End of training

README.md

@@ -1,11 +1,23 @@
 ---
 tags:
 - generated_from_trainer
+datasets:
+- Hartunka/processed_wikitext-103-raw-v1-rand-100_v2
 metrics:
 - accuracy
 model-index:
 - name: tiny_bert_rand_100_v2
-  results: []
+  results:
+  - task:
+      name: Masked Language Modeling
+      type: fill-mask
+    dataset:
+      name: Hartunka/processed_wikitext-103-raw-v1-rand-100_v2
+      type: Hartunka/processed_wikitext-103-raw-v1-rand-100_v2
+    metrics:
+    - name: Accuracy
+      type: accuracy
+      value: 0.15234220489348477
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -13,10 +25,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # tiny_bert_rand_100_v2
 
-This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
+This model is a fine-tuned version of [](https://huggingface.co/) on the Hartunka/processed_wikitext-103-raw-v1-rand-100_v2 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 13.4746
-- Accuracy: 0.1518
+- Loss: 10.8907
+- Accuracy: 0.1523
 
 ## Model description
 

all_results.json

@@ -1,16 +1,16 @@
 {
     "epoch": 25.0,
-    "eval_accuracy": 0.15262473865626944,
-    "eval_loss": 10.893350601196289,
-    "eval_runtime": 0.9406,
+    "eval_accuracy": 0.15234220489348477,
+    "eval_loss": 10.890739440917969,
+    "eval_runtime": 0.752,
     "eval_samples": 479,
-    "eval_samples_per_second": 509.273,
-    "eval_steps_per_second": 5.316,
-    "perplexity": 53817.31858838507,
+    "eval_samples_per_second": 636.945,
+    "eval_steps_per_second": 6.649,
+    "perplexity": 53676.97625170849,
     "total_flos": 3.058354515764736e+17,
-    "train_loss": 9.151931135944716,
-    "train_runtime": 12834.3964,
+    "train_loss": 9.169899891581883,
+    "train_runtime": 12212.2898,
     "train_samples": 228639,
-    "train_samples_per_second": 445.364,
-    "train_steps_per_second": 4.64
+    "train_samples_per_second": 468.051,
+    "train_steps_per_second": 4.876
 }

eval_results.json

@@ -1,10 +1,10 @@
 {
     "epoch": 25.0,
-    "eval_accuracy": 0.15262473865626944,
-    "eval_loss": 10.893350601196289,
-    "eval_runtime": 0.9406,
+    "eval_accuracy": 0.15234220489348477,
+    "eval_loss": 10.890739440917969,
+    "eval_runtime": 0.752,
     "eval_samples": 479,
-    "eval_samples_per_second": 509.273,
-    "eval_steps_per_second": 5.316,
-    "perplexity": 53817.31858838507
+    "eval_samples_per_second": 636.945,
+    "eval_steps_per_second": 6.649,
+    "perplexity": 53676.97625170849
 }

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 25.0,
     "total_flos": 3.058354515764736e+17,
-    "train_loss": 9.151931135944716,
-    "train_runtime": 12834.3964,
+    "train_loss": 9.169899891581883,
+    "train_runtime": 12212.2898,
     "train_samples": 228639,
-    "train_samples_per_second": 445.364,
-    "train_steps_per_second": 4.64
+    "train_samples_per_second": 468.051,
+    "train_steps_per_second": 4.876
 }

trainer_state.json

@@ -1,6 +1,5 @@
 {
-  "best_global_step": 10000,
-  "best_metric": 10.8783597946167,
+  "best_metric": 10.877917289733887,
   "best_model_checkpoint": "tiny_bert_rand_100_v2/checkpoint-10000",
   "epoch": 25.0,
   "eval_steps": 10000,
@@ -11,890 +10,890 @@
   "log_history": [
     {
       "epoch": 0.20990764063811923,
-      "grad_norm": 2.5572643280029297,
+      "grad_norm": 2.6227288246154785,
       "learning_rate": 5e-06,
-      "loss": 14.3831,
+      "loss": 14.3832,
       "step": 500
     },
     {
       "epoch": 0.41981528127623846,
-      "grad_norm": 1.9127840995788574,
+      "grad_norm": 1.8764026165008545,
       "learning_rate": 1e-05,
-      "loss": 12.8302,
+      "loss": 12.8296,
       "step": 1000
     },
     {
       "epoch": 0.6297229219143576,
-      "grad_norm": 1.3850857019424438,
+      "grad_norm": 1.4561502933502197,
       "learning_rate": 1.5e-05,
-      "loss": 11.8293,
+      "loss": 11.8295,
       "step": 1500
     },
     {
       "epoch": 0.8396305625524769,
-      "grad_norm": 1.447935700416565,
+      "grad_norm": 1.6282973289489746,
       "learning_rate": 2e-05,
-      "loss": 11.6299,
+      "loss": 11.628,
       "step": 2000
     },
     {
       "epoch": 1.0495382031905962,
-      "grad_norm": 1.5670098066329956,
+      "grad_norm": 1.4469887018203735,
       "learning_rate": 2.5e-05,
-      "loss": 11.5164,
+      "loss": 11.5129,
       "step": 2500
     },
     {
       "epoch": 1.2594458438287153,
