End of training

README.md

@@ -4,6 +4,7 @@ license: llama3.1
 base_model: meta-llama/Llama-3.1-8B
 tags:
 - llama-factory
+- full
 - generated_from_trainer
 model-index:
 - name: OH_DCFT_V3_wo_airoboros
@@ -15,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # OH_DCFT_V3_wo_airoboros
 
-This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on an unknown dataset.
+This model is a fine-tuned version of [meta-llama/Llama-3.1-8B](https://huggingface.co/meta-llama/Llama-3.1-8B) on the mlfoundations-dev/OH_DCFT_V3_wo_airoboros dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.6375
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 3.0,
-    "eval_loss": 0.6411369442939758,
-    "eval_runtime": 219.0232,
-    "eval_samples_per_second": 50.049,
-    "eval_steps_per_second": 0.393,
+    "eval_loss": 0.6375428438186646,
+    "eval_runtime": 218.1807,
+    "eval_samples_per_second": 50.243,
+    "eval_steps_per_second": 0.394,
     "total_flos": 2044801717370880.0,
-    "train_loss": 0.6174200418344977,
-    "train_runtime": 36857.6484,
-    "train_samples_per_second": 16.952,
+    "train_loss": 0.6189516479367608,
+    "train_runtime": 36570.9767,
+    "train_samples_per_second": 17.085,
     "train_steps_per_second": 0.033
 }

eval_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 3.0,
-    "eval_loss": 0.6411369442939758,
-    "eval_runtime": 219.0232,
-    "eval_samples_per_second": 50.049,
-    "eval_steps_per_second": 0.393
+    "eval_loss": 0.6375428438186646,
+    "eval_runtime": 218.1807,
+    "eval_samples_per_second": 50.243,
+    "eval_steps_per_second": 0.394
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.0,
     "total_flos": 2044801717370880.0,
-    "train_loss": 0.6174200418344977,
-    "train_runtime": 36857.6484,
-    "train_samples_per_second": 16.952,
+    "train_loss": 0.6189516479367608,
+    "train_runtime": 36570.9767,
+    "train_samples_per_second": 17.085,
     "train_steps_per_second": 0.033
 }

trainer_state.json

@@ -10,889 +10,889 @@
   "log_history": [
     {
       "epoch": 0.02457002457002457,
-      "grad_norm": 8.946310483633335,
+      "grad_norm": 6.1998985197847745,
       "learning_rate": 5e-06,
-      "loss": 0.9151,
+      "loss": 0.8819,
       "step": 10
     },
     {
       "epoch": 0.04914004914004914,
-      "grad_norm": 2.981455204527646,
+      "grad_norm": 58.73927160489986,
       "learning_rate": 5e-06,
-      "loss": 0.8186,
+      "loss": 0.7904,
       "step": 20
     },
     {
       "epoch": 0.07371007371007371,
-      "grad_norm": 1.32074337919512,
+      "grad_norm": 0.7748310375081953,
       "learning_rate": 5e-06,
-      "loss": 0.7773,
+      "loss": 0.7623,
       "step": 30
     },
     {
       "epoch": 0.09828009828009827,
-      "grad_norm": 3.907331439899627,
+      "grad_norm": 1.0183734838338003,
       "learning_rate": 5e-06,
-      "loss": 0.752,
+      "loss": 0.7382,
       "step": 40
     },
     {
       "epoch": 0.12285012285012285,
-      "grad_norm": 0.8221167183398617,
+      "grad_norm": 1.4797556953661242,
       "learning_rate": 5e-06,
-      "loss": 0.7311,
+      "loss": 0.7184,
       "step": 50
     },
     {
       "epoch": 0.14742014742014742,
-      "grad_norm": 4.06410196141404,
+      "grad_norm": 0.9599955498815156,
       "learning_rate": 5e-06,
-      "loss": 0.7249,
+      "loss": 0.714,
       "step": 60
     },
     {
       "epoch": 0.171990171990172,
-      "grad_norm": 1.6213863948934828,
+      "grad_norm": 1.0089273827026637,
       "learning_rate": 5e-06,
-      "loss": 0.7188,
+      "loss": 0.7048,
       "step": 70
     },
     {
       "epoch": 0.19656019656019655,
-      "grad_norm": 0.9311393114626392,
+      "grad_norm": 0.7283907495801997,
       "learning_rate": 5e-06,
-      "loss": 0.7116,
+      "loss": 0.7015,
       "step": 80
     },
     {
       "epoch": 0.22113022113022113,
