Training in progress, step 500, checkpoint

checkpoint-500/adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "v_proj",
-    "q_proj",
-    "gate_proj",
     "up_proj",
     "o_proj",
+    "v_proj",
+    "q_proj",
+    "down_proj",
     "k_proj",
-    "down_proj"
+    "gate_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

checkpoint-500/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:882c6a449a15a7c89b405a9b2bb953973aabb46d5bb332fd97e2636db7009c10
+oid sha256:77ad03a26563fb97d39b54488373b39fc25f88d455d31ef54c5aab28afa972cc
 size 167832240

checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec4b78ace1156f82710f9323015277767c2c7913e75b033470080d89dafce4ea
+oid sha256:5341d64ff65b343938bba55a46153c0ffa1516010b29f36f738cdc6ed7a1a7f8
 size 335922386

checkpoint-500/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f45fad1928be6a3cba447ba135798b1f889eac81fc87eb617b20212d56ec355
+oid sha256:ec4f25bf25febb815eb9a87cc696ea933d62d4744509c6c7ee46b8ee52dfff33
 size 14244

checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d3dd7bf6e53c14f5c31b3d93195825c804547bfedb7e1ffecff6ed079275331
+oid sha256:eb33de8afdd2f737b169d16a4787977e578963170b5dd53b43167e251c838ee3
 size 1064

checkpoint-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.5252100840336135,
+  "epoch": 0.5268703898840885,
   "eval_steps": 50,
   "global_step": 500,
   "is_hyper_param_search": false,
@@ -9,438 +9,438 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01050420168067227,
-      "grad_norm": 2.5582146644592285,
-      "learning_rate": 4.98249299719888e-05,
-      "loss": 1.6787,
+      "epoch": 0.01053740779768177,
+      "grad_norm": 3.070249080657959,
+      "learning_rate": 4.982437653670531e-05,
+      "loss": 1.7879,
       "step": 10
     },
     {
-      "epoch": 0.02100840336134454,
-      "grad_norm": 0.9345070719718933,
-      "learning_rate": 4.96498599439776e-05,
-      "loss": 0.518,
+      "epoch": 0.02107481559536354,
+      "grad_norm": 1.702326774597168,
+      "learning_rate": 4.964875307341061e-05,
+      "loss": 0.5567,
       "step": 20
     },
     {
-      "epoch": 0.031512605042016806,
-      "grad_norm": 1.6358414888381958,
-      "learning_rate": 4.947478991596639e-05,
-      "loss": 0.4604,
+      "epoch": 0.03161222339304531,
+      "grad_norm": 1.1947294473648071,
+      "learning_rate": 4.947312961011591e-05,
+      "loss": 0.4493,
       "step": 30
     },
     {
-      "epoch": 0.04201680672268908,
-      "grad_norm": 0.7778844237327576,
-      "learning_rate": 4.9299719887955186e-05,
-      "loss": 0.3771,
+      "epoch": 0.04214963119072708,
+      "grad_norm": 0.9556658267974854,
+      "learning_rate": 4.929750614682122e-05,
+      "loss": 0.3728,
       "step": 40
     },
     {
-      "epoch": 0.052521008403361345,
-      "grad_norm": 0.7006077766418457,
-      "learning_rate": 4.912464985994398e-05,
-      "loss": 0.3842,
+      "epoch": 0.05268703898840885,
+      "grad_norm": 0.7952510714530945,
+      "learning_rate": 4.9121882683526524e-05,
+      "loss": 0.3535,
       "step": 50
     },
     {
-      "epoch": 0.052521008403361345,
-      "eval_loss": 0.42603224515914917,
-      "eval_runtime": 13.5673,
-      "eval_samples_per_second": 35.379,
-      "eval_steps_per_second": 2.211,
+      "epoch": 0.05268703898840885,
+      "eval_loss": 0.4311191439628601,
+      "eval_runtime": 13.6539,
+      "eval_samples_per_second": 35.155,
+      "eval_steps_per_second": 2.197,
       "step": 50
     },
     {
-      "epoch": 0.06302521008403361,
-      "grad_norm": 0.6415153741836548,
-      "learning_rate": 4.8949579831932775e-05,
-      "loss": 0.3399,
+      "epoch": 0.06322444678609063,
+      "grad_norm": 0.6962826251983643,