-      "grad_norm": 1.5325168371200562,
+      "grad_norm": 1.505755066871643,
       "learning_rate": 3e-05,
-      "loss": 11.4095,
+      "loss": 11.4056,
       "step": 3000
     },
     {
       "epoch": 1.4693534844668346,
-      "grad_norm": 1.6217561960220337,
+      "grad_norm": 1.7616183757781982,
       "learning_rate": 3.5e-05,
-      "loss": 11.3235,
+      "loss": 11.3201,
       "step": 3500
     },
     {
       "epoch": 1.6792611251049538,
-      "grad_norm": 1.616829752922058,
+      "grad_norm": 1.473453164100647,
       "learning_rate": 4e-05,
-      "loss": 11.2435,
+      "loss": 11.2415,
       "step": 4000
     },
     {
       "epoch": 1.8891687657430731,
-      "grad_norm": 1.6852329969406128,
+      "grad_norm": 1.6783679723739624,
       "learning_rate": 4.5e-05,
-      "loss": 11.1679,
+      "loss": 11.1662,
       "step": 4500
     },
     {
       "epoch": 2.0990764063811924,
-      "grad_norm": 1.6728254556655884,
+      "grad_norm": 1.814674973487854,
       "learning_rate": 5e-05,
-      "loss": 11.0995,
+      "loss": 11.0977,
       "step": 5000
     },
     {
       "epoch": 2.3089840470193117,
-      "grad_norm": 1.8060897588729858,
+      "grad_norm": 1.694793701171875,
       "learning_rate": 5.500000000000001e-05,
-      "loss": 11.0497,
+      "loss": 11.051,
       "step": 5500
     },
     {
       "epoch": 2.5188916876574305,
-      "grad_norm": 1.8170562982559204,
+      "grad_norm": 1.7509475946426392,
       "learning_rate": 6e-05,
-      "loss": 11.0197,
+      "loss": 11.0192,
       "step": 6000
     },
     {
       "epoch": 2.72879932829555,
-      "grad_norm": 1.642697811126709,
+      "grad_norm": 1.7445567846298218,
       "learning_rate": 6.500000000000001e-05,
-      "loss": 10.9904,
+      "loss": 10.9921,
       "step": 6500
     },
     {
       "epoch": 2.938706968933669,
-      "grad_norm": 1.6969773769378662,
+      "grad_norm": 1.6910563707351685,
       "learning_rate": 7e-05,
-      "loss": 10.9666,
+      "loss": 10.9693,
       "step": 7000
     },
     {
       "epoch": 3.1486146095717884,
-      "grad_norm": 1.8204708099365234,
+      "grad_norm": 1.7993345260620117,
       "learning_rate": 7.500000000000001e-05,
-      "loss": 10.8992,
+      "loss": 10.9022,
       "step": 7500
     },
     {
       "epoch": 3.3585222502099077,
-      "grad_norm": 2.0726945400238037,
+      "grad_norm": 1.9512734413146973,
       "learning_rate": 8e-05,
-      "loss": 10.8626,
+      "loss": 10.8689,
       "step": 8000
     },
     {
       "epoch": 3.568429890848027,
-      "grad_norm": 2.0214316844940186,
+      "grad_norm": 2.025214672088623,
       "learning_rate": 8.5e-05,
-      "loss": 10.8541,
+      "loss": 10.8571,
       "step": 8500
     },
     {
       "epoch": 3.7783375314861463,
-      "grad_norm": 2.1393232345581055,
+      "grad_norm": 2.087209463119507,
       "learning_rate": 9e-05,
-      "loss": 10.8306,
+      "loss": 10.8382,
       "step": 9000
     },
     {
       "epoch": 3.988245172124265,
-      "grad_norm": 2.114140033721924,
+      "grad_norm": 2.030062198638916,
       "learning_rate": 9.5e-05,
-      "loss": 10.8188,
+      "loss": 10.821,
       "step": 9500
     },
     {
       "epoch": 4.198152812762385,
-      "grad_norm": 2.4008498191833496,
+      "grad_norm": 2.390781879425049,
       "learning_rate": 0.0001,
-      "loss": 10.6718,
+      "loss": 10.6784,
       "step": 10000
     },
     {
       "epoch": 4.198152812762385,
-      "eval_accuracy": 0.15022677314338184,
-      "eval_loss": 10.8783597946167,
-      "eval_runtime": 0.7987,
-      "eval_samples_per_second": 599.714,
-      "eval_steps_per_second": 6.26,
+      "eval_accuracy": 0.1500876485155402,
+      "eval_loss": 10.877917289733887,
+      "eval_runtime": 0.7308,
+      "eval_samples_per_second": 655.449,
+      "eval_steps_per_second": 6.842,
       "step": 10000
     },
     {
       "epoch": 4.408060453400504,
-      "grad_norm": 2.3720543384552,
+      "grad_norm": 2.3683950901031494,
       "learning_rate": 9.899091826437942e-05,
-      "loss": 10.6629,
+      "loss": 10.6692,
       "step": 10500
     },
     {
       "epoch": 4.617968094038623,
-      "grad_norm": 2.865995168685913,
+      "grad_norm": 2.8667867183685303,
       "learning_rate": 9.798183652875883e-05,
-      "loss": 10.6534,
+      "loss": 10.6567,
       "step": 11000
     },
