MHGanainy/8-clusters-balanced-lex-best-v2-4

all_results.json

@@ -1,13 +1,13 @@
 {
  "epoch": 1.0,
- "eval_loss": 1.9224168062210083,
- "eval_runtime": 63.692,
- "eval_samples_per_second": 15.591,
- "eval_steps_per_second": 1.963,
- "perplexity": 6.837463340656789,
+ "eval_loss": 1.922713041305542,
+ "eval_runtime": 63.7378,
+ "eval_samples_per_second": 15.579,
+ "eval_steps_per_second": 1.961,
+ "perplexity": 6.839489137228699,
  "total_flos": 1.0003539689472e+17,
- "train_loss": 2.061040549059463,
- "train_runtime": 1615.6801,
- "train_samples_per_second": 6.798,
- "train_steps_per_second": 3.399
+ "train_loss": 2.061430201075499,
+ "train_runtime": 1617.6118,
+ "train_samples_per_second": 6.79,
+ "train_steps_per_second": 3.395
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
  "epoch": 1.0,
- "eval_loss": 1.9224168062210083,
- "eval_runtime": 63.692,
- "eval_samples_per_second": 15.591,
- "eval_steps_per_second": 1.963,
- "perplexity": 6.837463340656789
+ "eval_loss": 1.922713041305542,
+ "eval_runtime": 63.7378,
+ "eval_samples_per_second": 15.579,
+ "eval_steps_per_second": 1.961,
+ "perplexity": 6.839489137228699
 }

train_results.json

@@ -1,8 +1,8 @@
 {
  "epoch": 1.0,
  "total_flos": 1.0003539689472e+17,
- "train_loss": 2.061040549059463,
- "train_runtime": 1615.6801,
- "train_samples_per_second": 6.798,
- "train_steps_per_second": 3.399
+ "train_loss": 2.061430201075499,
+ "train_runtime": 1617.6118,
+ "train_samples_per_second": 6.79,
+ "train_steps_per_second": 3.395
 }

trainer_state.json

@@ -10,398 +10,398 @@
  "log_history": [
  {
  "epoch": 0.01820830298616169,
- "grad_norm": 0.22019147872924805,
+ "grad_norm": 0.22140225768089294,
  "learning_rate": 3.642987249544627e-06,
  "loss": 2.4398,
  "step": 100
  },
  {
  "epoch": 0.03641660597232338,
- "grad_norm": 0.1777569204568863,
+ "grad_norm": 0.17718623578548431,
  "learning_rate": 7.285974499089254e-06,
  "loss": 2.4058,
  "step": 200
  },
  {
  "epoch": 0.05462490895848507,
- "grad_norm": 0.2692304253578186,
+ "grad_norm": 0.26951104402542114,
  "learning_rate": 1.0928961748633882e-05,
  "loss": 2.3914,
  "step": 300
  },
  {
  "epoch": 0.07283321194464676,
- "grad_norm": 0.36335742473602295,
+ "grad_norm": 0.36239442229270935,
  "learning_rate": 1.4571948998178507e-05,
- "loss": 2.3211,
+ "loss": 2.3212,
  "step": 400
  },
  {
  "epoch": 0.09104151493080845,
- "grad_norm": 0.5098339319229126,
+ "grad_norm": 0.505577564239502,
  "learning_rate": 1.8214936247723133e-05,
- "loss": 2.2797,
+ "loss": 2.2799,
  "step": 500
  },
  {
  "epoch": 0.10924981791697014,
- "grad_norm": 0.5050227642059326,
+ "grad_norm": 0.5019211173057556,
  "learning_rate": 1.999474720010985e-05,
- "loss": 2.2523,
+ "loss": 2.2525,
  "step": 600
  },
  {
  "epoch": 0.12745812090313183,
- "grad_norm": 0.6238083839416504,
+ "grad_norm": 0.6163275837898254,
  "learning_rate": 1.9953983978532914e-05,
- "loss": 2.2214,
+ "loss": 2.2219,
  "step": 700
  },
  {
  "epoch": 0.14566642388929352,
- "grad_norm": 0.6451362371444702,
+ "grad_norm": 0.6447716355323792,
  "learning_rate": 1.987302601308333e-05,
- "loss": 2.1591,
+ "loss": 2.1597,
  "step": 800
  },
  {
  "epoch": 0.1638747268754552,
- "grad_norm": 0.6711136102676392,
+ "grad_norm": 0.6667017936706543,
  "learning_rate": 1.9752200216552278e-05,