-      "grad_norm": 1.5115946625339785,
+      "grad_norm": 0.665386593224946,
       "learning_rate": 5e-06,
-      "loss": 0.7015,
+      "loss": 0.6941,
       "step": 90
     },
     {
       "epoch": 0.2457002457002457,
-      "grad_norm": 0.5710441461897465,
+      "grad_norm": 0.6931565576787211,
       "learning_rate": 5e-06,
-      "loss": 0.6967,
+      "loss": 0.6901,
       "step": 100
     },
     {
       "epoch": 0.2702702702702703,
-      "grad_norm": 0.4997697728974884,
+      "grad_norm": 0.5017251950296421,
       "learning_rate": 5e-06,
-      "loss": 0.6902,
+      "loss": 0.6844,
       "step": 110
     },
     {
       "epoch": 0.29484029484029484,
-      "grad_norm": 0.8729499790550864,
+      "grad_norm": 0.5071354456321502,
       "learning_rate": 5e-06,
-      "loss": 0.6934,
+      "loss": 0.6879,
       "step": 120
     },
     {
       "epoch": 0.3194103194103194,
-      "grad_norm": 0.6657672302515885,
+      "grad_norm": 0.6070247331517776,
       "learning_rate": 5e-06,
-      "loss": 0.6765,
+      "loss": 0.6715,
       "step": 130
     },
     {
       "epoch": 0.343980343980344,
-      "grad_norm": 0.6470656412612269,
+      "grad_norm": 0.627106293158071,
       "learning_rate": 5e-06,
-      "loss": 0.6819,
+      "loss": 0.6774,
       "step": 140
     },
     {
       "epoch": 0.36855036855036855,
-      "grad_norm": 0.5249575222315258,
+      "grad_norm": 0.7310651144047745,
       "learning_rate": 5e-06,
-      "loss": 0.6715,
+      "loss": 0.6675,
       "step": 150
     },
     {
       "epoch": 0.3931203931203931,
-      "grad_norm": 0.5548655642844337,
+      "grad_norm": 0.5686459982063066,
       "learning_rate": 5e-06,
-      "loss": 0.6793,
+      "loss": 0.6755,
       "step": 160
     },
     {
       "epoch": 0.4176904176904177,
-      "grad_norm": 0.6921445594896967,
+      "grad_norm": 0.5320144044432522,
       "learning_rate": 5e-06,
-      "loss": 0.6747,
+      "loss": 0.6711,
       "step": 170
     },
     {
       "epoch": 0.44226044226044225,
-      "grad_norm": 0.6728846658943244,
+      "grad_norm": 0.7374032221533481,
       "learning_rate": 5e-06,
-      "loss": 0.6814,
+      "loss": 0.6781,
       "step": 180
     },
     {
       "epoch": 0.4668304668304668,
-      "grad_norm": 0.8716022919185289,
+      "grad_norm": 0.7028348920446017,
       "learning_rate": 5e-06,
-      "loss": 0.6687,
+      "loss": 0.6655,
       "step": 190
     },
     {
       "epoch": 0.4914004914004914,
-      "grad_norm": 0.5542689146513952,
+      "grad_norm": 0.6462713543254518,
       "learning_rate": 5e-06,
-      "loss": 0.6757,
+      "loss": 0.6725,
       "step": 200
     },
     {
       "epoch": 0.515970515970516,
-      "grad_norm": 0.8273130681045643,
+      "grad_norm": 0.5063976505878001,
       "learning_rate": 5e-06,
-      "loss": 0.6652,
+      "loss": 0.6623,
       "step": 210
     },
     {
       "epoch": 0.5405405405405406,
-      "grad_norm": 0.6283391489980882,
+      "grad_norm": 0.5560125316540017,
       "learning_rate": 5e-06,
-      "loss": 0.661,
+      "loss": 0.6583,
       "step": 220
     },
     {
       "epoch": 0.5651105651105651,
-      "grad_norm": 0.6977475503407465,
+      "grad_norm": 0.7044183401160061,
       "learning_rate": 5e-06,
-      "loss": 0.6693,
+      "loss": 0.6669,
       "step": 230
     },
     {
       "epoch": 0.5896805896805897,
-      "grad_norm": 0.5708084653966776,
+      "grad_norm": 0.547110632127531,
       "learning_rate": 5e-06,
-      "loss": 0.6627,
+      "loss": 0.6604,
       "step": 240
     },
     {
       "epoch": 0.6142506142506142,
-      "grad_norm": 0.5693320867441568,
+      "grad_norm": 0.45751487478129493,
       "learning_rate": 5e-06,
-      "loss": 0.6516,
+      "loss": 0.6496,
       "step": 250
     },
     {
       "epoch": 0.6388206388206388,
-      "grad_norm": 0.5519412849870617,
