epoch": 0.5238744884038199,
+      "grad_norm": 0.6610264465102351,
+      "learning_rate": 9.846692739856023e-06,
+      "loss": 0.6861,
+      "step": 64
+    },
+    {
+      "epoch": 0.5320600272851296,
+      "grad_norm": 0.6519434892651405,
+      "learning_rate": 9.834739725916988e-06,
+      "loss": 0.6978,
+      "step": 65
+    },
+    {
+      "epoch": 0.5402455661664393,
+      "grad_norm": 0.6733480098205279,
+      "learning_rate": 9.822345875271884e-06,
+      "loss": 0.6861,
+      "step": 66
+    },
+    {
+      "epoch": 0.548431105047749,
+      "grad_norm": 0.6658797030885003,
+      "learning_rate": 9.80951231800518e-06,
+      "loss": 0.666,
+      "step": 67
+    },
+    {
+      "epoch": 0.5566166439290586,
+      "grad_norm": 0.6735946082472756,
+      "learning_rate": 9.79624022429427e-06,
+      "loss": 0.6678,
+      "step": 68
+    },
+    {
+      "epoch": 0.5648021828103683,
+      "grad_norm": 0.6744460007015742,
+      "learning_rate": 9.782530804302763e-06,
+      "loss": 0.6711,
+      "step": 69
+    },
+    {
+      "epoch": 0.572987721691678,
+      "grad_norm": 0.6620392972422307,
+      "learning_rate": 9.768385308070139e-06,
+      "loss": 0.6501,
+      "step": 70
+    },
+    {
+      "epoch": 0.5811732605729877,
+      "grad_norm": 0.6387994624033349,
+      "learning_rate": 9.75380502539778e-06,
+      "loss": 0.6382,
+      "step": 71
+    },
+    {
+      "epoch": 0.5893587994542974,
+      "grad_norm": 0.6772258322102017,
+      "learning_rate": 9.738791285731353e-06,
+      "loss": 0.6274,
+      "step": 72
+    },
+    {
+      "epoch": 0.597544338335607,
+      "grad_norm": 0.6825021235551098,
+      "learning_rate": 9.723345458039595e-06,
+      "loss": 0.6541,
+      "step": 73
+    },
+    {
+      "epoch": 0.6057298772169167,
+      "grad_norm": 0.6929813570605768,
+      "learning_rate": 9.70746895068949e-06,
+      "loss": 0.6442,
+      "step": 74
+    },
+    {
+      "epoch": 0.6139154160982264,
+      "grad_norm": 1.9433763442086747,
+      "learning_rate": 9.691163211317853e-06,
+      "loss": 0.697,
+      "step": 75
+    },
+    {
+      "epoch": 0.6221009549795361,
+      "grad_norm": 0.7013057534171012,
+      "learning_rate": 9.674429726699324e-06,
+      "loss": 0.6632,
+      "step": 76
+    },
+    {
+      "epoch": 0.6302864938608458,
+      "grad_norm": 0.6865730770709916,
+      "learning_rate": 9.657270022610814e-06,
+      "loss": 0.6654,
+      "step": 77
+    },
+    {
+      "epoch": 0.6384720327421555,
+      "grad_norm": 0.6247870338564917,
+      "learning_rate": 9.63968566369238e-06,
+      "loss": 0.6236,
+      "step": 78
+    },
+    {
+      "epoch": 0.6466575716234653,
+      "grad_norm": 0.682002762069422,
+      "learning_rate": 9.62167825330455e-06,
+      "loss": 0.6611,
+      "step": 79
+    },
+    {
+      "epoch": 0.654843110504775,
+      "grad_norm": 0.6725874189078757,
+      "learning_rate": 9.603249433382145e-06,
+      "loss": 0.6849,
+      "step": 80
+    },
+    {
+      "epoch": 0.6630286493860846,
+      "grad_norm": 0.6216501614963521,
+      "learning_rate": 9.584400884284546e-06,
+      "loss": 0.6457,
+      "step": 81
+    },
+    {
+      "epoch": 0.6712141882673943,
+      "grad_norm": 0.6882049477900337,
+      "learning_rate": 9.565134324642491e-06,
+      "loss": 0.6695,
+      "step": 82
+    },
+    {
+      "epoch": 0.679399727148704,
+      "grad_norm": 0.668128651644651,
+      "learning_rate": 9.545451511201365e-06,
+      "loss": 0.6657,
+      "step": 83
+    },
+    {
+      "epoch": 0.6875852660300137,
+      "grad_norm": 0.6794306150709426,
+      "learning_rate": 9.52535423866101e-06,
+      "loss": 0.6768,
+      "step": 84
+    },
+    {
+      "epoch": 0.6957708049113234,
+      "grad_norm": 0.631448511000336,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.6254,
+      "step": 85
+    },
+    {
+      "epoch": 0.703956343792633,
+      "grad_norm": 0.6877362534475567,
+      "learning_rate": 9.483923683869025e-06,
+      "loss": 0.6272,
+      "step": 86
+    },
+    {
+      "epoch": 0.7121418826739427,
+      "grad_norm": 0.7186232799606551,
+      "learning_rate": 9.462594179299408e-06,
+      "loss": 0.6652,
+      "step": 87
+    },
+    {
+      "epoch": 0.7203274215552524,
+      "grad_norm": 0.7066004569621066,
+      "learning_rate": 9.440857770650139e-06,
+      "loss": 0.7179,
+      "step": 88
+    },
+    {
+      "epoch": 0.7285129604365621,
+      "grad_norm": 0.8511137252282978,
+      "learning_rate": 9.418716439870056e-06,
+      "loss": 0.6228,
+      "step": 89
+    },
+    {
+      "epoch": 0.7366984993178718,
+      "grad_norm": 0.7048676512844883,
+      "learning_rate": 9.396172205829235e-06,
+      "loss": 0.6737,
+      "step": 90
+    },
+    {
+      "epoch": 0.7448840381991815,
+      "grad_norm": 0.727762016803116,
+      "learning_rate": 9.373227124134888e-06,
+      "loss": 0.6471,
+      "step": 91
+    },
+    {
+      "epoch": 0.7530695770804912,
+      "grad_norm": 0.6767761891949757,
+      "learning_rate": 9.349883286943951e-06,
+      "loss": 0.6328,
+      "step": 92
+    },
+    {
+      "epoch": 0.7612551159618008,
+      "grad_norm": 0.6862932738256654,
+      "learning_rate": 9.326142822772301e-06,
+      "loss": 0.665,
+      "step": 93
+    },
+    {
+      "epoch": 0.7694406548431105,
+      "grad_norm": 0.6532316480867286,
+      "learning_rate": 9.302007896300697e-06,
+      "loss": 0.6168,
+      "step": 94
+    },
+    {
+      "epoch": 0.7776261937244202,
+      "grad_norm": 0.65952295383208,
+      "learning_rate": 9.27748070817738e-06,
+      "loss": 0.6672,
+      "step": 95
+    },
+    {
+      "epoch": 0.7858117326057299,
+      "grad_norm": 0.6150250153464631,
+      "learning_rate": 9.252563494817426e-06,
+      "loss": 0.6642,
+      "step": 96
+    },
+    {
+      "epoch": 0.7939972714870396,
+      "grad_norm": 0.646477017301684,
+      "learning_rate": 9.227258528198832e-06,
+      "loss": 0.6556,
+      "step": 97
+    },
+    {
+      "epoch": 0.8021828103683493,
+      "grad_norm": 0.6868408186286113,
+      "learning_rate": 9.201568115655343e-06,
+      "loss": 0.6624,
+      "step": 98
+    },
+    {
+      "epoch": 0.810368349249659,
+      "grad_norm": 0.675658718340078,
+      "learning_rate": 9.175494599666078e-06,
+      "loss": 0.6602,
+      "step": 99
+    },
+    {
+      "epoch": 0.8185538881309686,
+      "grad_norm": 0.6507394927697661,
+      "learning_rate": 9.14904035764193e-06,
+      "loss": 0.6403,
+      "step": 100
+    },
+    {
+      "epoch": 0.8267394270122783,
+      "grad_norm": 0.6500323953371057,
+      "learning_rate": 9.122207801708802e-06,
+      "loss": 0.6958,
+      "step": 101
+    },
+    {
+      "epoch": 0.834924965893588,
+      "grad_norm": 0.6306391581625795,