+      "learning_rate": 4.894625922023183e-05,
+      "loss": 0.3507,
       "step": 60
     },
     {
-      "epoch": 0.07352941176470588,
-      "grad_norm": 0.6030780076980591,
-      "learning_rate": 4.877450980392157e-05,
-      "loss": 0.3447,
+      "epoch": 0.0737618545837724,
+      "grad_norm": 0.6941961646080017,
+      "learning_rate": 4.877063575693713e-05,
+      "loss": 0.3585,
       "step": 70
     },
     {
-      "epoch": 0.08403361344537816,
-      "grad_norm": 0.688852071762085,
-      "learning_rate": 4.859943977591036e-05,
-      "loss": 0.3219,
+      "epoch": 0.08429926238145416,
+      "grad_norm": 0.6864392757415771,
+      "learning_rate": 4.8595012293642434e-05,
+      "loss": 0.3496,
       "step": 80
     },
     {
-      "epoch": 0.09453781512605042,
-      "grad_norm": 0.6371557712554932,
-      "learning_rate": 4.8424369747899164e-05,
-      "loss": 0.3379,
+      "epoch": 0.09483667017913593,
+      "grad_norm": 0.7322937846183777,
+      "learning_rate": 4.841938883034774e-05,
+      "loss": 0.3295,
       "step": 90
     },
     {
-      "epoch": 0.10504201680672269,
-      "grad_norm": 0.7739270329475403,
-      "learning_rate": 4.824929971988796e-05,
-      "loss": 0.3177,
+      "epoch": 0.1053740779768177,
+      "grad_norm": 0.6921488046646118,
+      "learning_rate": 4.824376536705304e-05,
+      "loss": 0.3357,
       "step": 100
     },
     {
-      "epoch": 0.10504201680672269,
-      "eval_loss": 0.3801896274089813,
-      "eval_runtime": 13.6107,
-      "eval_samples_per_second": 35.266,
-      "eval_steps_per_second": 2.204,
+      "epoch": 0.1053740779768177,
+      "eval_loss": 0.39120009541511536,
+      "eval_runtime": 13.7031,
+      "eval_samples_per_second": 35.029,
+      "eval_steps_per_second": 2.189,
       "step": 100
     },
     {
-      "epoch": 0.11554621848739496,
-      "grad_norm": 0.649507462978363,
-      "learning_rate": 4.807422969187675e-05,
-      "loss": 0.3415,
+      "epoch": 0.11591148577449947,
+      "grad_norm": 0.6553240418434143,
+      "learning_rate": 4.8068141903758344e-05,
+      "loss": 0.3105,
       "step": 110
     },
     {
-      "epoch": 0.12605042016806722,
-      "grad_norm": 0.594717264175415,
-      "learning_rate": 4.7899159663865554e-05,
-      "loss": 0.3325,
+      "epoch": 0.12644889357218125,
+      "grad_norm": 0.5637819170951843,
+      "learning_rate": 4.789251844046364e-05,
+      "loss": 0.3164,
       "step": 120
     },
     {
-      "epoch": 0.13655462184873948,
-      "grad_norm": 0.627918541431427,
-      "learning_rate": 4.772408963585435e-05,
-      "loss": 0.3222,
+      "epoch": 0.136986301369863,
+      "grad_norm": 0.6341928839683533,
+      "learning_rate": 4.7716894977168955e-05,
+      "loss": 0.304,
       "step": 130
     },
     {
-      "epoch": 0.14705882352941177,
-      "grad_norm": 0.5384674668312073,
-      "learning_rate": 4.7549019607843135e-05,
-      "loss": 0.3426,
+      "epoch": 0.1475237091675448,
+      "grad_norm": 0.5917785167694092,
+      "learning_rate": 4.754127151387426e-05,
+      "loss": 0.3234,
       "step": 140
     },
     {
-      "epoch": 0.15756302521008403,
-      "grad_norm": 0.5673420429229736,
-      "learning_rate": 4.7373949579831936e-05,
-      "loss": 0.3061,
+      "epoch": 0.15806111696522657,
+      "grad_norm": 0.5884453654289246,
+      "learning_rate": 4.736564805057956e-05,
+      "loss": 0.317,
       "step": 150
     },
     {
-      "epoch": 0.15756302521008403,
-      "eval_loss": 0.3653399348258972,
-      "eval_runtime": 13.5947,
-      "eval_samples_per_second": 35.308,
-      "eval_steps_per_second": 2.207,
+      "epoch": 0.15806111696522657,
+      "eval_loss": 0.37688738107681274,