+      "grad_norm": 0.4993294423942927,
       "learning_rate": 5e-06,
-      "loss": 0.6652,
+      "loss": 0.663,
       "step": 260
     },
     {
       "epoch": 0.6633906633906634,
-      "grad_norm": 0.7202581055346592,
+      "grad_norm": 0.5942361162546903,
       "learning_rate": 5e-06,
-      "loss": 0.6701,
+      "loss": 0.6679,
       "step": 270
     },
     {
       "epoch": 0.687960687960688,
-      "grad_norm": 0.8023050470925295,
+      "grad_norm": 0.7411082671948905,
       "learning_rate": 5e-06,
-      "loss": 0.6617,
+      "loss": 0.66,
       "step": 280
     },
     {
       "epoch": 0.7125307125307125,
-      "grad_norm": 0.5163072877837536,
+      "grad_norm": 0.5402865637174907,
       "learning_rate": 5e-06,
-      "loss": 0.6577,
+      "loss": 0.656,
       "step": 290
     },
     {
       "epoch": 0.7371007371007371,
-      "grad_norm": 0.5823131054865358,
+      "grad_norm": 0.7940852714316793,
       "learning_rate": 5e-06,
-      "loss": 0.6577,
+      "loss": 0.6559,
       "step": 300
     },
     {
       "epoch": 0.7616707616707616,
-      "grad_norm": 0.5922229466469301,
+      "grad_norm": 0.4440401110733847,
       "learning_rate": 5e-06,
-      "loss": 0.6624,
+      "loss": 0.6609,
       "step": 310
     },
     {
       "epoch": 0.7862407862407862,
-      "grad_norm": 0.6006036496239568,
+      "grad_norm": 0.5659551226784064,
       "learning_rate": 5e-06,
-      "loss": 0.6511,
+      "loss": 0.6493,
       "step": 320
     },
     {
       "epoch": 0.8108108108108109,
-      "grad_norm": 0.6533699465325359,
+      "grad_norm": 0.5872037056746705,
       "learning_rate": 5e-06,
-      "loss": 0.6553,
+      "loss": 0.6537,
       "step": 330
     },
     {
       "epoch": 0.8353808353808354,
-      "grad_norm": 0.6882626303478473,
+      "grad_norm": 0.5670870147361556,
       "learning_rate": 5e-06,
-      "loss": 0.6536,
+      "loss": 0.6521,
       "step": 340
     },
     {
       "epoch": 0.85995085995086,
-      "grad_norm": 0.5251717254768736,
+      "grad_norm": 0.45655325491327986,
       "learning_rate": 5e-06,
-      "loss": 0.6546,
+      "loss": 0.6532,
       "step": 350
     },
     {
       "epoch": 0.8845208845208845,
-      "grad_norm": 0.5401311183676583,
+      "grad_norm": 0.4447712119508118,
       "learning_rate": 5e-06,
-      "loss": 0.6549,
+      "loss": 0.6535,
       "step": 360
     },
     {
       "epoch": 0.9090909090909091,
-      "grad_norm": 0.5423635183965629,
+      "grad_norm": 0.49332039079874884,
       "learning_rate": 5e-06,
-      "loss": 0.6558,
+      "loss": 0.6544,
       "step": 370
     },
     {
       "epoch": 0.9336609336609336,
-      "grad_norm": 0.4976652089187356,
+      "grad_norm": 0.4881054319905847,
       "learning_rate": 5e-06,
-      "loss": 0.6524,
+      "loss": 0.651,
       "step": 380
     },
     {
       "epoch": 0.9582309582309583,
-      "grad_norm": 0.8115237975091089,
+      "grad_norm": 0.5876239994992207,
       "learning_rate": 5e-06,
-      "loss": 0.6573,
+      "loss": 0.6559,
       "step": 390
     },
     {
       "epoch": 0.9828009828009828,
-      "grad_norm": 0.5179639159222823,
+      "grad_norm": 0.5399792337431831,
       "learning_rate": 5e-06,
-      "loss": 0.6513,
+      "loss": 0.65,
       "step": 400
     },
     {
       "epoch": 1.0,
-      "eval_loss": 0.6469405889511108,
-      "eval_runtime": 219.8729,
-      "eval_samples_per_second": 49.856,
-      "eval_steps_per_second": 0.391,
+      "eval_loss": 0.6456817984580994,
+      "eval_runtime": 217.9368,
+      "eval_samples_per_second": 50.299,
+      "eval_steps_per_second": 0.395,
       "step": 407
     },
     {
       "epoch": 1.0073710073710074,
-      "grad_norm": 1.133375985719492,
+      "grad_norm": 0.9036800627518505,
       "learning_rate": 5e-06,
-      "loss": 0.6392,
+      "loss": 0.6391,
       "step": 410
     },