+      "learning_rate": 9.094999378487659e-06,
+      "loss": 0.6553,
+      "step": 102
+    },
+    {
+      "epoch": 0.8431105047748977,
+      "grad_norm": 0.7566746505318314,
+      "learning_rate": 9.067417568871444e-06,
+      "loss": 0.6349,
+      "step": 103
+    },
+    {
+      "epoch": 0.8512960436562074,
+      "grad_norm": 0.68419614590406,
+      "learning_rate": 9.03946488779887e-06,
+      "loss": 0.6439,
+      "step": 104
+    },
+    {
+      "epoch": 0.859481582537517,
+      "grad_norm": 0.6929693472710352,
+      "learning_rate": 9.0111438840251e-06,
+      "loss": 0.6399,
+      "step": 105
+    },
+    {
+      "epoch": 0.8676671214188267,
+      "grad_norm": 0.6304833742834234,
+      "learning_rate": 8.982457139889358e-06,
+      "loss": 0.6235,
+      "step": 106
+    },
+    {
+      "epoch": 0.8758526603001364,
+      "grad_norm": 0.6432175058885881,
+      "learning_rate": 8.953407271079456e-06,
+      "loss": 0.6209,
+      "step": 107
+    },
+    {
+      "epoch": 0.8840381991814461,
+      "grad_norm": 0.6441099431808613,
+      "learning_rate": 8.923996926393306e-06,
+      "loss": 0.6434,
+      "step": 108
+    },
+    {
+      "epoch": 0.8922237380627558,
+      "grad_norm": 0.6546849172855025,
+      "learning_rate": 8.894228787497389e-06,
+      "loss": 0.648,
+      "step": 109
+    },
+    {
+      "epoch": 0.9004092769440655,
+      "grad_norm": 0.6522310096241787,
+      "learning_rate": 8.864105568682245e-06,
+      "loss": 0.616,
+      "step": 110
+    },
+    {
+      "epoch": 0.9085948158253752,
+      "grad_norm": 0.6505051090069522,
+      "learning_rate": 8.833630016614976e-06,
+      "loss": 0.66,
+      "step": 111
+    },
+    {
+      "epoch": 0.9167803547066848,
+      "grad_norm": 0.6164475760915742,
+      "learning_rate": 8.80280491008881e-06,
+      "loss": 0.6545,
+      "step": 112
+    },
+    {
+      "epoch": 0.9249658935879945,
+      "grad_norm": 0.6642637129874517,
+      "learning_rate": 8.771633059769712e-06,
+      "loss": 0.6477,
+      "step": 113
+    },
+    {
+      "epoch": 0.9331514324693042,
+      "grad_norm": 0.6225767272815247,
+      "learning_rate": 8.740117307940123e-06,
+      "loss": 0.6368,
+      "step": 114
+    },
+    {
+      "epoch": 0.9413369713506139,
+      "grad_norm": 0.6632125853712576,
+      "learning_rate": 8.708260528239788e-06,
+      "loss": 0.6829,
+      "step": 115
+    },
+    {
+      "epoch": 0.9495225102319236,
+      "grad_norm": 0.6343076535553938,
+      "learning_rate": 8.676065625403733e-06,
+      "loss": 0.6549,
+      "step": 116
+    },
+    {
+      "epoch": 0.9577080491132333,
+      "grad_norm": 0.6397774668350522,
+      "learning_rate": 8.64353553499741e-06,
+      "loss": 0.6278,
+      "step": 117
+    },
+    {
+      "epoch": 0.965893587994543,
+      "grad_norm": 0.6164487315135155,
+      "learning_rate": 8.610673223149036e-06,
+      "loss": 0.6373,
+      "step": 118
+    },
+    {
+      "epoch": 0.9740791268758526,
+      "grad_norm": 0.6664209310917184,
+      "learning_rate": 8.577481686279123e-06,
+      "loss": 0.6478,
+      "step": 119
+    },
+    {
+      "epoch": 0.9822646657571623,
+      "grad_norm": 0.663088073349988,
+      "learning_rate": 8.543963950827279e-06,
+      "loss": 0.6462,
+      "step": 120
+    },
+    {
+      "epoch": 0.990450204638472,
+      "grad_norm": 0.640336290815735,
+      "learning_rate": 8.51012307297624e-06,
+      "loss": 0.6635,
+      "step": 121
+    },
+    {
+      "epoch": 0.9986357435197817,
+      "grad_norm": 0.6440131782350959,
+      "learning_rate": 8.475962138373212e-06,
+      "loss": 0.6559,
+      "step": 122
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6440131782350959,
+      "learning_rate": 8.441484261848514e-06,
+      "loss": 0.6062,
+      "step": 123
+    },
+    {
+      "epoch": 1.0081855388813097,
+      "grad_norm": 1.5135652872806824,
+      "learning_rate": 8.406692587131569e-06,
+      "loss": 0.5367,
+      "step": 124
+    },
+    {
+      "epoch": 1.0163710777626194,
+      "grad_norm": 0.661432220591032,
+      "learning_rate": 8.371590286564247e-06,
+      "loss": 0.5697,
+      "step": 125
+    },
+    {
+      "epoch": 1.024556616643929,
+      "grad_norm": 0.6654816977894965,
+      "learning_rate": 8.336180560811619e-06,
+      "loss": 0.5299,
+      "step": 126
+    },
+    {
+      "epoch": 1.0327421555252387,
+      "grad_norm": 0.6287499014119071,
+      "learning_rate": 8.30046663857011e-06,
+      "loss": 0.558,
+      "step": 127
+    },
+    {
+      "epoch": 1.0409276944065484,
+      "grad_norm": 0.6293447887045929,
+      "learning_rate": 8.264451776273104e-06,
+      "loss": 0.5327,
+      "step": 128
+    },
+    {
+      "epoch": 1.049113233287858,
+      "grad_norm": 0.6667791830390026,
+      "learning_rate": 8.228139257794012e-06,
+      "loss": 0.529,
+      "step": 129
+    },
+    {
+      "epoch": 1.0572987721691678,
+      "grad_norm": 0.6699713976601184,
+      "learning_rate": 8.191532394146865e-06,
+      "loss": 0.5332,
+      "step": 130
+    },
+    {
+      "epoch": 1.0654843110504775,
+      "grad_norm": 0.7028538589270679,
+      "learning_rate": 8.154634523184389e-06,
+      "loss": 0.5505,
+      "step": 131
+    },
+    {
+      "epoch": 1.0736698499317872,
+      "grad_norm": 0.7021518320532332,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.5586,
+      "step": 132
+    },
+    {
+      "epoch": 1.0818553888130968,
+      "grad_norm": 0.7054628592464113,
+      "learning_rate": 8.07997924308938e-06,
+      "loss": 0.5167,
+      "step": 133
+    },
+    {
+      "epoch": 1.0900409276944065,
+      "grad_norm": 0.6688530715797821,
+      "learning_rate": 8.042228641104622e-06,
+      "loss": 0.529,
+      "step": 134
+    },
+    {
+      "epoch": 1.0982264665757162,
+      "grad_norm": 0.6745188963170561,
+      "learning_rate": 8.004200645479403e-06,
+      "loss": 0.5428,
+      "step": 135
+    },
+    {
+      "epoch": 1.106412005457026,
+      "grad_norm": 0.6418923060681222,
+      "learning_rate": 7.965898723646777e-06,
+      "loss": 0.5025,
+      "step": 136
+    },
+    {
+      "epoch": 1.1145975443383356,
+      "grad_norm": 0.6614931832576282,
+      "learning_rate": 7.927326368016677e-06,
+      "loss": 0.5204,
+      "step": 137
+    },
+    {
+      "epoch": 1.1227830832196453,
+      "grad_norm": 0.6742200529862905,
+      "learning_rate": 7.888487095657484e-06,
+      "loss": 0.5618,
+      "step": 138
+    },
+    {
+      "epoch": 1.130968622100955,
+      "grad_norm": 0.7432230779617753,
+      "learning_rate": 7.849384447975322e-06,
+      "loss": 0.5364,
+      "step": 139
+    },
+    {
+      "epoch": 1.1391541609822646,
+      "grad_norm": 0.6721951352204065,
+      "learning_rate": 7.810021990391163e-06,
+      "loss": 0.5339,