     {
       "epoch": 4.827875734676742,
-      "grad_norm": 2.6649460792541504,
+      "grad_norm": 2.78804874420166,
       "learning_rate": 9.697275479313825e-05,
-      "loss": 10.6545,
+      "loss": 10.6613,
       "step": 11500
     },
     {
       "epoch": 5.037783375314861,
-      "grad_norm": 2.761911153793335,
+      "grad_norm": 2.823746681213379,
       "learning_rate": 9.596367305751767e-05,
-      "loss": 10.6098,
+      "loss": 10.6069,
       "step": 12000
     },
     {
       "epoch": 5.247691015952981,
-      "grad_norm": 3.323977470397949,
+      "grad_norm": 3.034684419631958,
       "learning_rate": 9.495459132189707e-05,
-      "loss": 10.4387,
+      "loss": 10.442,
       "step": 12500
     },
     {
       "epoch": 5.4575986565911,
-      "grad_norm": 2.9472594261169434,
+      "grad_norm": 2.8574166297912598,
       "learning_rate": 9.394550958627649e-05,
-      "loss": 10.4492,
+      "loss": 10.4477,
       "step": 13000
     },
     {
       "epoch": 5.667506297229219,
-      "grad_norm": 3.004497766494751,
+      "grad_norm": 3.1504855155944824,
       "learning_rate": 9.293642785065591e-05,
-      "loss": 10.465,
+      "loss": 10.4651,
       "step": 13500
     },
     {
       "epoch": 5.877413937867338,
-      "grad_norm": 3.239577293395996,
+      "grad_norm": 3.020800828933716,
       "learning_rate": 9.192734611503532e-05,
-      "loss": 10.4677,
+      "loss": 10.4799,
       "step": 14000
     },
     {
       "epoch": 6.087321578505458,
-      "grad_norm": 2.9033613204956055,
+      "grad_norm": 3.036743640899658,
       "learning_rate": 9.091826437941474e-05,
-      "loss": 10.3753,
+      "loss": 10.3752,
       "step": 14500
     },
     {
       "epoch": 6.297229219143577,
-      "grad_norm": 3.0392324924468994,
+      "grad_norm": 3.0110838413238525,
       "learning_rate": 8.990918264379415e-05,
-      "loss": 10.2735,
+      "loss": 10.272,
       "step": 15000
     },
     {
       "epoch": 6.507136859781696,
-      "grad_norm": 3.534576177597046,
+      "grad_norm": 3.3376975059509277,
       "learning_rate": 8.890010090817356e-05,
-      "loss": 10.2972,
+      "loss": 10.2982,
       "step": 15500
     },
     {
       "epoch": 6.717044500419815,
-      "grad_norm": 3.014953136444092,
+      "grad_norm": 3.2656025886535645,
       "learning_rate": 8.789101917255298e-05,
-      "loss": 10.3233,
+      "loss": 10.3262,
       "step": 16000
     },
     {
       "epoch": 6.926952141057934,
-      "grad_norm": 3.170473098754883,
+      "grad_norm": 3.2516350746154785,
       "learning_rate": 8.68819374369324e-05,
-      "loss": 10.3444,
+      "loss": 10.3482,
       "step": 16500
     },
     {
       "epoch": 7.136859781696054,
-      "grad_norm": 3.1827826499938965,
+      "grad_norm": 3.389404535293579,
       "learning_rate": 8.58728557013118e-05,
-      "loss": 10.1951,
+      "loss": 10.1959,
       "step": 17000
     },
     {
       "epoch": 7.346767422334173,
-      "grad_norm": 3.821049928665161,
+      "grad_norm": 3.69683837890625,
       "learning_rate": 8.486377396569122e-05,
-      "loss": 10.1439,
+      "loss": 10.1418,
       "step": 17500
     },
     {
       "epoch": 7.5566750629722925,
-      "grad_norm": 3.252129554748535,
+      "grad_norm": 3.3357648849487305,
       "learning_rate": 8.385469223007064e-05,
-      "loss": 10.1822,
+      "loss": 10.1816,
       "step": 18000
     },
     {
       "epoch": 7.766582703610411,
-      "grad_norm": 3.332376480102539,
+      "grad_norm": 3.374206066131592,
       "learning_rate": 8.284561049445006e-05,
-      "loss": 10.2191,
+      "loss": 10.2132,
       "step": 18500
     },
     {
       "epoch": 7.97649034424853,
-      "grad_norm": 3.5262091159820557,
+      "grad_norm": 3.306196928024292,
       "learning_rate": 8.183652875882947e-05,
-      "loss": 10.2294,
+      "loss": 10.2229,
       "step": 19000
     },
     {
       "epoch": 8.18639798488665,
-      "grad_norm": 3.345799684524536,
+      "grad_norm": 3.6772420406341553,
       "learning_rate": 8.082744702320889e-05,
-      "loss": 10.0135,
+      "loss": 10.0083,
       "step": 19500
     },
     {
       "epoch": 8.39630562552477,
-      "grad_norm": 3.6675608158111572,
+      "grad_norm": 3.798260450363159,
       "learning_rate": 7.98183652875883e-05,
-      "loss": 10.0251,
+      "loss": 10.0232,
       "step": 20000
     },
     {