- "loss": 2.1624,
+ "loss": 2.1629,
  "step": 900
  },
  {
  "epoch": 0.1820830298616169,
- "grad_norm": 0.7889479994773865,
+ "grad_norm": 0.7740277647972107,
  "learning_rate": 1.9591994490261997e-05,
- "loss": 2.0842,
+ "loss": 2.0851,
  "step": 1000
  },
  {
  "epoch": 0.20029133284777859,
- "grad_norm": 0.8510493040084839,
+ "grad_norm": 0.8541315197944641,
  "learning_rate": 1.9393055753893e-05,
- "loss": 2.1171,
+ "loss": 2.1176,
  "step": 1100
  },
  {
  "epoch": 0.21849963583394028,
- "grad_norm": 0.741717517375946,
+ "grad_norm": 0.7334938645362854,
  "learning_rate": 1.915618733318621e-05,
- "loss": 2.1071,
+ "loss": 2.1079,
  "step": 1200
  },
  {
  "epoch": 0.23670793882010197,
- "grad_norm": 0.7924832701683044,
+ "grad_norm": 0.7850305438041687,
  "learning_rate": 1.8882345716068708e-05,
- "loss": 2.0552,
+ "loss": 2.0558,
  "step": 1300
  },
  {
  "epoch": 0.25491624180626365,
- "grad_norm": 0.8211630582809448,
+ "grad_norm": 0.8258129358291626,
  "learning_rate": 1.8572636690301997e-05,
- "loss": 2.0649,
+ "loss": 2.0657,
  "step": 1400
  },
  {
  "epoch": 0.27312454479242537,
- "grad_norm": 0.7808334231376648,
+ "grad_norm": 0.7766979932785034,
  "learning_rate": 1.8228310878249212e-05,
- "loss": 2.0604,
+ "loss": 2.061,
  "step": 1500
  },
  {
  "epoch": 0.29133284777858703,
- "grad_norm": 1.0906625986099243,
+ "grad_norm": 1.0766857862472534,
  "learning_rate": 1.7850758686792054e-05,
- "loss": 2.08,
+ "loss": 2.0807,
  "step": 1600
  },
  {
  "epoch": 0.30954115076474875,
- "grad_norm": 0.7606624960899353,
+ "grad_norm": 0.7605693936347961,
  "learning_rate": 1.7441504692790104e-05,
- "loss": 2.0447,
+ "loss": 2.0452,
  "step": 1700
  },
  {
  "epoch": 0.3277494537509104,
- "grad_norm": 0.8482190370559692,
+ "grad_norm": 0.8467194437980652,
  "learning_rate": 1.700220148675417e-05,
- "loss": 2.0584,
+ "loss": 2.0589,
  "step": 1800
  },
  {
  "epoch": 0.34595775673707213,
- "grad_norm": 0.8170859813690186,
+ "grad_norm": 0.8118539452552795,
  "learning_rate": 1.6534622999593437e-05,
- "loss": 2.0788,
+ "loss": 2.0793,
  "step": 1900
  },
  {
  "epoch": 0.3641660597232338,
- "grad_norm": 0.9151561260223389,
+ "grad_norm": 0.9152652621269226,
  "learning_rate": 1.6040657339383255e-05,
- "loss": 2.0458,
+ "loss": 2.0462,
  "step": 2000
  },
  {
  "epoch": 0.3823743627093955,
- "grad_norm": 0.7852152585983276,
+ "grad_norm": 0.7831888794898987,
  "learning_rate": 1.5522299167079173e-05,
- "loss": 2.0271,
+ "loss": 2.0275,
  "step": 2100
  },
  {
  "epoch": 0.40058266569555717,
- "grad_norm": 1.1423839330673218,
+ "grad_norm": 1.125390887260437,
  "learning_rate": 1.4981641641964437e-05,
- "loss": 2.0153,
+ "loss": 2.0157,
  "step": 2200
  },
  {
  "epoch": 0.4187909686817189,
- "grad_norm": 0.9260895252227783,
+ "grad_norm": 0.9301549196243286,
  "learning_rate": 1.44208679693558e-05,
- "loss": 2.038,
+ "loss": 2.0384,
  "step": 2300
  },
  {
  "epoch": 0.43699927166788055,
- "grad_norm": 0.8138246536254883,
+ "grad_norm": 0.8163829445838928,
  "learning_rate": 1.384224258469838e-05,
- "loss": 2.0325,
+ "loss": 2.0331,
  "step": 2400
  },
  {
  "epoch": 0.45520757465404227,
- "grad_norm": 0.8293213844299316,
+ "grad_norm": 0.8261762261390686,
  "learning_rate": 1.3248102009648686e-05,