+      "step": 140
+    },
+    {
+      "epoch": 1.1473396998635743,
+      "grad_norm": 0.6775122343869695,
+      "learning_rate": 7.77040331201572e-06,
+      "loss": 0.544,
+      "step": 141
+    },
+    {
+      "epoch": 1.155525238744884,
+      "grad_norm": 0.6977446290741449,
+      "learning_rate": 7.73053202532219e-06,
+      "loss": 0.5176,
+      "step": 142
+    },
+    {
+      "epoch": 1.1637107776261937,
+      "grad_norm": 0.7195762022745771,
+      "learning_rate": 7.690411765816864e-06,
+      "loss": 0.5199,
+      "step": 143
+    },
+    {
+      "epoch": 1.1718963165075034,
+      "grad_norm": 0.7052036741882443,
+      "learning_rate": 7.650046191707641e-06,
+      "loss": 0.52,
+      "step": 144
+    },
+    {
+      "epoch": 1.180081855388813,
+      "grad_norm": 0.6993817480227583,
+      "learning_rate": 7.609438983570461e-06,
+      "loss": 0.5398,
+      "step": 145
+    },
+    {
+      "epoch": 1.1882673942701227,
+      "grad_norm": 0.6693930742306715,
+      "learning_rate": 7.5685938440137185e-06,
+      "loss": 0.5352,
+      "step": 146
+    },
+    {
+      "epoch": 1.1964529331514324,
+      "grad_norm": 0.6891988814575712,
+      "learning_rate": 7.527514497340642e-06,
+      "loss": 0.5001,
+      "step": 147
+    },
+    {
+      "epoch": 1.204638472032742,
+      "grad_norm": 0.6792075266848643,
+      "learning_rate": 7.486204689209719e-06,
+      "loss": 0.5085,
+      "step": 148
+    },
+    {
+      "epoch": 1.2128240109140518,
+      "grad_norm": 0.7068616251465075,
+      "learning_rate": 7.444668186293153e-06,
+      "loss": 0.5325,
+      "step": 149
+    },
+    {
+      "epoch": 1.2210095497953615,
+      "grad_norm": 0.6795815065929743,
+      "learning_rate": 7.402908775933419e-06,
+      "loss": 0.5429,
+      "step": 150
+    },
+    {
+      "epoch": 1.2291950886766712,
+      "grad_norm": 0.6812900626623999,
+      "learning_rate": 7.360930265797934e-06,
+      "loss": 0.5141,
+      "step": 151
+    },
+    {
+      "epoch": 1.2373806275579808,
+      "grad_norm": 0.6922781081584425,
+      "learning_rate": 7.318736483531861e-06,
+      "loss": 0.5039,
+      "step": 152
+    },
+    {
+      "epoch": 1.2455661664392905,
+      "grad_norm": 0.6971735097481496,
+      "learning_rate": 7.2763312764091055e-06,
+      "loss": 0.515,
+      "step": 153
+    },
+    {
+      "epoch": 1.2537517053206002,
+      "grad_norm": 0.7190442324772383,
+      "learning_rate": 7.23371851098152e-06,
+      "loss": 0.5347,
+      "step": 154
+    },
+    {
+      "epoch": 1.26193724420191,
+      "grad_norm": 0.6798326549238114,
+      "learning_rate": 7.190902072726336e-06,
+      "loss": 0.5072,
+      "step": 155
+    },
+    {
+      "epoch": 1.2701227830832196,
+      "grad_norm": 0.6373536362037905,
+      "learning_rate": 7.147885865691899e-06,
+      "loss": 0.5058,
+      "step": 156
+    },
+    {
+      "epoch": 1.2783083219645293,
+      "grad_norm": 0.7409453004444784,
+      "learning_rate": 7.104673812141676e-06,
+      "loss": 0.5326,
+      "step": 157
+    },
+    {
+      "epoch": 1.286493860845839,
+      "grad_norm": 0.708927994034785,
+      "learning_rate": 7.061269852196633e-06,
+      "loss": 0.51,
+      "step": 158
+    },
+    {
+      "epoch": 1.2946793997271486,
+      "grad_norm": 0.7378291795102794,
+      "learning_rate": 7.017677943475962e-06,
+      "loss": 0.5506,
+      "step": 159
+    },
+    {
+      "epoch": 1.3028649386084583,
+      "grad_norm": 0.6684781087704905,
+      "learning_rate": 6.973902060736226e-06,
+      "loss": 0.5408,
+      "step": 160
+    },
+    {
+      "epoch": 1.311050477489768,
+      "grad_norm": 0.7303626826476006,
+      "learning_rate": 6.929946195508933e-06,
+      "loss": 0.5059,
+      "step": 161
+    },
+    {
+      "epoch": 1.3192360163710777,
+      "grad_norm": 0.7233810086483591,
+      "learning_rate": 6.8858143557365865e-06,
+      "loss": 0.5266,
+      "step": 162
+    },
+    {
+      "epoch": 1.3274215552523874,
+      "grad_norm": 0.6517028104171342,
+      "learning_rate": 6.841510565407235e-06,
+      "loss": 0.5511,
+      "step": 163
+    },
+    {
+      "epoch": 1.3356070941336973,
+      "grad_norm": 0.6797794166831331,
+      "learning_rate": 6.797038864187564e-06,
+      "loss": 0.5514,
+      "step": 164
+    },
+    {
+      "epoch": 1.3437926330150067,
+      "grad_norm": 0.6544424076735298,
+      "learning_rate": 6.752403307054549e-06,
+      "loss": 0.5265,
+      "step": 165
+    },
+    {
+      "epoch": 1.3519781718963166,
+      "grad_norm": 0.7213289608772747,
+      "learning_rate": 6.707607963925725e-06,
+      "loss": 0.5016,
+      "step": 166
+    },
+    {
+      "epoch": 1.360163710777626,
+      "grad_norm": 0.6712362344838894,
+      "learning_rate": 6.66265691928808e-06,
+      "loss": 0.5307,
+      "step": 167
+    },
+    {
+      "epoch": 1.368349249658936,
+      "grad_norm": 0.6774741054792113,
+      "learning_rate": 6.617554271825636e-06,
+      "loss": 0.4987,
+      "step": 168
+    },
+    {
+      "epoch": 1.3765347885402455,
+      "grad_norm": 0.6953202992195601,
+      "learning_rate": 6.5723041340457175e-06,
+      "loss": 0.5601,
+      "step": 169
+    },
+    {
+      "epoch": 1.3847203274215554,
+      "grad_norm": 0.6514741526426598,
+      "learning_rate": 6.526910631903973e-06,
+      "loss": 0.518,
+      "step": 170
+    },
+    {
+      "epoch": 1.3929058663028648,
+      "grad_norm": 0.6911638142925786,
+      "learning_rate": 6.481377904428171e-06,
+      "loss": 0.544,
+      "step": 171
+    },
+    {
+      "epoch": 1.4010914051841747,
+      "grad_norm": 0.6707693839058396,
+      "learning_rate": 6.435710103340787e-06,
+      "loss": 0.53,
+      "step": 172
+    },
+    {
+      "epoch": 1.4092769440654842,
+      "grad_norm": 0.6969660470990657,
+      "learning_rate": 6.3899113926804565e-06,
+      "loss": 0.529,
+      "step": 173
+    },
+    {
+      "epoch": 1.4174624829467941,
+      "grad_norm": 0.6524912464535223,
+      "learning_rate": 6.3439859484222874e-06,
+      "loss": 0.5019,
+      "step": 174
+    },
+    {
+      "epoch": 1.4256480218281036,
+      "grad_norm": 0.6931835129968518,
+      "learning_rate": 6.297937958097094e-06,
+      "loss": 0.5374,
+      "step": 175
+    },
+    {
+      "epoch": 1.4338335607094135,
+      "grad_norm": 0.6584397219393618,
+      "learning_rate": 6.251771620409563e-06,
+      "loss": 0.491,
+      "step": 176
+    },
+    {
+      "epoch": 1.442019099590723,
+      "grad_norm": 0.6896633651029679,
+      "learning_rate": 6.205491144855432e-06,
+      "loss": 0.5166,
+      "step": 177
+    },
+    {
+      "epoch": 1.4502046384720328,
+      "grad_norm": 0.6839099660636084,
+      "learning_rate": 6.1591007513376425e-06,
+      "loss": 0.5387,
+      "step": 178