       "epoch": 8.39630562552477,
-      "eval_accuracy": 0.15292236033558743,
-      "eval_loss": 11.145699501037598,
-      "eval_runtime": 0.7129,
-      "eval_samples_per_second": 671.95,
-      "eval_steps_per_second": 7.014,
+      "eval_accuracy": 0.15295041948427285,
+      "eval_loss": 11.155789375305176,
+      "eval_runtime": 0.7248,
+      "eval_samples_per_second": 660.903,
+      "eval_steps_per_second": 6.899,
       "step": 20000
     },
     {
       "epoch": 8.606213266162888,
-      "grad_norm": 3.721390962600708,
+      "grad_norm": 3.7763068675994873,
       "learning_rate": 7.880928355196771e-05,
-      "loss": 10.0494,
+      "loss": 10.0465,
       "step": 20500
     },
     {
       "epoch": 8.816120906801007,
-      "grad_norm": 3.681361436843872,
+      "grad_norm": 3.530777931213379,
       "learning_rate": 7.780020181634713e-05,
-      "loss": 10.0923,
+      "loss": 10.0932,
       "step": 21000
     },
     {
       "epoch": 9.026028547439127,
-      "grad_norm": 3.9859206676483154,
+      "grad_norm": 3.9749209880828857,
       "learning_rate": 7.679112008072655e-05,
-      "loss": 10.0599,
+      "loss": 10.0535,
       "step": 21500
     },
     {
       "epoch": 9.235936188077247,
-      "grad_norm": 4.25050163269043,
+      "grad_norm": 4.1896843910217285,
       "learning_rate": 7.578203834510595e-05,
-      "loss": 9.8323,
+      "loss": 9.824,
       "step": 22000
     },
     {
       "epoch": 9.445843828715365,
-      "grad_norm": 4.300194263458252,
+      "grad_norm": 4.39830207824707,
       "learning_rate": 7.477295660948537e-05,
-      "loss": 9.8658,
+      "loss": 9.8569,
       "step": 22500
     },
     {
       "epoch": 9.655751469353484,
-      "grad_norm": 4.0486931800842285,
+      "grad_norm": 4.277210712432861,
       "learning_rate": 7.376387487386479e-05,
-      "loss": 9.9066,
+      "loss": 9.9004,
       "step": 23000
     },
     {
       "epoch": 9.865659109991604,
-      "grad_norm": 4.532042026519775,
+      "grad_norm": 4.306820392608643,
       "learning_rate": 7.27547931382442e-05,
-      "loss": 9.9182,
+      "loss": 9.9117,
       "step": 23500
     },
     {
       "epoch": 10.075566750629722,
-      "grad_norm": 4.684360027313232,
+      "grad_norm": 4.395482063293457,
       "learning_rate": 7.174571140262362e-05,
-      "loss": 9.8172,
+      "loss": 9.8127,
       "step": 24000
     },
     {
       "epoch": 10.285474391267842,
-      "grad_norm": 4.4530439376831055,
+      "grad_norm": 4.56743049621582,
       "learning_rate": 7.073662966700304e-05,
-      "loss": 9.6398,
+      "loss": 9.6301,
       "step": 24500
     },
     {
       "epoch": 10.495382031905962,
-      "grad_norm": 5.032015800476074,
+      "grad_norm": 4.907061576843262,
       "learning_rate": 6.972754793138244e-05,
-      "loss": 9.6586,
+      "loss": 9.6473,
       "step": 25000
     },
     {
       "epoch": 10.705289672544081,
-      "grad_norm": 4.901517868041992,
+      "grad_norm": 4.961501121520996,
       "learning_rate": 6.871846619576186e-05,
-      "loss": 9.6916,
+      "loss": 9.6806,
       "step": 25500
     },
     {
       "epoch": 10.9151973131822,
-      "grad_norm": 5.025380611419678,
+      "grad_norm": 5.273646354675293,
       "learning_rate": 6.770938446014128e-05,
-      "loss": 9.7161,
+      "loss": 9.7101,
       "step": 26000
     },
     {
       "epoch": 11.125104953820319,
-      "grad_norm": 5.682083606719971,
+      "grad_norm": 5.380168914794922,
       "learning_rate": 6.670030272452068e-05,
-      "loss": 9.5129,
+      "loss": 9.5095,
       "step": 26500
     },
     {
       "epoch": 11.335012594458439,
-      "grad_norm": 6.0306243896484375,
+      "grad_norm": 5.709550857543945,
       "learning_rate": 6.56912209889001e-05,
-      "loss": 9.3883,
+      "loss": 9.3969,
       "step": 27000
     },
     {
       "epoch": 11.544920235096557,
-      "grad_norm": 5.714005470275879,
+      "grad_norm": 5.763561248779297,
       "learning_rate": 6.468213925327952e-05,
-      "loss": 9.441,
+      "loss": 9.4286,
       "step": 27500
     },
     {
       "epoch": 11.754827875734676,
-      "grad_norm": 5.998770236968994,
+      "grad_norm": 5.631028175354004,
       "learning_rate": 6.367305751765893e-05,
-      "loss": 9.4526,
+      "loss": 9.4604,
       "step": 28000
     },
     {
       "epoch": 11.964735516372796,