- "loss": 1.9852,
+ "loss": 1.9856,
  "step": 2500
  },
  {
  "epoch": 0.47341587764020393,
- "grad_norm": 1.0487293004989624,
+ "grad_norm": 1.0488537549972534,
  "learning_rate": 1.2640845417069571e-05,
- "loss": 2.0304,
+ "loss": 2.031,
  "step": 2600
  },
  {
  "epoch": 0.49162418062636565,
- "grad_norm": 1.4424030780792236,
+ "grad_norm": 1.400754451751709,
  "learning_rate": 1.2022924943036024e-05,
- "loss": 2.0349,
+ "loss": 2.0352,
  "step": 2700
  },
  {
  "epoch": 0.5098324836125273,
- "grad_norm": 0.930513322353363,
+ "grad_norm": 0.9314346313476562,
  "learning_rate": 1.139683578497262e-05,
- "loss": 2.0298,
+ "loss": 2.0303,
  "step": 2800
  },
  {
  "epoch": 0.528040786598689,
- "grad_norm": 1.3939330577850342,
+ "grad_norm": 1.420285701751709,
  "learning_rate": 1.0765106125906782e-05,
  "loss": 2.0071,
  "step": 2900
  },
  {
  "epoch": 0.5462490895848507,
- "grad_norm": 1.0249812602996826,
+ "grad_norm": 1.0269055366516113,
  "learning_rate": 1.0130286925524367e-05,
- "loss": 1.9692,
+ "loss": 1.9697,
  "step": 3000
  },
  {
  "epoch": 0.5644573925710124,
- "grad_norm": 0.9683905243873596,
+ "grad_norm": 0.9681551456451416,
  "learning_rate": 9.494941619251817e-06,
- "loss": 2.0243,
+ "loss": 2.0246,
  "step": 3100
  },
  {
  "epoch": 0.5826656955571741,
- "grad_norm": 1.1064997911453247,
+ "grad_norm": 1.1162570714950562,
  "learning_rate": 8.861635766960579e-06,
- "loss": 1.9983,
+ "loss": 1.9988,
  "step": 3200
  },
  {
  "epoch": 0.6008739985433358,
- "grad_norm": 0.9611035585403442,
+ "grad_norm": 0.9726856350898743,
  "learning_rate": 8.232926693092881e-06,
- "loss": 1.9898,
+ "loss": 1.9904,
  "step": 3300
  },
  {
  "epoch": 0.6190823015294975,
- "grad_norm": 1.2558187246322632,
+ "grad_norm": 1.2489055395126343,
  "learning_rate": 7.611353160042658e-06,
- "loss": 1.9698,
+ "loss": 1.9702,
  "step": 3400
  },
  {
  "epoch": 0.6372906045156591,
- "grad_norm": 1.1066709756851196,
+ "grad_norm": 1.1108543872833252,
  "learning_rate": 6.99942511649105e-06,
- "loss": 2.0468,
+ "loss": 2.0471,
  "step": 3500
  },
  {
  "epoch": 0.6554989075018208,
- "grad_norm": 1.0138076543807983,
+ "grad_norm": 1.0132871866226196,
  "learning_rate": 6.399613562093272e-06,
- "loss": 2.0535,
+ "loss": 2.0539,
  "step": 3600
  },
  {
  "epoch": 0.6737072104879825,
- "grad_norm": 1.125917673110962,
+ "grad_norm": 1.1130536794662476,
  "learning_rate": 5.814340569443867e-06,
- "loss": 2.009,
+ "loss": 2.0095,
  "step": 3700
  },
  {
  "epoch": 0.6919155134741443,
- "grad_norm": 0.7838294506072998,
+ "grad_norm": 0.7807123064994812,
  "learning_rate": 5.245969503612125e-06,
- "loss": 1.9229,
+ "loss": 1.9232,
  "step": 3800
  },
  {
  "epoch": 0.7101238164603059,
- "grad_norm": 1.051643967628479,
+ "grad_norm": 1.0500433444976807,
  "learning_rate": 4.696795478741786e-06,
- "loss": 1.9857,
+ "loss": 1.986,
  "step": 3900
  },
  {
  "epoch": 0.7283321194464676,
- "grad_norm": 0.9563839435577393,
+ "grad_norm": 0.9565054774284363,
  "learning_rate": 4.169036090251809e-06,
- "loss": 2.0503,
+ "loss": 2.0505,
  "step": 4000
  },
  {
  "epoch": 0.7465404224326293,
- "grad_norm": 0.8592619299888611,
+ "grad_norm": 0.8539314270019531,
  "learning_rate": 3.6648224600620653e-06,
- "loss": 2.0066,
+ "loss": 2.0069,
  "step": 4100
  },
  {