     {
       "epoch": 1.031941031941032,
-      "grad_norm": 0.5662644763508548,
+      "grad_norm": 0.5211566701261678,
       "learning_rate": 5e-06,
-      "loss": 0.6049,
+      "loss": 0.6082,
       "step": 420
     },
     {
       "epoch": 1.0565110565110565,
-      "grad_norm": 0.6059886351419264,
+      "grad_norm": 0.49201394538696464,
       "learning_rate": 5e-06,
-      "loss": 0.6009,
+      "loss": 0.6047,
       "step": 430
     },
     {
       "epoch": 1.0810810810810811,
-      "grad_norm": 0.5539569535460708,
+      "grad_norm": 0.5230677761506219,
       "learning_rate": 5e-06,
-      "loss": 0.6014,
+      "loss": 0.605,
       "step": 440
     },
     {
       "epoch": 1.1056511056511056,
-      "grad_norm": 0.5322655590369197,
+      "grad_norm": 0.5061353049548358,
       "learning_rate": 5e-06,
-      "loss": 0.6081,
+      "loss": 0.6117,
       "step": 450
     },
     {
       "epoch": 1.1302211302211302,
-      "grad_norm": 0.5043434437693305,
+      "grad_norm": 0.5308414282748481,
       "learning_rate": 5e-06,
-      "loss": 0.6001,
+      "loss": 0.6038,
       "step": 460
     },
     {
       "epoch": 1.154791154791155,
-      "grad_norm": 0.5717057363415977,
+      "grad_norm": 0.5669662460358014,
       "learning_rate": 5e-06,
-      "loss": 0.6014,
+      "loss": 0.6046,
       "step": 470
     },
     {
       "epoch": 1.1793611793611793,
-      "grad_norm": 0.5403530749972654,
+      "grad_norm": 0.4422388402451124,
       "learning_rate": 5e-06,
-      "loss": 0.6057,
+      "loss": 0.6094,
       "step": 480
     },
     {
       "epoch": 1.203931203931204,
-      "grad_norm": 0.658623685891949,
+      "grad_norm": 0.6204673536874101,
       "learning_rate": 5e-06,
-      "loss": 0.6067,
+      "loss": 0.6098,
       "step": 490
     },
     {
       "epoch": 1.2285012285012284,
-      "grad_norm": 0.6157017033677983,
+      "grad_norm": 0.5083801522633119,
       "learning_rate": 5e-06,
-      "loss": 0.6023,
+      "loss": 0.6054,
       "step": 500
     },
     {
       "epoch": 1.253071253071253,
-      "grad_norm": 0.6102954603718583,
+      "grad_norm": 0.6670278402124389,
       "learning_rate": 5e-06,
-      "loss": 0.6085,
+      "loss": 0.6119,
       "step": 510
     },
     {
       "epoch": 1.2776412776412776,
-      "grad_norm": 0.5505779133780798,
+      "grad_norm": 0.4939088110674659,
       "learning_rate": 5e-06,
-      "loss": 0.6133,
+      "loss": 0.6165,
       "step": 520
     },
     {
       "epoch": 1.3022113022113022,
-      "grad_norm": 0.6781032596462336,
+      "grad_norm": 0.560793275781157,
       "learning_rate": 5e-06,
-      "loss": 0.6012,
+      "loss": 0.6045,
       "step": 530
     },
     {
       "epoch": 1.3267813267813269,
-      "grad_norm": 0.8958613423952794,
+      "grad_norm": 0.7550429966767501,
       "learning_rate": 5e-06,
-      "loss": 0.6099,
+      "loss": 0.6131,
       "step": 540
     },
     {
       "epoch": 1.3513513513513513,
-      "grad_norm": 0.5921228562589916,
+      "grad_norm": 0.5972489241212914,
       "learning_rate": 5e-06,
-      "loss": 0.6053,
+      "loss": 0.6086,
       "step": 550
     },
     {
       "epoch": 1.375921375921376,
-      "grad_norm": 0.6350136073582923,
+      "grad_norm": 0.5508921516494931,
       "learning_rate": 5e-06,
-      "loss": 0.6058,
+      "loss": 0.6093,
       "step": 560
     },
     {
       "epoch": 1.4004914004914004,
-      "grad_norm": 0.551237787133293,
+      "grad_norm": 0.5721594827498228,
       "learning_rate": 5e-06,
-      "loss": 0.6052,
+      "loss": 0.6083,
       "step": 570
     },
     {
       "epoch": 1.425061425061425,
-      "grad_norm": 0.5121185284640027,
+      "grad_norm": 0.44567791335112333,
       "learning_rate": 5e-06,
-      "loss": 0.6058,
+      "loss": 0.6088,
       "step": 580
     },
     {
       "epoch": 1.4496314496314495,