+    },
+    {
+      "epoch": 1.4583901773533423,
+      "grad_norm": 0.7005580566040835,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.5109,
+      "step": 179
+    },
+    {
+      "epoch": 1.4665757162346522,
+      "grad_norm": 0.6921812967470553,
+      "learning_rate": 6.066007139749351e-06,
+      "loss": 0.5065,
+      "step": 180
+    },
+    {
+      "epoch": 1.4747612551159617,
+      "grad_norm": 0.6688682619844354,
+      "learning_rate": 6.019312410053286e-06,
+      "loss": 0.5353,
+      "step": 181
+    },
+    {
+      "epoch": 1.4829467939972716,
+      "grad_norm": 0.7624212958880006,
+      "learning_rate": 5.972524738368452e-06,
+      "loss": 0.5339,
+      "step": 182
+    },
+    {
+      "epoch": 1.491132332878581,
+      "grad_norm": 0.6425569450279436,
+      "learning_rate": 5.925648390844476e-06,
+      "loss": 0.5241,
+      "step": 183
+    },
+    {
+      "epoch": 1.499317871759891,
+      "grad_norm": 0.7031579623697819,
+      "learning_rate": 5.878687641716539e-06,
+      "loss": 0.5416,
+      "step": 184
+    },
+    {
+      "epoch": 1.5075034106412004,
+      "grad_norm": 0.671036143518173,
+      "learning_rate": 5.831646772915651e-06,
+      "loss": 0.5304,
+      "step": 185
+    },
+    {
+      "epoch": 1.5156889495225103,
+      "grad_norm": 0.6965643037878884,
+      "learning_rate": 5.7845300736782205e-06,
+      "loss": 0.5603,
+      "step": 186
+    },
+    {
+      "epoch": 1.5238744884038198,
+      "grad_norm": 0.9636732668433035,
+      "learning_rate": 5.7373418401549565e-06,
+      "loss": 0.5128,
+      "step": 187
+    },
+    {
+      "epoch": 1.5320600272851297,
+      "grad_norm": 0.6809777005177087,
+      "learning_rate": 5.690086375019135e-06,
+      "loss": 0.5662,
+      "step": 188
+    },
+    {
+      "epoch": 1.5402455661664392,
+      "grad_norm": 0.6995628233043069,
+      "learning_rate": 5.642767987074288e-06,
+      "loss": 0.5132,
+      "step": 189
+    },
+    {
+      "epoch": 1.548431105047749,
+      "grad_norm": 0.6397896952913672,
+      "learning_rate": 5.595390990861311e-06,
+      "loss": 0.5266,
+      "step": 190
+    },
+    {
+      "epoch": 1.5566166439290585,
+      "grad_norm": 0.7260279515386082,
+      "learning_rate": 5.547959706265068e-06,
+      "loss": 0.5122,
+      "step": 191
+    },
+    {
+      "epoch": 1.5648021828103684,
+      "grad_norm": 0.6671500094333921,
+      "learning_rate": 5.500478458120493e-06,
+      "loss": 0.523,
+      "step": 192
+    },
+    {
+      "epoch": 1.5729877216916779,
+      "grad_norm": 0.6279438494285721,
+      "learning_rate": 5.45295157581825e-06,
+      "loss": 0.4994,
+      "step": 193
+    },
+    {
+      "epoch": 1.5811732605729878,
+      "grad_norm": 0.6666263298068601,
+      "learning_rate": 5.405383392909973e-06,
+      "loss": 0.5345,
+      "step": 194
+    },
+    {
+      "epoch": 1.5893587994542973,
+      "grad_norm": 0.6615143998334845,
+      "learning_rate": 5.357778246713131e-06,
+      "loss": 0.4999,
+      "step": 195
+    },
+    {
+      "epoch": 1.5975443383356072,
+      "grad_norm": 0.6968103870177652,
+      "learning_rate": 5.310140477915544e-06,
+      "loss": 0.5315,
+      "step": 196
+    },
+    {
+      "epoch": 1.6057298772169166,
+      "grad_norm": 0.673129914192788,
+      "learning_rate": 5.262474430179597e-06,
+      "loss": 0.5103,
+      "step": 197
+    },
+    {
+      "epoch": 1.6139154160982265,
+      "grad_norm": 0.6980507997253446,
+      "learning_rate": 5.2147844497461745e-06,
+      "loss": 0.5235,
+      "step": 198
+    },
+    {
+      "epoch": 1.622100954979536,
+      "grad_norm": 0.7296195245827997,
+      "learning_rate": 5.1670748850383734e-06,
+      "loss": 0.5306,
+      "step": 199
+    },
+    {
+      "epoch": 1.630286493860846,
+      "grad_norm": 0.6909338740528714,
+      "learning_rate": 5.1193500862650045e-06,
+      "loss": 0.5312,
+      "step": 200
+    },
+    {
+      "epoch": 1.6384720327421554,
+      "grad_norm": 0.7196816573312542,
+      "learning_rate": 5.071614405023938e-06,
+      "loss": 0.4967,
+      "step": 201
+    },
+    {
+      "epoch": 1.6466575716234653,
+      "grad_norm": 0.7000487180066235,
+      "learning_rate": 5.023872193905316e-06,
+      "loss": 0.522,
+      "step": 202
+    },
+    {
+      "epoch": 1.654843110504775,
+      "grad_norm": 0.6846805681540516,
+      "learning_rate": 4.976127806094685e-06,
+      "loss": 0.5001,
+      "step": 203
+    },
+    {
+      "epoch": 1.6630286493860846,
+      "grad_norm": 0.6634128402084197,
+      "learning_rate": 4.928385594976063e-06,
+      "loss": 0.4972,
+      "step": 204
+    },
+    {
+      "epoch": 1.6712141882673943,
+      "grad_norm": 0.6641383584022369,
+      "learning_rate": 4.880649913734996e-06,
+      "loss": 0.5417,
+      "step": 205
+    },
+    {
+      "epoch": 1.679399727148704,
+      "grad_norm": 0.686001010727504,
+      "learning_rate": 4.832925114961629e-06,
+      "loss": 0.5443,
+      "step": 206
+    },
+    {
+      "epoch": 1.6875852660300137,
+      "grad_norm": 0.6467569918335102,
+      "learning_rate": 4.785215550253826e-06,
+      "loss": 0.4966,
+      "step": 207
+    },
+    {
+      "epoch": 1.6957708049113234,
+      "grad_norm": 0.6736896402505433,
+      "learning_rate": 4.737525569820405e-06,
+      "loss": 0.5149,
+      "step": 208
+    },
+    {
+      "epoch": 1.703956343792633,
+      "grad_norm": 0.6547008477244521,
+      "learning_rate": 4.689859522084457e-06,
+      "loss": 0.5185,
+      "step": 209
+    },
+    {
+      "epoch": 1.7121418826739427,
+      "grad_norm": 0.6779152973329728,
+      "learning_rate": 4.64222175328687e-06,
+      "loss": 0.5112,
+      "step": 210
+    },
+    {
+      "epoch": 1.7203274215552524,
+      "grad_norm": 0.6881760025662235,
+      "learning_rate": 4.594616607090028e-06,
+      "loss": 0.5319,
+      "step": 211
+    },
+    {
+      "epoch": 1.728512960436562,
+      "grad_norm": 0.6934876690870692,
+      "learning_rate": 4.547048424181751e-06,
+      "loss": 0.5036,
+      "step": 212
+    },
+    {
+      "epoch": 1.7366984993178718,
+      "grad_norm": 0.6564578983689818,
+      "learning_rate": 4.499521541879508e-06,
+      "loss": 0.5342,
+      "step": 213
+    },
+    {
+      "epoch": 1.7448840381991815,
+      "grad_norm": 0.6705466743944972,
+      "learning_rate": 4.452040293734934e-06,
+      "loss": 0.5208,
+      "step": 214
+    },
+    {
+      "epoch": 1.7530695770804912,
+      "grad_norm": 0.6505369843152567,
+      "learning_rate": 4.40460900913869e-06,
+      "loss": 0.5416,
+      "step": 215
+    },
+    {
+      "epoch": 1.7612551159618008,
+      "grad_norm": 0.6990568514053161,
+      "learning_rate": 4.357232012925714e-06,
+      "loss": 0.5231,
+      "step": 216
+    },
+    {