-      "grad_norm": 5.728001117706299,
+      "grad_norm": 5.918311595916748,
       "learning_rate": 6.266397578203835e-05,
-      "loss": 9.4967,
+      "loss": 9.4802,
       "step": 28500
     },
     {
       "epoch": 12.174643157010916,
-      "grad_norm": 6.085269451141357,
+      "grad_norm": 6.19346809387207,
       "learning_rate": 6.165489404641777e-05,
-      "loss": 9.1696,
+      "loss": 9.173,
       "step": 29000
     },
     {
       "epoch": 12.384550797649034,
-      "grad_norm": 6.195425510406494,
+      "grad_norm": 6.1046061515808105,
       "learning_rate": 6.064581231079718e-05,
-      "loss": 9.1282,
+      "loss": 9.138,
       "step": 29500
     },
     {
       "epoch": 12.594458438287154,
-      "grad_norm": 6.641568183898926,
+      "grad_norm": 6.77022647857666,
       "learning_rate": 5.963673057517659e-05,
-      "loss": 9.1754,
+      "loss": 9.1752,
       "step": 30000
     },
     {
       "epoch": 12.594458438287154,
-      "eval_accuracy": 0.15297870554327447,
-      "eval_loss": 11.742791175842285,
-      "eval_runtime": 0.7274,
-      "eval_samples_per_second": 658.554,
-      "eval_steps_per_second": 6.874,
+      "eval_accuracy": 0.1537062428295604,
+      "eval_loss": 11.716165542602539,
+      "eval_runtime": 0.7187,
+      "eval_samples_per_second": 666.444,
+      "eval_steps_per_second": 6.957,
       "step": 30000
     },
     {
       "epoch": 12.804366078925273,
-      "grad_norm": 7.108124256134033,
+      "grad_norm": 6.768683910369873,
       "learning_rate": 5.8627648839556e-05,
-      "loss": 9.2082,
+      "loss": 9.208,
       "step": 30500
     },
     {
       "epoch": 13.014273719563391,
-      "grad_norm": 6.683660507202148,
+      "grad_norm": 7.0972208976745605,
       "learning_rate": 5.7618567103935416e-05,
-      "loss": 9.2022,
+      "loss": 9.1954,
       "step": 31000
     },
     {
       "epoch": 13.224181360201511,
-      "grad_norm": 7.303939342498779,
+      "grad_norm": 6.531957626342773,
       "learning_rate": 5.660948536831484e-05,
-      "loss": 8.8255,
+      "loss": 8.8417,
       "step": 31500
     },
     {
       "epoch": 13.43408900083963,
-      "grad_norm": 7.398545742034912,
+      "grad_norm": 7.456687927246094,
       "learning_rate": 5.5600403632694254e-05,
-      "loss": 8.869,
+      "loss": 8.8852,
       "step": 32000
     },
     {
       "epoch": 13.64399664147775,
-      "grad_norm": 7.806242942810059,
+      "grad_norm": 7.331772804260254,
       "learning_rate": 5.4591321897073666e-05,
-      "loss": 8.9316,
+      "loss": 8.9337,
       "step": 32500
     },
     {
       "epoch": 13.853904282115868,
-      "grad_norm": 7.273571491241455,
+      "grad_norm": 7.22735071182251,
       "learning_rate": 5.358224016145308e-05,
-      "loss": 8.9545,
+      "loss": 8.9578,
       "step": 33000
     },
     {
       "epoch": 14.063811922753988,
-      "grad_norm": 7.572352886199951,
+      "grad_norm": 7.088395118713379,
       "learning_rate": 5.257315842583249e-05,
-      "loss": 8.8332,
+      "loss": 8.8495,
       "step": 33500
     },
     {
       "epoch": 14.273719563392108,
-      "grad_norm": 8.740715980529785,
+      "grad_norm": 8.663612365722656,
       "learning_rate": 5.15640766902119e-05,
-      "loss": 8.5831,
+      "loss": 8.5942,
       "step": 34000
     },
     {
       "epoch": 14.483627204030226,
-      "grad_norm": 8.140837669372559,
+      "grad_norm": 7.5975847244262695,
       "learning_rate": 5.055499495459133e-05,
-      "loss": 8.6245,
+      "loss": 8.6454,
       "step": 34500
     },
     {
       "epoch": 14.693534844668346,
-      "grad_norm": 8.795963287353516,
+      "grad_norm": 7.2629876136779785,
       "learning_rate": 4.954591321897074e-05,
-      "loss": 8.6655,
+      "loss": 8.6736,
       "step": 35000
     },
     {
       "epoch": 14.903442485306465,
-      "grad_norm": 8.24455451965332,
+      "grad_norm": 7.846585273742676,
       "learning_rate": 4.853683148335015e-05,
-      "loss": 8.6897,
+      "loss": 8.7,
       "step": 35500
     },
     {
       "epoch": 15.113350125944585,
-      "grad_norm": 8.989808082580566,
+      "grad_norm": 8.747940063476562,
       "learning_rate": 4.7527749747729566e-05,
-      "loss": 8.4835,
+      "loss": 8.5133,
       "step": 36000
     },
     {
       "epoch": 15.323257766582703,
-      "grad_norm": 9.389223098754883,
+      "grad_norm": 9.021889686584473,