-      "grad_norm": 0.572102367341818,
+      "grad_norm": 0.4714094356929579,
       "learning_rate": 5e-06,
-      "loss": 0.6049,
+      "loss": 0.608,
       "step": 590
     },
     {
       "epoch": 1.4742014742014742,
-      "grad_norm": 0.49778553563528477,
+      "grad_norm": 0.45010217795717233,
       "learning_rate": 5e-06,
-      "loss": 0.5995,
+      "loss": 0.6023,
       "step": 600
     },
     {
       "epoch": 1.4987714987714988,
-      "grad_norm": 0.5039385054824482,
+      "grad_norm": 0.4507933071597356,
       "learning_rate": 5e-06,
-      "loss": 0.608,
+      "loss": 0.6111,
       "step": 610
     },
     {
       "epoch": 1.5233415233415233,
-      "grad_norm": 0.5349681092470842,
+      "grad_norm": 0.5050862616941423,
       "learning_rate": 5e-06,
-      "loss": 0.6063,
+      "loss": 0.6093,
       "step": 620
     },
     {
       "epoch": 1.547911547911548,
-      "grad_norm": 0.5543225135434735,
+      "grad_norm": 0.4815368134390591,
       "learning_rate": 5e-06,
-      "loss": 0.6087,
+      "loss": 0.612,
       "step": 630
     },
     {
       "epoch": 1.5724815724815726,
-      "grad_norm": 0.5233855996883069,
+      "grad_norm": 0.4550986333720554,
       "learning_rate": 5e-06,
-      "loss": 0.6105,
+      "loss": 0.6134,
       "step": 640
     },
     {
       "epoch": 1.597051597051597,
-      "grad_norm": 0.5328844425719327,
+      "grad_norm": 0.49348414741752084,
       "learning_rate": 5e-06,
-      "loss": 0.6095,
+      "loss": 0.6126,
       "step": 650
     },
     {
       "epoch": 1.6216216216216215,
-      "grad_norm": 0.5229727655454878,
+      "grad_norm": 0.46259072485753394,
       "learning_rate": 5e-06,
-      "loss": 0.6069,
+      "loss": 0.6102,
       "step": 660
     },
     {
       "epoch": 1.6461916461916462,
-      "grad_norm": 0.4961791614318175,
+      "grad_norm": 0.4489893330738159,
       "learning_rate": 5e-06,
-      "loss": 0.6028,
+      "loss": 0.6059,
       "step": 670
     },
     {
       "epoch": 1.6707616707616708,
-      "grad_norm": 0.5603987512874191,
+      "grad_norm": 0.532020034080249,
       "learning_rate": 5e-06,
-      "loss": 0.5994,
+      "loss": 0.6021,
       "step": 680
     },
     {
       "epoch": 1.6953316953316953,
-      "grad_norm": 0.5860389068572665,
+      "grad_norm": 0.5291299108019797,
       "learning_rate": 5e-06,
-      "loss": 0.6066,
+      "loss": 0.6092,
       "step": 690
     },
     {
       "epoch": 1.71990171990172,
-      "grad_norm": 0.5150718455508834,
+      "grad_norm": 0.4977355779144285,
       "learning_rate": 5e-06,
-      "loss": 0.6078,
+      "loss": 0.6105,
       "step": 700
     },
     {
       "epoch": 1.7444717444717446,
-      "grad_norm": 0.5145664220876448,
+      "grad_norm": 0.422516384063118,
       "learning_rate": 5e-06,
-      "loss": 0.6036,
+      "loss": 0.6066,
       "step": 710
     },
     {
       "epoch": 1.769041769041769,
-      "grad_norm": 0.5321485867383878,
+      "grad_norm": 0.5230305107651955,
       "learning_rate": 5e-06,
-      "loss": 0.6128,
+      "loss": 0.616,
       "step": 720
     },
     {
       "epoch": 1.7936117936117935,
-      "grad_norm": 0.6017429292341754,
+      "grad_norm": 0.6480321772002665,
       "learning_rate": 5e-06,
-      "loss": 0.6029,
+      "loss": 0.6053,
       "step": 730
     },
     {
       "epoch": 1.8181818181818183,
-      "grad_norm": 0.5261140082202593,
+      "grad_norm": 0.46981041595278134,
       "learning_rate": 5e-06,
-      "loss": 0.6119,
+      "loss": 0.6147,
       "step": 740
     },
     {
       "epoch": 1.8427518427518428,
-      "grad_norm": 0.568154347478099,
+      "grad_norm": 0.5219860110928298,
       "learning_rate": 5e-06,
-      "loss": 0.6034,
+      "loss": 0.6061,
       "step": 750
     },
     {
       "epoch": 1.8673218673218672,
-      "grad_norm": 0.7163771337990886,