+      "eval_runtime": 13.6535,
+      "eval_samples_per_second": 35.156,
+      "eval_steps_per_second": 2.197,
       "step": 150
     },
     {
-      "epoch": 0.16806722689075632,
-      "grad_norm": 0.6111018657684326,
-      "learning_rate": 4.719887955182073e-05,
-      "loss": 0.3271,
+      "epoch": 0.16859852476290832,
+      "grad_norm": 0.5819964408874512,
+      "learning_rate": 4.7190024587284866e-05,
+      "loss": 0.2992,
       "step": 160
     },
     {
-      "epoch": 0.17857142857142858,
-      "grad_norm": 0.7422594428062439,
-      "learning_rate": 4.7023809523809525e-05,
-      "loss": 0.315,
+      "epoch": 0.1791359325605901,
+      "grad_norm": 0.689468264579773,
+      "learning_rate": 4.7014401123990165e-05,
+      "loss": 0.3168,
       "step": 170
     },
     {
-      "epoch": 0.18907563025210083,
-      "grad_norm": 0.7226534485816956,
-      "learning_rate": 4.684873949579832e-05,
-      "loss": 0.3031,
+      "epoch": 0.18967334035827185,
+      "grad_norm": 0.6950872540473938,
+      "learning_rate": 4.683877766069547e-05,
+      "loss": 0.3041,
       "step": 180
     },
     {
-      "epoch": 0.19957983193277312,
-      "grad_norm": 0.6302976012229919,
-      "learning_rate": 4.667366946778712e-05,
-      "loss": 0.3161,
+      "epoch": 0.20021074815595363,
+      "grad_norm": 0.8322122097015381,
+      "learning_rate": 4.6663154197400776e-05,
+      "loss": 0.3028,
       "step": 190
     },
     {
-      "epoch": 0.21008403361344538,
-      "grad_norm": 0.6225076913833618,
-      "learning_rate": 4.6498599439775914e-05,
-      "loss": 0.3038,
+      "epoch": 0.2107481559536354,
+      "grad_norm": 0.5850774645805359,
+      "learning_rate": 4.6487530734106075e-05,
+      "loss": 0.2992,
       "step": 200
     },
     {
-      "epoch": 0.21008403361344538,
-      "eval_loss": 0.35061606764793396,
-      "eval_runtime": 13.5616,
-      "eval_samples_per_second": 35.394,
-      "eval_steps_per_second": 2.212,
+      "epoch": 0.2107481559536354,
+      "eval_loss": 0.36230018734931946,
+      "eval_runtime": 13.6165,
+      "eval_samples_per_second": 35.251,
+      "eval_steps_per_second": 2.203,
       "step": 200
     },
     {
-      "epoch": 0.22058823529411764,
-      "grad_norm": 0.6001319885253906,
-      "learning_rate": 4.632352941176471e-05,
-      "loss": 0.3129,
+      "epoch": 0.22128556375131717,
+      "grad_norm": 0.6668715476989746,
+      "learning_rate": 4.631190727081138e-05,
+      "loss": 0.2924,
       "step": 210
     },
     {
-      "epoch": 0.23109243697478993,
-      "grad_norm": 0.5385990142822266,
-      "learning_rate": 4.61484593837535e-05,
-      "loss": 0.2991,
+      "epoch": 0.23182297154899895,
+      "grad_norm": 0.4749641418457031,
+      "learning_rate": 4.6136283807516686e-05,
+      "loss": 0.3017,
       "step": 220
     },
     {
-      "epoch": 0.2415966386554622,
-      "grad_norm": 0.4513624906539917,
-      "learning_rate": 4.59733893557423e-05,
-      "loss": 0.2896,
+      "epoch": 0.24236037934668073,
+      "grad_norm": 0.6381515860557556,
+      "learning_rate": 4.596066034422199e-05,
+      "loss": 0.2887,
       "step": 230
     },
     {
-      "epoch": 0.25210084033613445,
-      "grad_norm": 0.6142160892486572,
-      "learning_rate": 4.579831932773109e-05,
-      "loss": 0.3059,
+      "epoch": 0.2528977871443625,
+      "grad_norm": 0.49952977895736694,
+      "learning_rate": 4.57850368809273e-05,
+      "loss": 0.2833,
       "step": 240
     },
     {
-      "epoch": 0.26260504201680673,
-      "grad_norm": 0.6714802384376526,
-      "learning_rate": 4.562324929971989e-05,
-      "loss": 0.2897,
+      "epoch": 0.26343519494204426,
+      "grad_norm": 0.699518084526062,