+      "epoch": 1.7694406548431105,
+      "grad_norm": 0.682171935793151,
+      "learning_rate": 4.309913624980866e-06,
+      "loss": 0.5131,
+      "step": 217
+    },
+    {
+      "epoch": 1.7776261937244202,
+      "grad_norm": 0.6888009240380015,
+      "learning_rate": 4.262658159845046e-06,
+      "loss": 0.5562,
+      "step": 218
+    },
+    {
+      "epoch": 1.78581173260573,
+      "grad_norm": 0.6631729470775671,
+      "learning_rate": 4.2154699263217794e-06,
+      "loss": 0.5388,
+      "step": 219
+    },
+    {
+      "epoch": 1.7939972714870396,
+      "grad_norm": 0.6913357855941524,
+      "learning_rate": 4.1683532270843505e-06,
+      "loss": 0.5257,
+      "step": 220
+    },
+    {
+      "epoch": 1.8021828103683493,
+      "grad_norm": 0.7600974747556154,
+      "learning_rate": 4.121312358283464e-06,
+      "loss": 0.5594,
+      "step": 221
+    },
+    {
+      "epoch": 1.810368349249659,
+      "grad_norm": 0.7405495207001502,
+      "learning_rate": 4.074351609155527e-06,
+      "loss": 0.5265,
+      "step": 222
+    },
+    {
+      "epoch": 1.8185538881309686,
+      "grad_norm": 0.6737231563100519,
+      "learning_rate": 4.0274752616315485e-06,
+      "loss": 0.5119,
+      "step": 223
+    },
+    {
+      "epoch": 1.8267394270122783,
+      "grad_norm": 0.6701649580766764,
+      "learning_rate": 3.980687589946715e-06,
+      "loss": 0.5466,
+      "step": 224
+    },
+    {
+      "epoch": 1.834924965893588,
+      "grad_norm": 0.6725451869274366,
+      "learning_rate": 3.9339928602506505e-06,
+      "loss": 0.5302,
+      "step": 225
+    },
+    {
+      "epoch": 1.8431105047748977,
+      "grad_norm": 0.6331432125437578,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.4998,
+      "step": 226
+    },
+    {
+      "epoch": 1.8512960436562074,
+      "grad_norm": 0.6873175907256674,
+      "learning_rate": 3.840899248662358e-06,
+      "loss": 0.546,
+      "step": 227
+    },
+    {
+      "epoch": 1.859481582537517,
+      "grad_norm": 0.6637408125371841,
+      "learning_rate": 3.7945088551445698e-06,
+      "loss": 0.5404,
+      "step": 228
+    },
+    {
+      "epoch": 1.8676671214188267,
+      "grad_norm": 0.6559384775809904,
+      "learning_rate": 3.748228379590438e-06,
+      "loss": 0.5278,
+      "step": 229
+    },
+    {
+      "epoch": 1.8758526603001364,
+      "grad_norm": 0.6684145829249631,
+      "learning_rate": 3.7020620419029095e-06,
+      "loss": 0.5084,
+      "step": 230
+    },
+    {
+      "epoch": 1.884038199181446,
+      "grad_norm": 0.6890056694528337,
+      "learning_rate": 3.656014051577713e-06,
+      "loss": 0.4891,
+      "step": 231
+    },
+    {
+      "epoch": 1.8922237380627558,
+      "grad_norm": 0.6904343790721232,
+      "learning_rate": 3.610088607319544e-06,
+      "loss": 0.5311,
+      "step": 232
+    },
+    {
+      "epoch": 1.9004092769440655,
+      "grad_norm": 0.7268616356810529,
+      "learning_rate": 3.5642898966592145e-06,
+      "loss": 0.5268,
+      "step": 233
+    },
+    {
+      "epoch": 1.9085948158253752,
+      "grad_norm": 0.6610468431905363,
+      "learning_rate": 3.518622095571831e-06,
+      "loss": 0.5163,
+      "step": 234
+    },
+    {
+      "epoch": 1.9167803547066848,
+      "grad_norm": 0.7007954364884253,
+      "learning_rate": 3.4730893680960267e-06,
+      "loss": 0.5166,
+      "step": 235
+    },
+    {
+      "epoch": 1.9249658935879945,
+      "grad_norm": 0.7565901496990775,
+      "learning_rate": 3.4276958659542838e-06,
+      "loss": 0.5041,
+      "step": 236
+    },
+    {
+      "epoch": 1.9331514324693042,
+      "grad_norm": 0.6964549914506967,
+      "learning_rate": 3.382445728174365e-06,
+      "loss": 0.5368,
+      "step": 237
+    },
+    {
+      "epoch": 1.941336971350614,
+      "grad_norm": 0.700682306258137,
+      "learning_rate": 3.3373430807119212e-06,
+      "loss": 0.525,
+      "step": 238
+    },
+    {
+      "epoch": 1.9495225102319236,
+      "grad_norm": 0.6866654675874281,
+      "learning_rate": 3.292392036074277e-06,
+      "loss": 0.5603,
+      "step": 239
+    },
+    {
+      "epoch": 1.9577080491132333,
+      "grad_norm": 0.6728811050097208,
+      "learning_rate": 3.2475966929454505e-06,
+      "loss": 0.5259,
+      "step": 240
+    },
+    {
+      "epoch": 1.965893587994543,
+      "grad_norm": 0.6591241401446245,
+      "learning_rate": 3.202961135812437e-06,
+      "loss": 0.5047,
+      "step": 241
+    },
+    {
+      "epoch": 1.9740791268758526,
+      "grad_norm": 0.6374754047711063,
+      "learning_rate": 3.1584894345927663e-06,
+      "loss": 0.4998,
+      "step": 242
+    },
+    {
+      "epoch": 1.9822646657571623,
+      "grad_norm": 0.6631893411714516,
+      "learning_rate": 3.114185644263415e-06,
+      "loss": 0.5356,
+      "step": 243
+    },
+    {
+      "epoch": 1.990450204638472,
+      "grad_norm": 0.6924356519329616,
+      "learning_rate": 3.0700538044910684e-06,
+      "loss": 0.5246,
+      "step": 244
+    },
+    {
+      "epoch": 1.9986357435197817,
+      "grad_norm": 0.864095503348459,
+      "learning_rate": 3.0260979392637753e-06,
+      "loss": 0.5096,
+      "step": 245
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.864095503348459,
+      "learning_rate": 2.9823220565240396e-06,
+      "loss": 0.4995,
+      "step": 246
+    },
+    {
+      "epoch": 2.00818553888131,
+      "grad_norm": 1.8027596551233536,
+      "learning_rate": 2.9387301478033694e-06,
+      "loss": 0.4139,
+      "step": 247
+    },
+    {
+      "epoch": 2.0163710777626194,
+      "grad_norm": 0.7619387743115077,
+      "learning_rate": 2.8953261878583263e-06,
+      "loss": 0.4119,
+      "step": 248
+    },
+    {
+      "epoch": 2.0245566166439293,
+      "grad_norm": 0.792250852956364,
+      "learning_rate": 2.852114134308104e-06,
+      "loss": 0.4299,
+      "step": 249
+    },
+    {
+      "epoch": 2.0327421555252387,
+      "grad_norm": 0.8331234294016224,
+      "learning_rate": 2.8090979272736663e-06,
+      "loss": 0.4173,
+      "step": 250
+    },
+    {
+      "epoch": 2.0409276944065486,
+      "grad_norm": 0.7837271669310566,
+      "learning_rate": 2.766281489018482e-06,
+      "loss": 0.4125,
+      "step": 251
+    },
+    {
+      "epoch": 2.049113233287858,
+      "grad_norm": 0.732909381181695,
+      "learning_rate": 2.7236687235908953e-06,
+      "loss": 0.4269,
+      "step": 252
+    },
+    {
+      "epoch": 2.057298772169168,
+      "grad_norm": 0.7716051532871122,
+      "learning_rate": 2.681263516468139e-06,
+      "loss": 0.4238,
+      "step": 253
+    },
+    {
+      "epoch": 2.0654843110504775,
+      "grad_norm": 0.7542777472731785,
+      "learning_rate": 2.6390697342020665e-06,
+      "loss": 0.4184,
+      "step": 254
+    },
+    {
+      "epoch": 2.0736698499317874,