+      "grad_norm": 0.6090315148152918,
       "learning_rate": 5e-06,
-      "loss": 0.6035,
+      "loss": 0.6062,
       "step": 760
     },
     {
       "epoch": 1.8918918918918919,
-      "grad_norm": 0.49318241144283287,
+      "grad_norm": 0.45264061321802845,
       "learning_rate": 5e-06,
-      "loss": 0.6056,
+      "loss": 0.6086,
       "step": 770
     },
     {
       "epoch": 1.9164619164619165,
-      "grad_norm": 0.5300451609828752,
+      "grad_norm": 0.4862492186635227,
       "learning_rate": 5e-06,
-      "loss": 0.6034,
+      "loss": 0.6062,
       "step": 780
     },
     {
       "epoch": 1.941031941031941,
-      "grad_norm": 0.5583192342180996,
+      "grad_norm": 0.4262515523099187,
       "learning_rate": 5e-06,
-      "loss": 0.6063,
+      "loss": 0.609,
       "step": 790
     },
     {
       "epoch": 1.9656019656019657,
-      "grad_norm": 0.6078304009681306,
+      "grad_norm": 0.540857759125553,
       "learning_rate": 5e-06,
-      "loss": 0.5979,
+      "loss": 0.6003,
       "step": 800
     },
     {
       "epoch": 1.9901719901719903,
-      "grad_norm": 0.5602383199809017,
+      "grad_norm": 0.5033832069244489,
       "learning_rate": 5e-06,
-      "loss": 0.6048,
+      "loss": 0.6074,
       "step": 810
     },
     {
       "epoch": 2.0,
-      "eval_loss": 0.6370313763618469,
-      "eval_runtime": 219.5312,
-      "eval_samples_per_second": 49.934,
-      "eval_steps_per_second": 0.392,
+      "eval_loss": 0.6354637742042542,
+      "eval_runtime": 218.2757,
+      "eval_samples_per_second": 50.221,
+      "eval_steps_per_second": 0.394,
       "step": 814
     },
     {
       "epoch": 2.0147420147420148,
-      "grad_norm": 0.7747334433956021,
+      "grad_norm": 0.7307156038164634,
       "learning_rate": 5e-06,
-      "loss": 0.5762,
+      "loss": 0.5822,
       "step": 820
     },
     {
       "epoch": 2.039312039312039,
-      "grad_norm": 0.5870617062342867,
+      "grad_norm": 0.5232318359619684,
       "learning_rate": 5e-06,
-      "loss": 0.5581,
+      "loss": 0.5664,
       "step": 830
     },
     {
       "epoch": 2.063882063882064,
-      "grad_norm": 0.5793099927447433,
+      "grad_norm": 0.5745655618201481,
       "learning_rate": 5e-06,
-      "loss": 0.5542,
+      "loss": 0.5624,
       "step": 840
     },
     {
       "epoch": 2.0884520884520885,
-      "grad_norm": 0.5392551354617476,
+      "grad_norm": 0.4968769160797791,
       "learning_rate": 5e-06,
-      "loss": 0.5526,
+      "loss": 0.5606,
       "step": 850
     },
     {
       "epoch": 2.113022113022113,
-      "grad_norm": 0.5892486652162464,
+      "grad_norm": 0.5327370268989237,
       "learning_rate": 5e-06,
-      "loss": 0.556,
+      "loss": 0.5643,
       "step": 860
     },
     {
       "epoch": 2.1375921375921374,
-      "grad_norm": 0.5537768340448994,
+      "grad_norm": 0.561176488765928,
       "learning_rate": 5e-06,
-      "loss": 0.5503,
+      "loss": 0.5582,
       "step": 870
     },
     {
       "epoch": 2.1621621621621623,
-      "grad_norm": 0.6064695378614886,
+      "grad_norm": 0.5155315136467447,
       "learning_rate": 5e-06,
-      "loss": 0.5501,
+      "loss": 0.5579,
       "step": 880
     },
     {
       "epoch": 2.1867321867321867,
-      "grad_norm": 0.5839237852242206,
+      "grad_norm": 0.5150732586456199,
       "learning_rate": 5e-06,
-      "loss": 0.5603,
+      "loss": 0.5687,
       "step": 890
     },
     {
       "epoch": 2.211302211302211,
-      "grad_norm": 0.4817575055188907,
+      "grad_norm": 0.44289319818662953,
       "learning_rate": 5e-06,
-      "loss": 0.5513,
+      "loss": 0.559,
       "step": 900
     },
     {
       "epoch": 2.235872235872236,
-      "grad_norm": 0.5601962025930167,
+      "grad_norm": 0.4685578082551758,
       "learning_rate": 5e-06,
-      "loss": 0.5625,
+      "loss": 0.5702,
       "step": 910