+      "learning_rate": 4.5609413417632596e-05,
+      "loss": 0.2859,
       "step": 250
     },
     {
-      "epoch": 0.26260504201680673,
-      "eval_loss": 0.3456435203552246,
-      "eval_runtime": 13.5552,
-      "eval_samples_per_second": 35.411,
-      "eval_steps_per_second": 2.213,
+      "epoch": 0.26343519494204426,
+      "eval_loss": 0.3563433885574341,
+      "eval_runtime": 13.6275,
+      "eval_samples_per_second": 35.223,
+      "eval_steps_per_second": 2.201,
       "step": 250
     },
     {
-      "epoch": 0.27310924369747897,
-      "grad_norm": 0.6518235206604004,
-      "learning_rate": 4.5448179271708687e-05,
-      "loss": 0.312,
+      "epoch": 0.273972602739726,
+      "grad_norm": 0.5912085175514221,
+      "learning_rate": 4.54337899543379e-05,
+      "loss": 0.302,
       "step": 260
     },
     {
-      "epoch": 0.28361344537815125,
-      "grad_norm": 0.6250632405281067,
-      "learning_rate": 4.527310924369748e-05,
-      "loss": 0.2959,
+      "epoch": 0.2845100105374078,
+      "grad_norm": 0.6353363990783691,
+      "learning_rate": 4.525816649104321e-05,
+      "loss": 0.3112,
       "step": 270
     },
     {
-      "epoch": 0.29411764705882354,
-      "grad_norm": 0.5683826804161072,
-      "learning_rate": 4.5098039215686275e-05,
-      "loss": 0.3027,
+      "epoch": 0.2950474183350896,
+      "grad_norm": 0.5483567118644714,
+      "learning_rate": 4.5082543027748506e-05,
+      "loss": 0.2808,
       "step": 280
     },
     {
-      "epoch": 0.30462184873949577,
-      "grad_norm": 0.560312807559967,
-      "learning_rate": 4.4922969187675076e-05,
-      "loss": 0.3002,
+      "epoch": 0.3055848261327713,
+      "grad_norm": 0.5003193616867065,
+      "learning_rate": 4.490691956445381e-05,
+      "loss": 0.3185,
       "step": 290
     },
     {
-      "epoch": 0.31512605042016806,
-      "grad_norm": 0.66291743516922,
-      "learning_rate": 4.474789915966387e-05,
-      "loss": 0.2925,
+      "epoch": 0.31612223393045313,
+      "grad_norm": 0.4919240176677704,
+      "learning_rate": 4.473129610115912e-05,
+      "loss": 0.2883,
       "step": 300
     },
     {
-      "epoch": 0.31512605042016806,
-      "eval_loss": 0.3431606888771057,
-      "eval_runtime": 13.5629,
-      "eval_samples_per_second": 35.391,
-      "eval_steps_per_second": 2.212,
+      "epoch": 0.31612223393045313,
+      "eval_loss": 0.35315924882888794,
+      "eval_runtime": 13.6133,
+      "eval_samples_per_second": 35.26,
+      "eval_steps_per_second": 2.204,
       "step": 300
     },
     {
-      "epoch": 0.32563025210084034,
-      "grad_norm": 0.6478439569473267,
-      "learning_rate": 4.4572829131652665e-05,
-      "loss": 0.2893,
+      "epoch": 0.3266596417281349,
+      "grad_norm": 0.5145038366317749,
+      "learning_rate": 4.455567263786442e-05,
+      "loss": 0.3084,
       "step": 310
     },
     {
-      "epoch": 0.33613445378151263,
-      "grad_norm": 0.5832348465919495,
-      "learning_rate": 4.439775910364146e-05,
-      "loss": 0.2842,
+      "epoch": 0.33719704952581664,
+      "grad_norm": 0.6343855261802673,
+      "learning_rate": 4.438004917456973e-05,
+      "loss": 0.295,
       "step": 320
     },
     {
-      "epoch": 0.34663865546218486,
-      "grad_norm": 0.525932252407074,
-      "learning_rate": 4.422268907563025e-05,
-      "loss": 0.2837,
+      "epoch": 0.34773445732349845,
+      "grad_norm": 0.5336400270462036,
+      "learning_rate": 4.420442571127503e-05,
+      "loss": 0.2882,
       "step": 330
     },
     {
-      "epoch": 0.35714285714285715,
-      "grad_norm": 0.5487508177757263,
-      "learning_rate": 4.404761904761905e-05,
-      "loss": 0.2706,
+      "epoch": 0.3582718651211802,