       "learning_rate": 4.6518668012108985e-05,
-      "loss": 8.3403,
+      "loss": 8.3672,
       "step": 36500
     },
     {
       "epoch": 15.533165407220823,
-      "grad_norm": 9.29480266571045,
+      "grad_norm": 8.786920547485352,
       "learning_rate": 4.55095862764884e-05,
-      "loss": 8.392,
+      "loss": 8.4053,
       "step": 37000
     },
     {
       "epoch": 15.743073047858942,
-      "grad_norm": 8.583866119384766,
+      "grad_norm": 8.324206352233887,
       "learning_rate": 4.450050454086781e-05,
-      "loss": 8.4257,
+      "loss": 8.4422,
       "step": 37500
     },
     {
       "epoch": 15.95298068849706,
-      "grad_norm": 9.255289077758789,
+      "grad_norm": 9.694009780883789,
       "learning_rate": 4.349142280524723e-05,
-      "loss": 8.4621,
+      "loss": 8.4855,
       "step": 38000
     },
     {
       "epoch": 16.162888329135182,
-      "grad_norm": 9.531908988952637,
+      "grad_norm": 9.064590454101562,
       "learning_rate": 4.248234106962664e-05,
-      "loss": 8.1496,
+      "loss": 8.192,
       "step": 38500
     },
     {
       "epoch": 16.3727959697733,
-      "grad_norm": 9.915640830993652,
+      "grad_norm": 9.216103553771973,
       "learning_rate": 4.147325933400605e-05,
-      "loss": 8.1239,
+      "loss": 8.1601,
       "step": 39000
     },
     {
       "epoch": 16.582703610411418,
-      "grad_norm": 9.33653736114502,
+      "grad_norm": 9.250482559204102,
       "learning_rate": 4.046417759838547e-05,
-      "loss": 8.1522,
+      "loss": 8.1841,
       "step": 39500
     },
     {
       "epoch": 16.79261125104954,
-      "grad_norm": 10.401065826416016,
+      "grad_norm": 9.507728576660156,
       "learning_rate": 3.9455095862764884e-05,
-      "loss": 8.1981,
+      "loss": 8.2175,
       "step": 40000
     },
     {
       "epoch": 16.79261125104954,
-      "eval_accuracy": 0.15049904676460693,
-      "eval_loss": 12.571419715881348,
-      "eval_runtime": 0.7087,
-      "eval_samples_per_second": 675.895,
-      "eval_steps_per_second": 7.055,
+      "eval_accuracy": 0.15125602781204442,
+      "eval_loss": 12.658378601074219,
+      "eval_runtime": 0.7467,
+      "eval_samples_per_second": 641.519,
+      "eval_steps_per_second": 6.696,
       "step": 40000
     },
     {
       "epoch": 17.002518891687657,
-      "grad_norm": 9.926044464111328,
+      "grad_norm": 8.521580696105957,
       "learning_rate": 3.84460141271443e-05,
-      "loss": 8.2202,
+      "loss": 8.252,
       "step": 40500
     },
     {
       "epoch": 17.212426532325775,
-      "grad_norm": 9.641388893127441,
+      "grad_norm": 8.822457313537598,
       "learning_rate": 3.7436932391523715e-05,
-      "loss": 7.868,
+      "loss": 7.9128,
       "step": 41000
     },
     {
       "epoch": 17.422334172963897,
-      "grad_norm": 10.477396011352539,
+      "grad_norm": 10.097407341003418,
       "learning_rate": 3.642785065590313e-05,
-      "loss": 7.9082,
+      "loss": 7.9455,
       "step": 41500
     },
     {
       "epoch": 17.632241813602015,
-      "grad_norm": 10.177766799926758,
+      "grad_norm": 10.7813081741333,
       "learning_rate": 3.541876892028255e-05,
-      "loss": 7.9471,
+      "loss": 7.9851,
       "step": 42000
     },
     {
       "epoch": 17.842149454240133,
-      "grad_norm": 10.878178596496582,
+      "grad_norm": 9.292275428771973,
       "learning_rate": 3.440968718466196e-05,
-      "loss": 7.9808,
+      "loss": 8.0182,
       "step": 42500
     },
     {
       "epoch": 18.052057094878254,
-      "grad_norm": 9.570734977722168,
+      "grad_norm": 9.53646183013916,
       "learning_rate": 3.340060544904137e-05,
-      "loss": 7.898,
+      "loss": 7.9434,
       "step": 43000
     },
     {
       "epoch": 18.261964735516372,
-      "grad_norm": 10.084707260131836,
+      "grad_norm": 10.176555633544922,
       "learning_rate": 3.239152371342079e-05,
-      "loss": 7.6738,
+      "loss": 7.7276,
       "step": 43500
     },
     {
       "epoch": 18.471872376154494,
-      "grad_norm": 10.34295654296875,
+      "grad_norm": 10.262212753295898,
       "learning_rate": 3.13824419778002e-05,
-      "loss": 7.7283,
+      "loss": 7.7723,
       "step": 44000
     },
     {
       "epoch": 18.68178001679261,
-      "grad_norm": 10.675642967224121,
+      "grad_norm": 9.775254249572754,
       "learning_rate": 3.0373360242179615e-05,