     },
     {
       "epoch": 2.2604422604422605,
-      "grad_norm": 0.606649078810016,
+      "grad_norm": 0.5785484819826471,
       "learning_rate": 5e-06,
-      "loss": 0.5669,
+      "loss": 0.5748,
       "step": 920
     },
     {
       "epoch": 2.285012285012285,
-      "grad_norm": 0.6223780920860711,
+      "grad_norm": 0.5830747436786318,
       "learning_rate": 5e-06,
-      "loss": 0.5637,
+      "loss": 0.5714,
       "step": 930
     },
     {
       "epoch": 2.30958230958231,
-      "grad_norm": 0.5969281028682913,
+      "grad_norm": 0.6088954402907469,
       "learning_rate": 5e-06,
-      "loss": 0.5625,
+      "loss": 0.5699,
       "step": 940
     },
     {
       "epoch": 2.3341523341523343,
-      "grad_norm": 0.6041411635437814,
+      "grad_norm": 0.5065740936571793,
       "learning_rate": 5e-06,
-      "loss": 0.5626,
+      "loss": 0.57,
       "step": 950
     },
     {
       "epoch": 2.3587223587223587,
-      "grad_norm": 0.540177728950405,
+      "grad_norm": 0.4385965614068684,
       "learning_rate": 5e-06,
-      "loss": 0.5623,
+      "loss": 0.5698,
       "step": 960
     },
     {
       "epoch": 2.383292383292383,
-      "grad_norm": 0.5972894547360773,
+      "grad_norm": 0.5105642133338127,
       "learning_rate": 5e-06,
-      "loss": 0.5584,
+      "loss": 0.5658,
       "step": 970
     },
     {
       "epoch": 2.407862407862408,
-      "grad_norm": 0.5077820328697467,
+      "grad_norm": 0.44745755786742697,
       "learning_rate": 5e-06,
-      "loss": 0.5566,
+      "loss": 0.5638,
       "step": 980
     },
     {
       "epoch": 2.4324324324324325,
-      "grad_norm": 0.5684297480394552,
+      "grad_norm": 0.4744160372716773,
       "learning_rate": 5e-06,
-      "loss": 0.5605,
+      "loss": 0.5676,
       "step": 990
     },
     {
       "epoch": 2.457002457002457,
-      "grad_norm": 0.5653223780605167,
+      "grad_norm": 0.5308528263789132,
       "learning_rate": 5e-06,
-      "loss": 0.5589,
+      "loss": 0.5658,
       "step": 1000
     },
     {
       "epoch": 2.4815724815724813,
-      "grad_norm": 0.5974364658021912,
+      "grad_norm": 0.5707200384455066,
       "learning_rate": 5e-06,
-      "loss": 0.5662,
+      "loss": 0.5734,
       "step": 1010
     },
     {
       "epoch": 2.506142506142506,
-      "grad_norm": 0.5968576754173772,
+      "grad_norm": 0.5519286283869067,
       "learning_rate": 5e-06,
-      "loss": 0.5583,
+      "loss": 0.5654,
       "step": 1020
     },
     {
       "epoch": 2.5307125307125307,
-      "grad_norm": 0.5404588062522825,
+      "grad_norm": 0.4588518825063357,
       "learning_rate": 5e-06,
-      "loss": 0.5648,
+      "loss": 0.572,
       "step": 1030
     },
     {
       "epoch": 2.555282555282555,
-      "grad_norm": 0.5434813520512509,
+      "grad_norm": 0.4919724825790031,
       "learning_rate": 5e-06,
-      "loss": 0.5574,
+      "loss": 0.5646,
       "step": 1040
     },
     {
       "epoch": 2.57985257985258,
-      "grad_norm": 0.5674865141570866,
+      "grad_norm": 0.4839147831831718,
       "learning_rate": 5e-06,
-      "loss": 0.5618,
+      "loss": 0.5686,
       "step": 1050
     },
     {
       "epoch": 2.6044226044226044,
-      "grad_norm": 0.5322758462790178,
+      "grad_norm": 0.4855007349272119,
       "learning_rate": 5e-06,
-      "loss": 0.5639,
+      "loss": 0.5711,
       "step": 1060
     },
     {
       "epoch": 2.628992628992629,
-      "grad_norm": 0.5577388218869722,
+      "grad_norm": 0.4652801035899528,
       "learning_rate": 5e-06,
-      "loss": 0.5585,
+      "loss": 0.5655,
       "step": 1070
     },
     {
       "epoch": 2.6535626535626538,
-      "grad_norm": 0.547813203437415,
+      "grad_norm": 0.4816575537658345,
       "learning_rate": 5e-06,
-      "loss": 0.5603,
+      "loss": 0.5672,
       "step": 1080
     },
     {
       "epoch": 2.678132678132678,