+      "grad_norm": 0.49547308683395386,
+      "learning_rate": 4.4028802247980333e-05,
+      "loss": 0.3084,
       "step": 340
     },
     {
-      "epoch": 0.36764705882352944,
-      "grad_norm": 0.5392388701438904,
-      "learning_rate": 4.387254901960784e-05,
-      "loss": 0.2835,
+      "epoch": 0.36880927291886195,
+      "grad_norm": 0.5037292242050171,
+      "learning_rate": 4.385317878468563e-05,
+      "loss": 0.3084,
       "step": 350
     },
     {
-      "epoch": 0.36764705882352944,
-      "eval_loss": 0.33528536558151245,
-      "eval_runtime": 13.5508,
-      "eval_samples_per_second": 35.422,
-      "eval_steps_per_second": 2.214,
+      "epoch": 0.36880927291886195,
+      "eval_loss": 0.34608179330825806,
+      "eval_runtime": 13.6537,
+      "eval_samples_per_second": 35.155,
+      "eval_steps_per_second": 2.197,
       "step": 350
     },
     {
-      "epoch": 0.37815126050420167,
-      "grad_norm": 0.6706260442733765,
-      "learning_rate": 4.369747899159664e-05,
-      "loss": 0.2844,
+      "epoch": 0.3793466807165437,
+      "grad_norm": 0.5956543684005737,
+      "learning_rate": 4.367755532139094e-05,
+      "loss": 0.2763,
       "step": 360
     },
     {
-      "epoch": 0.38865546218487396,
-      "grad_norm": 0.6042625904083252,
-      "learning_rate": 4.352240896358544e-05,
-      "loss": 0.2758,
+      "epoch": 0.3898840885142255,
+      "grad_norm": 0.6263634562492371,
+      "learning_rate": 4.3501931858096244e-05,
+      "loss": 0.3025,
       "step": 370
     },
     {
-      "epoch": 0.39915966386554624,
-      "grad_norm": 0.534008264541626,
-      "learning_rate": 4.334733893557423e-05,
-      "loss": 0.2918,
+      "epoch": 0.40042149631190727,
+      "grad_norm": 0.4832920730113983,
+      "learning_rate": 4.332630839480154e-05,
+      "loss": 0.2792,
       "step": 380
     },
     {
-      "epoch": 0.4096638655462185,
-      "grad_norm": 0.48162588477134705,
-      "learning_rate": 4.317226890756303e-05,
-      "loss": 0.273,
+      "epoch": 0.410958904109589,
+      "grad_norm": 0.4969714879989624,
+      "learning_rate": 4.3150684931506855e-05,
+      "loss": 0.2704,
       "step": 390
     },
     {
-      "epoch": 0.42016806722689076,
-      "grad_norm": 0.5669644474983215,
-      "learning_rate": 4.2997198879551826e-05,
-      "loss": 0.285,
+      "epoch": 0.4214963119072708,
+      "grad_norm": 0.568900465965271,
+      "learning_rate": 4.297506146821216e-05,
+      "loss": 0.2734,
       "step": 400
     },
     {
-      "epoch": 0.42016806722689076,
-      "eval_loss": 0.3348632752895355,
-      "eval_runtime": 13.5507,
-      "eval_samples_per_second": 35.423,
-      "eval_steps_per_second": 2.214,
+      "epoch": 0.4214963119072708,
+      "eval_loss": 0.34151414036750793,
+      "eval_runtime": 13.664,
+      "eval_samples_per_second": 35.129,
+      "eval_steps_per_second": 2.196,
       "step": 400
     },
     {
-      "epoch": 0.43067226890756305,
-      "grad_norm": 0.6257824897766113,
-      "learning_rate": 4.2822128851540614e-05,
-      "loss": 0.299,
+      "epoch": 0.4320337197049526,
+      "grad_norm": 0.5023282170295715,
+      "learning_rate": 4.279943800491746e-05,
+      "loss": 0.2845,
       "step": 410
     },
     {
-      "epoch": 0.4411764705882353,
-      "grad_norm": 0.5430576205253601,
-      "learning_rate": 4.2647058823529415e-05,
-      "loss": 0.2868,
+      "epoch": 0.44257112750263433,
+      "grad_norm": 0.531538188457489,
+      "learning_rate": 4.2623814541622765e-05,
+      "loss": 0.276,
       "step": 420
     },
     {
-      "epoch": 0.45168067226890757,
-      "grad_norm": 0.5633955597877502,
-      "learning_rate": 4.247198879551821e-05,
-      "loss": 0.2589,