-      "loss": 7.7656,
+      "loss": 7.806,
       "step": 44500
     },
     {
       "epoch": 18.89168765743073,
-      "grad_norm": 10.677101135253906,
+      "grad_norm": 10.265592575073242,
       "learning_rate": 2.9364278506559034e-05,
-      "loss": 7.7995,
+      "loss": 7.8424,
       "step": 45000
     },
     {
       "epoch": 19.10159529806885,
-      "grad_norm": 10.481788635253906,
+      "grad_norm": 10.479218482971191,
       "learning_rate": 2.8355196770938446e-05,
-      "loss": 7.6326,
+      "loss": 7.6881,
       "step": 45500
     },
     {
       "epoch": 19.31150293870697,
-      "grad_norm": 10.14302921295166,
+      "grad_norm": 11.256647109985352,
       "learning_rate": 2.734611503531786e-05,
-      "loss": 7.527,
+      "loss": 7.5807,
       "step": 46000
     },
     {
       "epoch": 19.521410579345087,
-      "grad_norm": 11.806954383850098,
+      "grad_norm": 12.51082706451416,
       "learning_rate": 2.6337033299697278e-05,
-      "loss": 7.5512,
+      "loss": 7.6111,
       "step": 46500
     },
     {
       "epoch": 19.73131821998321,
-      "grad_norm": 11.329198837280273,
+      "grad_norm": 9.883224487304688,
       "learning_rate": 2.532795156407669e-05,
-      "loss": 7.5728,
+      "loss": 7.6202,
       "step": 47000
     },
     {
       "epoch": 19.941225860621326,
-      "grad_norm": 11.316740036010742,
+      "grad_norm": 10.54744815826416,
       "learning_rate": 2.4318869828456106e-05,
-      "loss": 7.6052,
+      "loss": 7.6449,
       "step": 47500
     },
     {
       "epoch": 20.151133501259444,
-      "grad_norm": 10.515312194824219,
+      "grad_norm": 9.121170997619629,
       "learning_rate": 2.330978809283552e-05,
-      "loss": 7.437,
+      "loss": 7.4934,
       "step": 48000
     },
     {
       "epoch": 20.361041141897566,
-      "grad_norm": 12.769679069519043,
+      "grad_norm": 11.935457229614258,
       "learning_rate": 2.2300706357214934e-05,
-      "loss": 7.3799,
+      "loss": 7.4323,
       "step": 48500
     },
     {
       "epoch": 20.570948782535684,
-      "grad_norm": 10.36323356628418,
+      "grad_norm": 10.347271919250488,
       "learning_rate": 2.129162462159435e-05,
-      "loss": 7.4071,
+      "loss": 7.4545,
       "step": 49000
     },
     {
       "epoch": 20.7808564231738,
-      "grad_norm": 10.92757511138916,
+      "grad_norm": 10.799528121948242,
       "learning_rate": 2.0282542885973765e-05,
-      "loss": 7.429,
+      "loss": 7.4741,
       "step": 49500
     },
     {
       "epoch": 20.990764063811923,
-      "grad_norm": 10.121776580810547,
+      "grad_norm": 9.884795188903809,
       "learning_rate": 1.927346115035318e-05,
-      "loss": 7.4533,
+      "loss": 7.4989,
       "step": 50000
     },
     {
       "epoch": 20.990764063811923,
-      "eval_accuracy": 0.1508957860206914,
-      "eval_loss": 13.652094841003418,
-      "eval_runtime": 0.7044,
-      "eval_samples_per_second": 680.021,
-      "eval_steps_per_second": 7.098,
+      "eval_accuracy": 0.15184904813973701,
+      "eval_loss": 13.474550247192383,
+      "eval_runtime": 0.7228,
+      "eval_samples_per_second": 662.696,
+      "eval_steps_per_second": 6.917,
       "step": 50000
     },
     {
       "epoch": 21.20067170445004,
-      "grad_norm": 11.646376609802246,
+      "grad_norm": 10.711160659790039,
       "learning_rate": 1.8264379414732593e-05,
-      "loss": 7.2613,
+      "loss": 7.3127,
       "step": 50500
     },
     {
       "epoch": 21.410579345088163,
-      "grad_norm": 10.331342697143555,
+      "grad_norm": 10.422394752502441,
       "learning_rate": 1.725529767911201e-05,
-      "loss": 7.279,
+      "loss": 7.3273,
       "step": 51000
     },
     {
       "epoch": 21.62048698572628,
-      "grad_norm": 11.476698875427246,
+      "grad_norm": 11.552720069885254,
       "learning_rate": 1.6246215943491424e-05,
-      "loss": 7.2961,
+      "loss": 7.3415,
       "step": 51500
     },
     {
       "epoch": 21.8303946263644,
-      "grad_norm": 12.156853675842285,
+      "grad_norm": 12.2216215133667,
       "learning_rate": 1.5237134207870837e-05,
-      "loss": 7.2957,
+      "loss": 7.3598,
       "step": 52000
     },
     {
       "epoch": 22.04030226700252,
-      "grad_norm": 8.92335319519043,
+      "grad_norm": 11.112517356872559,
       "learning_rate": 1.4228052472250252e-05,
-      "loss": 7.2764,