-      "grad_norm": 0.5519720050717444,
+      "grad_norm": 0.4924974787000159,
       "learning_rate": 5e-06,
-      "loss": 0.5649,
+      "loss": 0.5717,
       "step": 1090
     },
     {
       "epoch": 2.7027027027027026,
-      "grad_norm": 0.6068150251362087,
+      "grad_norm": 0.46678604918391425,
       "learning_rate": 5e-06,
-      "loss": 0.5693,
+      "loss": 0.5761,
       "step": 1100
     },
     {
       "epoch": 2.7272727272727275,
-      "grad_norm": 0.5733078031441013,
+      "grad_norm": 0.4815650588760393,
       "learning_rate": 5e-06,
-      "loss": 0.553,
+      "loss": 0.5592,
       "step": 1110
     },
     {
       "epoch": 2.751842751842752,
-      "grad_norm": 0.5261275498168718,
+      "grad_norm": 0.451576443537888,
       "learning_rate": 5e-06,
-      "loss": 0.5649,
+      "loss": 0.5714,
       "step": 1120
     },
     {
       "epoch": 2.7764127764127764,
-      "grad_norm": 0.5106351805286377,
+      "grad_norm": 0.49058626754375095,
       "learning_rate": 5e-06,
-      "loss": 0.5574,
+      "loss": 0.564,
       "step": 1130
     },
     {
       "epoch": 2.800982800982801,
-      "grad_norm": 0.6930658633743803,
+      "grad_norm": 0.5407476269825076,
       "learning_rate": 5e-06,
-      "loss": 0.5649,
+      "loss": 0.5715,
       "step": 1140
     },
     {
       "epoch": 2.8255528255528253,
-      "grad_norm": 0.5845919218726604,
+      "grad_norm": 0.5145580657005562,
       "learning_rate": 5e-06,
-      "loss": 0.5645,
+      "loss": 0.5713,
       "step": 1150
     },
     {
       "epoch": 2.85012285012285,
-      "grad_norm": 0.6569459469454167,
+      "grad_norm": 0.5608100121904279,
       "learning_rate": 5e-06,
-      "loss": 0.5579,
+      "loss": 0.5646,
       "step": 1160
     },
     {
       "epoch": 2.8746928746928746,
-      "grad_norm": 0.5569018454240479,
+      "grad_norm": 0.5060915278089009,
       "learning_rate": 5e-06,
-      "loss": 0.5617,
+      "loss": 0.5682,
       "step": 1170
     },
     {
       "epoch": 2.899262899262899,
-      "grad_norm": 0.7644913160268996,
+      "grad_norm": 0.6967844756678558,
       "learning_rate": 5e-06,
-      "loss": 0.5602,
+      "loss": 0.5666,
       "step": 1180
     },
     {
       "epoch": 2.923832923832924,
-      "grad_norm": 0.5508164107092175,
+      "grad_norm": 0.5006587440021466,
       "learning_rate": 5e-06,
-      "loss": 0.5632,
+      "loss": 0.5698,
       "step": 1190
     },
     {
       "epoch": 2.9484029484029484,
-      "grad_norm": 0.5499111308444584,
+      "grad_norm": 0.47107358105339453,
       "learning_rate": 5e-06,
-      "loss": 0.5641,
+      "loss": 0.5706,
       "step": 1200
     },
     {
       "epoch": 2.972972972972973,
-      "grad_norm": 0.5038444819979239,
+      "grad_norm": 0.46151336241822494,
       "learning_rate": 5e-06,
-      "loss": 0.5597,
+      "loss": 0.5665,
       "step": 1210
     },
     {
       "epoch": 2.9975429975429977,
-      "grad_norm": 0.5642492029878866,
+      "grad_norm": 0.5264094237400985,
       "learning_rate": 5e-06,
-      "loss": 0.5545,
+      "loss": 0.561,
       "step": 1220
     },
     {
       "epoch": 3.0,
-      "eval_loss": 0.6411369442939758,
-      "eval_runtime": 220.2966,
-      "eval_samples_per_second": 49.76,
-      "eval_steps_per_second": 0.39,
+      "eval_loss": 0.6375428438186646,
+      "eval_runtime": 218.7931,
+      "eval_samples_per_second": 50.102,
+      "eval_steps_per_second": 0.393,
       "step": 1221
     },
     {
       "epoch": 3.0,
       "step": 1221,
       "total_flos": 2044801717370880.0,
-      "train_loss": 0.6174200418344977,
-      "train_runtime": 36857.6484,
-      "train_samples_per_second": 16.952,
+      "train_loss": 0.6189516479367608,
+      "train_runtime": 36570.9767,
+      "train_samples_per_second": 17.085,
       "train_steps_per_second": 0.033
     }
   ],