+      "grad_norm": 0.8158429899004254,
+      "learning_rate": 2.5970912240665815e-06,
+      "loss": 0.4346,
+      "step": 255
+    },
+    {
+      "epoch": 2.081855388813097,
+      "grad_norm": 0.9160533796282962,
+      "learning_rate": 2.5553318137068473e-06,
+      "loss": 0.4048,
+      "step": 256
+    },
+    {
+      "epoch": 2.0900409276944067,
+      "grad_norm": 0.8800259349255135,
+      "learning_rate": 2.5137953107902814e-06,
+      "loss": 0.4099,
+      "step": 257
+    },
+    {
+      "epoch": 2.098226466575716,
+      "grad_norm": 0.81536917483078,
+      "learning_rate": 2.472485502659358e-06,
+      "loss": 0.4106,
+      "step": 258
+    },
+    {
+      "epoch": 2.106412005457026,
+      "grad_norm": 0.7818869480203516,
+      "learning_rate": 2.4314061559862836e-06,
+      "loss": 0.3983,
+      "step": 259
+    },
+    {
+      "epoch": 2.1145975443383356,
+      "grad_norm": 0.7435500830249482,
+      "learning_rate": 2.3905610164295394e-06,
+      "loss": 0.4007,
+      "step": 260
+    },
+    {
+      "epoch": 2.1227830832196455,
+      "grad_norm": 0.7365477010338396,
+      "learning_rate": 2.3499538082923607e-06,
+      "loss": 0.4247,
+      "step": 261
+    },
+    {
+      "epoch": 2.130968622100955,
+      "grad_norm": 0.7734902356583239,
+      "learning_rate": 2.309588234183137e-06,
+      "loss": 0.4117,
+      "step": 262
+    },
+    {
+      "epoch": 2.139154160982265,
+      "grad_norm": 0.7939085324855263,
+      "learning_rate": 2.2694679746778116e-06,
+      "loss": 0.3701,
+      "step": 263
+    },
+    {
+      "epoch": 2.1473396998635743,
+      "grad_norm": 0.7460548804297111,
+      "learning_rate": 2.22959668798428e-06,
+      "loss": 0.3924,
+      "step": 264
+    },
+    {
+      "epoch": 2.155525238744884,
+      "grad_norm": 0.8499262881402114,
+      "learning_rate": 2.1899780096088375e-06,
+      "loss": 0.4061,
+      "step": 265
+    },
+    {
+      "epoch": 2.1637107776261937,
+      "grad_norm": 0.7609355875240096,
+      "learning_rate": 2.1506155520246795e-06,
+      "loss": 0.4094,
+      "step": 266
+    },
+    {
+      "epoch": 2.1718963165075036,
+      "grad_norm": 0.7824457857456796,
+      "learning_rate": 2.1115129043425188e-06,
+      "loss": 0.4186,
+      "step": 267
+    },
+    {
+      "epoch": 2.180081855388813,
+      "grad_norm": 0.7912164425233928,
+      "learning_rate": 2.072673631983323e-06,
+      "loss": 0.3787,
+      "step": 268
+    },
+    {
+      "epoch": 2.188267394270123,
+      "grad_norm": 0.8076115925038775,
+      "learning_rate": 2.0341012763532243e-06,
+      "loss": 0.3779,
+      "step": 269
+    },
+    {
+      "epoch": 2.1964529331514324,
+      "grad_norm": 0.7792124545324812,
+      "learning_rate": 1.995799354520598e-06,
+      "loss": 0.3973,
+      "step": 270
+    },
+    {
+      "epoch": 2.2046384720327423,
+      "grad_norm": 0.7556374773757739,
+      "learning_rate": 1.9577713588953797e-06,
+      "loss": 0.3861,
+      "step": 271
+    },
+    {
+      "epoch": 2.212824010914052,
+      "grad_norm": 0.715008006933338,
+      "learning_rate": 1.9200207569106216e-06,
+      "loss": 0.3936,
+      "step": 272
+    },
+    {
+      "epoch": 2.2210095497953617,
+      "grad_norm": 0.7807990202837165,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.4023,
+      "step": 273
+    },
+    {
+      "epoch": 2.229195088676671,
+      "grad_norm": 0.7587528540992885,
+      "learning_rate": 1.8453654768156138e-06,
+      "loss": 0.4275,
+      "step": 274
+    },
+    {
+      "epoch": 2.237380627557981,
+      "grad_norm": 0.8220738353200976,
+      "learning_rate": 1.8084676058531376e-06,
+      "loss": 0.3933,
+      "step": 275
+    },
+    {
+      "epoch": 2.2455661664392905,
+      "grad_norm": 0.7901348024570493,
+      "learning_rate": 1.771860742205988e-06,
+      "loss": 0.4317,
+      "step": 276
+    },
+    {
+      "epoch": 2.2537517053206004,
+      "grad_norm": 0.8140221241566415,
+      "learning_rate": 1.7355482237268983e-06,
+      "loss": 0.4366,
+      "step": 277
+    },
+    {
+      "epoch": 2.26193724420191,
+      "grad_norm": 0.7792400569651549,
+      "learning_rate": 1.6995333614298908e-06,
+      "loss": 0.4082,
+      "step": 278
+    },
+    {
+      "epoch": 2.27012278308322,
+      "grad_norm": 0.7916799443290311,
+      "learning_rate": 1.6638194391883822e-06,
+      "loss": 0.3857,
+      "step": 279
+    },
+    {
+      "epoch": 2.2783083219645293,
+      "grad_norm": 0.7731098725686014,
+      "learning_rate": 1.6284097134357535e-06,
+      "loss": 0.3958,
+      "step": 280
+    },
+    {
+      "epoch": 2.286493860845839,
+      "grad_norm": 0.732230108667779,
+      "learning_rate": 1.5933074128684333e-06,
+      "loss": 0.4102,
+      "step": 281
+    },
+    {
+      "epoch": 2.2946793997271486,
+      "grad_norm": 0.7884350213329203,
+      "learning_rate": 1.5585157381514875e-06,
+      "loss": 0.4152,
+      "step": 282
+    },
+    {
+      "epoch": 2.3028649386084585,
+      "grad_norm": 0.7700379905456493,
+      "learning_rate": 1.5240378616267887e-06,
+      "loss": 0.4174,
+      "step": 283
+    },
+    {
+      "epoch": 2.311050477489768,
+      "grad_norm": 0.7957616626164629,
+      "learning_rate": 1.4898769270237611e-06,
+      "loss": 0.3886,
+      "step": 284
+    },
+    {
+      "epoch": 2.319236016371078,
+      "grad_norm": 0.7893635109056786,
+      "learning_rate": 1.4560360491727233e-06,
+      "loss": 0.3913,
+      "step": 285
+    },
+    {
+      "epoch": 2.3274215552523874,
+      "grad_norm": 0.7902911589168176,
+      "learning_rate": 1.4225183137208775e-06,
+      "loss": 0.4141,
+      "step": 286
+    },
+    {
+      "epoch": 2.3356070941336973,
+      "grad_norm": 0.7817906842176934,
+      "learning_rate": 1.389326776850966e-06,
+      "loss": 0.4221,
+      "step": 287
+    },
+    {
+      "epoch": 2.3437926330150067,
+      "grad_norm": 0.7462547571612814,
+      "learning_rate": 1.3564644650025894e-06,
+      "loss": 0.4116,
+      "step": 288
+    },
+    {
+      "epoch": 2.3519781718963166,
+      "grad_norm": 0.7831968047397042,
+      "learning_rate": 1.323934374596268e-06,
+      "loss": 0.3932,
+      "step": 289
+    },
+    {
+      "epoch": 2.360163710777626,
+      "grad_norm": 0.7221724236305611,
+      "learning_rate": 1.2917394717602123e-06,
+      "loss": 0.3853,
+      "step": 290
+    },
+    {
+      "epoch": 2.368349249658936,
+      "grad_norm": 0.7461225498645182,
+      "learning_rate": 1.2598826920598773e-06,
+      "loss": 0.3862,
+      "step": 291
+    },
+    {
+      "epoch": 2.3765347885402455,
+      "grad_norm": 1.046414381109276,
+      "learning_rate": 1.2283669402302878e-06,
+      "loss": 0.4309,
+      "step": 292
+    },
+    {
+      "epoch": 2.3847203274215554,