  "epoch": 0.764748725418791,
- "grad_norm": 0.9203991293907166,
+ "grad_norm": 0.9163336157798767,
  "learning_rate": 3.1861906310038825e-06,
- "loss": 1.9719,
+ "loss": 1.9723,
  "step": 4200
  },
  {
  "epoch": 0.7829570284049526,
- "grad_norm": 0.9755929112434387,
+ "grad_norm": 0.9746565222740173,
  "learning_rate": 2.735073345165228e-06,
- "loss": 1.9785,
+ "loss": 1.9787,
  "step": 4300
  },
  {
  "epoch": 0.8011653313911143,
- "grad_norm": 1.030912160873413,
+ "grad_norm": 1.0261231660842896,
  "learning_rate": 2.313292239370102e-06,
- "loss": 2.0122,
+ "loss": 2.0124,
  "step": 4400
  },
  {
  "epoch": 0.8193736343772761,
- "grad_norm": 1.2122974395751953,
+ "grad_norm": 1.207588791847229,
  "learning_rate": 1.9225504893071823e-06,
- "loss": 1.9747,
+ "loss": 1.9749,
  "step": 4500
  },
  {
  "epoch": 0.8375819373634378,
- "grad_norm": 1.0663396120071411,
+ "grad_norm": 1.0550912618637085,
  "learning_rate": 1.5644259320111733e-06,
- "loss": 1.9379,
+ "loss": 1.9382,
  "step": 4600
  },
  {
  "epoch": 0.8557902403495994,
- "grad_norm": 0.9615539908409119,
+ "grad_norm": 0.9657886624336243,
  "learning_rate": 1.2403646944686198e-06,
  "loss": 1.9893,
  "step": 4700
  },
  {
  "epoch": 0.8739985433357611,
- "grad_norm": 1.184589147567749,
+ "grad_norm": 1.1723967790603638,
  "learning_rate": 9.516753540762868e-07,
- "loss": 1.9812,
+ "loss": 1.9816,
  "step": 4800
  },
  {
  "epoch": 0.8922068463219228,
- "grad_norm": 1.2580232620239258,
+ "grad_norm": 1.271134376525879,
  "learning_rate": 6.995236545324624e-07,
- "loss": 1.948,
+ "loss": 1.9485,
  "step": 4900
  },
  {
  "epoch": 0.9104151493080845,
- "grad_norm": 0.958463728427887,
+ "grad_norm": 0.9425839185714722,
  "learning_rate": 4.849277984987221e-07,
- "loss": 1.9752,
+ "loss": 1.9754,
  "step": 5000
  },
  {
  "epoch": 0.9286234522942461,
- "grad_norm": 1.0817508697509766,
+ "grad_norm": 1.077772617340088,
  "learning_rate": 3.0875433604064976e-07,
- "loss": 2.0005,
+ "loss": 2.0007,
  "step": 5100
  },
  {
  "epoch": 0.9468317552804079,
- "grad_norm": 1.1332553625106812,
+ "grad_norm": 1.13583242893219,
  "learning_rate": 1.7171466545021665e-07,
- "loss": 1.9255,
+ "loss": 1.926,
  "step": 5200
  },
  {
  "epoch": 0.9650400582665696,
- "grad_norm": 0.9218988418579102,
+ "grad_norm": 0.9153927564620972,
  "learning_rate": 7.436216057970735e-08,
- "loss": 2.0052,
+ "loss": 2.0055,
  "step": 5300
  },
  {
  "epoch": 0.9832483612527313,
- "grad_norm": 1.2976170778274536,
+ "grad_norm": 1.2890238761901855,
  "learning_rate": 1.708993628716016e-08,
- "loss": 2.0214,
+ "loss": 2.0217,
  "step": 5400
  },
  {
  "epoch": 1.0,
- "eval_loss": 1.9224168062210083,
- "eval_runtime": 63.6773,
- "eval_samples_per_second": 15.594,
- "eval_steps_per_second": 1.963,
+ "eval_loss": 1.922713041305542,
+ "eval_runtime": 63.7429,
+ "eval_samples_per_second": 15.578,
+ "eval_steps_per_second": 1.961,
  "step": 5492
  },
  {
  "epoch": 1.0,
  "step": 5492,
  "total_flos": 1.0003539689472e+17,
- "train_loss": 2.061040549059463,
- "train_runtime": 1615.6801,
- "train_samples_per_second": 6.798,
- "train_steps_per_second": 3.399
+ "train_loss": 2.061430201075499,
+ "train_runtime": 1617.6118,
+ "train_samples_per_second": 6.79,
+ "train_steps_per_second": 3.395
  }
  ],
  "logging_steps": 100,