+      "loss": 7.332,
       "step": 52500
     },
     {
       "epoch": 22.250209907640638,
-      "grad_norm": 10.145243644714355,
+      "grad_norm": 10.92021369934082,
       "learning_rate": 1.3218970736629668e-05,
-      "loss": 7.1601,
+      "loss": 7.2086,
       "step": 53000
     },
     {
       "epoch": 22.460117548278756,
-      "grad_norm": 10.481016159057617,
+      "grad_norm": 10.774788856506348,
       "learning_rate": 1.2209889001009082e-05,
-      "loss": 7.1669,
+      "loss": 7.2187,
       "step": 53500
     },
     {
       "epoch": 22.670025188916878,
-      "grad_norm": 10.638428688049316,
+      "grad_norm": 10.409795761108398,
       "learning_rate": 1.1200807265388496e-05,
-      "loss": 7.1843,
+      "loss": 7.2431,
       "step": 54000
     },
     {
       "epoch": 22.879932829554996,
-      "grad_norm": 9.991191864013672,
+      "grad_norm": 9.242880821228027,
       "learning_rate": 1.0191725529767912e-05,
-      "loss": 7.1918,
+      "loss": 7.2409,
       "step": 54500
     },
     {
       "epoch": 23.089840470193113,
-      "grad_norm": 9.825872421264648,
+      "grad_norm": 10.372599601745605,
       "learning_rate": 9.182643794147326e-06,
-      "loss": 7.1576,
+      "loss": 7.199,
       "step": 55000
     },
     {
       "epoch": 23.299748110831235,
-      "grad_norm": 11.133133888244629,
+      "grad_norm": 11.163091659545898,
       "learning_rate": 8.173562058526741e-06,
-      "loss": 7.0867,
+      "loss": 7.1352,
       "step": 55500
     },
     {
       "epoch": 23.509655751469353,
-      "grad_norm": 10.233381271362305,
+      "grad_norm": 11.205037117004395,
       "learning_rate": 7.164480322906156e-06,
-      "loss": 7.1159,
+      "loss": 7.1597,
       "step": 56000
     },
     {
       "epoch": 23.719563392107474,
-      "grad_norm": 10.7654390335083,
+      "grad_norm": 10.741216659545898,
       "learning_rate": 6.155398587285571e-06,
-      "loss": 7.0929,
+      "loss": 7.145,
       "step": 56500
     },
     {
       "epoch": 23.929471032745592,
-      "grad_norm": 10.611351013183594,
+      "grad_norm": 10.837223052978516,
       "learning_rate": 5.146316851664986e-06,
-      "loss": 7.1108,
+      "loss": 7.1618,
       "step": 57000
     },
     {
       "epoch": 24.13937867338371,
-      "grad_norm": 10.949281692504883,
+      "grad_norm": 10.488913536071777,
       "learning_rate": 4.1372351160444e-06,
-      "loss": 7.0685,
+      "loss": 7.1065,
       "step": 57500
     },
     {
       "epoch": 24.349286314021832,
-      "grad_norm": 10.407085418701172,
+      "grad_norm": 11.610832214355469,
       "learning_rate": 3.1281533804238145e-06,
-      "loss": 7.0581,
+      "loss": 7.0994,
       "step": 58000
     },
     {
       "epoch": 24.55919395465995,
-      "grad_norm": 10.60808277130127,
+      "grad_norm": 12.08505916595459,
       "learning_rate": 2.1190716448032293e-06,
-      "loss": 7.0645,
+      "loss": 7.121,
       "step": 58500
     },
     {
       "epoch": 24.769101595298068,
-      "grad_norm": 10.884474754333496,
+      "grad_norm": 10.379035949707031,
       "learning_rate": 1.109989909182644e-06,
-      "loss": 7.0465,
+      "loss": 7.0947,
       "step": 59000
     },
     {
       "epoch": 24.97900923593619,
-      "grad_norm": 10.9041748046875,
+      "grad_norm": 10.820743560791016,
       "learning_rate": 1.0090817356205853e-07,
-      "loss": 7.0428,
+      "loss": 7.0874,
       "step": 59500
     },
     {
       "epoch": 25.0,
       "step": 59550,
       "total_flos": 3.058354515764736e+17,
-      "train_loss": 9.151931135944716,
-      "train_runtime": 12834.3964,
-      "train_samples_per_second": 445.364,
-      "train_steps_per_second": 4.64
+      "train_loss": 9.169899891581883,
+      "train_runtime": 12212.2898,
+      "train_samples_per_second": 468.051,
+      "train_steps_per_second": 4.876
     }
   ],
   "logging_steps": 500,
@@ -902,18 +901,6 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 25,
   "save_steps": 10000,
-  "stateful_callbacks": {
-    "TrainerControl": {
-      "args": {
-        "should_epoch_stop": false,
-        "should_evaluate": false,
-        "should_log": false,
-        "should_save": true,
-        "should_training_stop": true
-      },
-      "attributes": {}
-    }
-  },
   "total_flos": 3.058354515764736e+17,
   "train_batch_size": 96,
   "trial_name": null,