+      "grad_norm": 0.9489345130797908,
+      "learning_rate": 1.197195089911191e-06,
+      "loss": 0.3929,
+      "step": 293
+    },
+    {
+      "epoch": 2.392905866302865,
+      "grad_norm": 0.7903351074244309,
+      "learning_rate": 1.166369983385024e-06,
+      "loss": 0.3881,
+      "step": 294
+    },
+    {
+      "epoch": 2.4010914051841747,
+      "grad_norm": 0.7541482648516513,
+      "learning_rate": 1.1358944313177566e-06,
+      "loss": 0.419,
+      "step": 295
+    },
+    {
+      "epoch": 2.409276944065484,
+      "grad_norm": 0.7784727134402827,
+      "learning_rate": 1.1057712125026116e-06,
+      "loss": 0.4125,
+      "step": 296
+    },
+    {
+      "epoch": 2.417462482946794,
+      "grad_norm": 0.7831320352654086,
+      "learning_rate": 1.0760030736066952e-06,
+      "loss": 0.4295,
+      "step": 297
+    },
+    {
+      "epoch": 2.4256480218281036,
+      "grad_norm": 0.7242161841038534,
+      "learning_rate": 1.0465927289205452e-06,
+      "loss": 0.3715,
+      "step": 298
+    },
+    {
+      "epoch": 2.4338335607094135,
+      "grad_norm": 0.7538861750254986,
+      "learning_rate": 1.0175428601106441e-06,
+      "loss": 0.4174,
+      "step": 299
+    },
+    {
+      "epoch": 2.442019099590723,
+      "grad_norm": 0.7533427868456628,
+      "learning_rate": 9.888561159748995e-07,
+      "loss": 0.3972,
+      "step": 300
+    },
+    {
+      "epoch": 2.450204638472033,
+      "grad_norm": 0.7475572855082404,
+      "learning_rate": 9.605351122011308e-07,
+      "loss": 0.4157,
+      "step": 301
+    },
+    {
+      "epoch": 2.4583901773533423,
+      "grad_norm": 0.8126758147330377,
+      "learning_rate": 9.325824311285564e-07,
+      "loss": 0.3962,
+      "step": 302
+    },
+    {
+      "epoch": 2.466575716234652,
+      "grad_norm": 0.7585622565258165,
+      "learning_rate": 9.050006215123419e-07,
+      "loss": 0.4263,
+      "step": 303
+    },
+    {
+      "epoch": 2.4747612551159617,
+      "grad_norm": 0.7599274947242529,
+      "learning_rate": 8.777921982911996e-07,
+      "loss": 0.4052,
+      "step": 304
+    },
+    {
+      "epoch": 2.4829467939972716,
+      "grad_norm": 0.7542201080325229,
+      "learning_rate": 8.509596423580712e-07,
+      "loss": 0.4034,
+      "step": 305
+    },
+    {
+      "epoch": 2.491132332878581,
+      "grad_norm": 0.7273725054331669,
+      "learning_rate": 8.245054003339247e-07,
+      "loss": 0.3882,
+      "step": 306
+    },
+    {
+      "epoch": 2.499317871759891,
+      "grad_norm": 0.7539559642233192,
+      "learning_rate": 7.984318843446593e-07,
+      "loss": 0.4005,
+      "step": 307
+    },
+    {
+      "epoch": 2.5075034106412004,
+      "grad_norm": 0.8036228808002506,
+      "learning_rate": 7.727414718011706e-07,
+      "loss": 0.4003,
+      "step": 308
+    },
+    {
+      "epoch": 2.5156889495225103,
+      "grad_norm": 0.7421038405236332,
+      "learning_rate": 7.474365051825749e-07,
+      "loss": 0.4158,
+      "step": 309
+    },
+    {
+      "epoch": 2.52387448840382,
+      "grad_norm": 0.7808353000074717,
+      "learning_rate": 7.225192918226215e-07,
+      "loss": 0.3879,
+      "step": 310
+    },
+    {
+      "epoch": 2.5320600272851297,
+      "grad_norm": 0.7606258935305734,
+      "learning_rate": 6.979921036993042e-07,
+      "loss": 0.4158,
+      "step": 311
+    },
+    {
+      "epoch": 2.540245566166439,
+      "grad_norm": 0.7045765025498434,
+      "learning_rate": 6.738571772276997e-07,
+      "loss": 0.415,
+      "step": 312
+    },
+    {
+      "epoch": 2.548431105047749,
+      "grad_norm": 0.7444529768374089,
+      "learning_rate": 6.501167130560515e-07,
+      "loss": 0.3929,
+      "step": 313
+    },
+    {
+      "epoch": 2.5566166439290585,
+      "grad_norm": 0.7584143919424828,
+      "learning_rate": 6.267728758651131e-07,
+      "loss": 0.4123,
+      "step": 314
+    },
+    {
+      "epoch": 2.5648021828103684,
+      "grad_norm": 0.7479945247594694,
+      "learning_rate": 6.038277941707671e-07,
+      "loss": 0.3821,
+      "step": 315
+    },
+    {
+      "epoch": 2.572987721691678,
+      "grad_norm": 0.7561067208148362,
+      "learning_rate": 5.812835601299438e-07,
+      "loss": 0.4023,
+      "step": 316
+    },
+    {
+      "epoch": 2.581173260572988,
+      "grad_norm": 0.7751562482156354,
+      "learning_rate": 5.591422293498633e-07,
+      "loss": 0.4124,
+      "step": 317
+    },
+    {
+      "epoch": 2.5893587994542973,
+      "grad_norm": 0.7992598569451743,
+      "learning_rate": 5.374058207005945e-07,
+      "loss": 0.3929,
+      "step": 318
+    },
+    {
+      "epoch": 2.597544338335607,
+      "grad_norm": 0.750932391285096,
+      "learning_rate": 5.160763161309768e-07,
+      "loss": 0.3859,
+      "step": 319
+    },
+    {
+      "epoch": 2.6057298772169166,
+      "grad_norm": 0.7772903134930654,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.4166,
+      "step": 320
+    },
+    {
+      "epoch": 2.6139154160982265,
+      "grad_norm": 0.750452009708911,
+      "learning_rate": 4.7464576133899043e-07,
+      "loss": 0.3857,
+      "step": 321
+    },
+    {
+      "epoch": 2.622100954979536,
+      "grad_norm": 0.7372642532039464,
+      "learning_rate": 4.545484887986368e-07,
+      "loss": 0.4134,
+      "step": 322
+    },
+    {
+      "epoch": 2.630286493860846,
+      "grad_norm": 0.8460806837403229,
+      "learning_rate": 4.348656753575092e-07,
+      "loss": 0.4038,
+      "step": 323
+    },
+    {
+      "epoch": 2.6384720327421554,
+      "grad_norm": 0.7694305883370988,
+      "learning_rate": 4.1559911571545544e-07,
+      "loss": 0.4126,
+      "step": 324
+    },
+    {
+      "epoch": 2.6466575716234653,
+      "grad_norm": 0.7956542721679164,
+      "learning_rate": 3.9675056661785563e-07,
+      "loss": 0.4049,
+      "step": 325
+    },
+    {
+      "epoch": 2.6548431105047747,
+      "grad_norm": 0.7703211713211958,
+      "learning_rate": 3.783217466954503e-07,
+      "loss": 0.393,
+      "step": 326
+    },
+    {
+      "epoch": 2.6630286493860846,
+      "grad_norm": 0.7755050016922161,
+      "learning_rate": 3.603143363076217e-07,
+      "loss": 0.4036,
+      "step": 327
+    },
+    {
+      "epoch": 2.6712141882673945,
+      "grad_norm": 0.731356046504214,
+      "learning_rate": 3.427299773891868e-07,
+      "loss": 0.3863,
+      "step": 328
+    },
+    {
+      "epoch": 2.679399727148704,
+      "grad_norm": 0.777189147789671,
+      "learning_rate": 3.255702733006766e-07,
+      "loss": 0.3967,
+      "step": 329
+    },
+    {
+      "epoch": 2.6875852660300135,
+      "grad_norm": 0.840829025972722,
+      "learning_rate": 3.088367886821481e-07,
+      "loss": 0.3953,
+      "step": 330
+    },
+    {
+      "epoch": 2.6957708049113234,
+      "grad_norm": 0.7242832376127195,