+      "epoch": 0.45310853530031614,
+      "grad_norm": 0.6686979532241821,
+      "learning_rate": 4.2448191078328064e-05,
+      "loss": 0.283,
       "step": 430
     },
     {
-      "epoch": 0.46218487394957986,
-      "grad_norm": 0.5294789671897888,
-      "learning_rate": 4.2296918767507e-05,
-      "loss": 0.2777,
+      "epoch": 0.4636459430979979,
+      "grad_norm": 0.5194190144538879,
+      "learning_rate": 4.227256761503337e-05,
+      "loss": 0.2636,
       "step": 440
     },
     {
-      "epoch": 0.4726890756302521,
-      "grad_norm": 0.5480856895446777,
-      "learning_rate": 4.21218487394958e-05,
-      "loss": 0.2704,
+      "epoch": 0.47418335089567965,
+      "grad_norm": 0.6105541586875916,
+      "learning_rate": 4.2096944151738675e-05,
+      "loss": 0.2713,
       "step": 450
     },
     {
-      "epoch": 0.4726890756302521,
-      "eval_loss": 0.329515278339386,
-      "eval_runtime": 13.5423,
-      "eval_samples_per_second": 35.445,
-      "eval_steps_per_second": 2.215,
+      "epoch": 0.47418335089567965,
+      "eval_loss": 0.33879777789115906,
+      "eval_runtime": 13.6666,
+      "eval_samples_per_second": 35.122,
+      "eval_steps_per_second": 2.195,
       "step": 450
     },
     {
-      "epoch": 0.4831932773109244,
-      "grad_norm": 0.5051332116127014,
-      "learning_rate": 4.19467787114846e-05,
-      "loss": 0.2438,
+      "epoch": 0.48472075869336145,
+      "grad_norm": 0.4929138123989105,
+      "learning_rate": 4.1921320688443974e-05,
+      "loss": 0.261,
       "step": 460
     },
     {
-      "epoch": 0.49369747899159666,
-      "grad_norm": 0.6251511573791504,
-      "learning_rate": 4.177170868347339e-05,
-      "loss": 0.2748,
+      "epoch": 0.4952581664910432,
+      "grad_norm": 0.584095299243927,
+      "learning_rate": 4.174569722514928e-05,
+      "loss": 0.2714,
       "step": 470
     },
     {
-      "epoch": 0.5042016806722689,
-      "grad_norm": 0.4729413092136383,
-      "learning_rate": 4.159663865546219e-05,
-      "loss": 0.2689,
+      "epoch": 0.505795574288725,
+      "grad_norm": 0.5386167764663696,
+      "learning_rate": 4.1570073761854585e-05,
+      "loss": 0.2855,
       "step": 480
     },
     {
-      "epoch": 0.5147058823529411,
-      "grad_norm": 0.5220003724098206,
-      "learning_rate": 4.142156862745099e-05,
-      "loss": 0.2899,
+      "epoch": 0.5163329820864068,
+      "grad_norm": 0.5819774270057678,
+      "learning_rate": 4.139445029855989e-05,
+      "loss": 0.2911,
       "step": 490
     },
     {
-      "epoch": 0.5252100840336135,
-      "grad_norm": 0.54283207654953,
-      "learning_rate": 4.1246498599439776e-05,
-      "loss": 0.272,
+      "epoch": 0.5268703898840885,
+      "grad_norm": 0.5875944495201111,
+      "learning_rate": 4.12188268352652e-05,
+      "loss": 0.2619,
       "step": 500
     },
     {
-      "epoch": 0.5252100840336135,
-      "eval_loss": 0.32714489102363586,
-      "eval_runtime": 13.5497,
-      "eval_samples_per_second": 35.425,
-      "eval_steps_per_second": 2.214,
+      "epoch": 0.5268703898840885,
+      "eval_loss": 0.3364439010620117,
+      "eval_runtime": 13.6731,
+      "eval_samples_per_second": 35.105,
+      "eval_steps_per_second": 2.194,
       "step": 500
     }
   ],
   "logging_steps": 10,
-  "max_steps": 2856,
+  "max_steps": 2847,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 100,
@@ -456,7 +456,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.977409027624141e+16,
+  "total_flos": 5.975090629863014e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d5894a90f0aacad19de132730666f8b4647a0c4aa14309866a5f87d3723ce6a7
+oid sha256:b5fa9d8b62d1ebe6967a504a7decdb5eeee2bb4aac96e7414f3930f9adcff095
 size 5880