+      "learning_rate": 2.925310493105099e-07,
+      "loss": 0.4203,
+      "step": 331
+    },
+    {
+      "epoch": 2.7039563437926333,
+      "grad_norm": 0.7565477747954039,
+      "learning_rate": 2.7665454196040665e-07,
+      "loss": 0.4053,
+      "step": 332
+    },
+    {
+      "epoch": 2.7121418826739427,
+      "grad_norm": 0.7749638803941362,
+      "learning_rate": 2.6120871426864866e-07,
+      "loss": 0.3805,
+      "step": 333
+    },
+    {
+      "epoch": 2.720327421555252,
+      "grad_norm": 0.771106186784679,
+      "learning_rate": 2.4619497460222184e-07,
+      "loss": 0.4104,
+      "step": 334
+    },
+    {
+      "epoch": 2.728512960436562,
+      "grad_norm": 0.7431816659046924,
+      "learning_rate": 2.316146919298623e-07,
+      "loss": 0.3869,
+      "step": 335
+    },
+    {
+      "epoch": 2.736698499317872,
+      "grad_norm": 0.7874811270318532,
+      "learning_rate": 2.1746919569723858e-07,
+      "loss": 0.4145,
+      "step": 336
+    },
+    {
+      "epoch": 2.7448840381991815,
+      "grad_norm": 0.7940636084356335,
+      "learning_rate": 2.037597757057297e-07,
+      "loss": 0.4082,
+      "step": 337
+    },
+    {
+      "epoch": 2.753069577080491,
+      "grad_norm": 0.7656364975895665,
+      "learning_rate": 1.9048768199481983e-07,
+      "loss": 0.3917,
+      "step": 338
+    },
+    {
+      "epoch": 2.761255115961801,
+      "grad_norm": 0.761990226097845,
+      "learning_rate": 1.776541247281177e-07,
+      "loss": 0.4099,
+      "step": 339
+    },
+    {
+      "epoch": 2.7694406548431107,
+      "grad_norm": 0.8341929047827159,
+      "learning_rate": 1.6526027408301227e-07,
+      "loss": 0.4033,
+      "step": 340
+    },
+    {
+      "epoch": 2.77762619372442,
+      "grad_norm": 0.8041923010242235,
+      "learning_rate": 1.5330726014397668e-07,
+      "loss": 0.412,
+      "step": 341
+    },
+    {
+      "epoch": 2.7858117326057297,
+      "grad_norm": 0.7666115759780836,
+      "learning_rate": 1.417961727995254e-07,
+      "loss": 0.4121,
+      "step": 342
+    },
+    {
+      "epoch": 2.7939972714870396,
+      "grad_norm": 0.7685299729666355,
+      "learning_rate": 1.307280616428336e-07,
+      "loss": 0.427,
+      "step": 343
+    },
+    {
+      "epoch": 2.8021828103683495,
+      "grad_norm": 0.7302009650736616,
+      "learning_rate": 1.2010393587603975e-07,
+      "loss": 0.4314,
+      "step": 344
+    },
+    {
+      "epoch": 2.810368349249659,
+      "grad_norm": 0.75693817420565,
+      "learning_rate": 1.0992476421822052e-07,
+      "loss": 0.4014,
+      "step": 345
+    },
+    {
+      "epoch": 2.8185538881309684,
+      "grad_norm": 0.7445337074971496,
+      "learning_rate": 1.0019147481706626e-07,
+      "loss": 0.3793,
+      "step": 346
+    },
+    {
+      "epoch": 2.8267394270122783,
+      "grad_norm": 0.7491472902145769,
+      "learning_rate": 9.090495516424713e-08,
+      "loss": 0.3935,
+      "step": 347
+    },
+    {
+      "epoch": 2.8349249658935882,
+      "grad_norm": 0.762122248462177,
+      "learning_rate": 8.206605201449447e-08,
+      "loss": 0.4103,
+      "step": 348
+    },
+    {
+      "epoch": 2.8431105047748977,
+      "grad_norm": 0.7952386558480641,
+      "learning_rate": 7.367557130838921e-08,
+      "loss": 0.4194,
+      "step": 349
+    },
+    {
+      "epoch": 2.851296043656207,
+      "grad_norm": 0.7193624590061964,
+      "learning_rate": 6.573427809888067e-08,
+      "loss": 0.4058,
+      "step": 350
+    },
+    {
+      "epoch": 2.859481582537517,
+      "grad_norm": 0.7746451418380024,
+      "learning_rate": 5.824289648152126e-08,
+      "loss": 0.4043,
+      "step": 351
+    },
+    {
+      "epoch": 2.867667121418827,
+      "grad_norm": 0.763755453581624,
+      "learning_rate": 5.120210952844873e-08,
+      "loss": 0.4079,
+      "step": 352
+    },
+    {
+      "epoch": 2.8758526603001364,
+      "grad_norm": 0.7679191261981648,
+      "learning_rate": 4.461255922609986e-08,
+      "loss": 0.4114,
+      "step": 353
+    },
+    {
+      "epoch": 2.884038199181446,
+      "grad_norm": 0.759091588522436,
+      "learning_rate": 3.8474846416672874e-08,
+      "loss": 0.3987,
+      "step": 354
+    },
+    {
+      "epoch": 2.892223738062756,
+      "grad_norm": 0.7542406340165235,
+      "learning_rate": 3.278953074334512e-08,
+      "loss": 0.3947,
+      "step": 355
+    },
+    {
+      "epoch": 2.9004092769440657,
+      "grad_norm": 0.7601715694682697,
+      "learning_rate": 2.75571305992417e-08,
+      "loss": 0.3892,
+      "step": 356
+    },
+    {
+      "epoch": 2.908594815825375,
+      "grad_norm": 0.7260489305968849,
+      "learning_rate": 2.2778123080167136e-08,
+      "loss": 0.4151,
+      "step": 357
+    },
+    {
+      "epoch": 2.9167803547066846,
+      "grad_norm": 0.7764596030153685,
+      "learning_rate": 1.845294394110686e-08,
+      "loss": 0.3986,
+      "step": 358
+    },
+    {
+      "epoch": 2.9249658935879945,
+      "grad_norm": 0.7553548836840808,
+      "learning_rate": 1.4581987556490095e-08,
+      "loss": 0.4214,
+      "step": 359
+    },
+    {
+      "epoch": 2.9331514324693044,
+      "grad_norm": 0.7488951591031938,
+      "learning_rate": 1.1165606884234182e-08,
+      "loss": 0.4103,
+      "step": 360
+    },
+    {
+      "epoch": 2.941336971350614,
+      "grad_norm": 0.7487421289826812,
+      "learning_rate": 8.204113433559202e-09,
+      "loss": 0.3706,
+      "step": 361
+    },
+    {
+      "epoch": 2.9495225102319234,
+      "grad_norm": 0.7105264155773031,
+      "learning_rate": 5.6977772365857105e-09,
+      "loss": 0.3931,
+      "step": 362
+    },
+    {
+      "epoch": 2.9577080491132333,
+      "grad_norm": 0.7505520876979219,
+      "learning_rate": 3.6468268237105364e-09,
+      "loss": 0.411,
+      "step": 363
+    },
+    {
+      "epoch": 2.965893587994543,
+      "grad_norm": 0.7891148389786055,
+      "learning_rate": 2.0514492027728928e-09,
+      "loss": 0.4077,
+      "step": 364
+    },
+    {
+      "epoch": 2.9740791268758526,
+      "grad_norm": 0.7571146450027982,
+      "learning_rate": 9.117898419991333e-10,
+      "loss": 0.3988,
+      "step": 365
+    },
+    {
+      "epoch": 2.982264665757162,
+      "grad_norm": 0.7366594553416043,
+      "learning_rate": 2.2795265674113721e-10,
+      "loss": 0.3827,
+      "step": 366
+    },
+    {
+      "epoch": 2.982264665757162,
+      "step": 366,
+      "total_flos": 5.395725452134318e+17,
+      "train_loss": 0.5498430520458951,
+      "train_runtime": 28891.7819,
+      "train_samples_per_second": 0.609,
+      "train_steps_per_second": 0.013
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 366,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.395725452134318e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}