End of training

README.md

@@ -4,6 +4,8 @@ license: apache-2.0
 base_model: albert/albert-base-v2
 tags:
 - generated_from_trainer
+datasets:
+- squad
 model-index:
 - name: squad_albert_finetuned
   results: []
@@ -14,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 # squad_albert_finetuned
 
-This model is a fine-tuned version of [albert/albert-base-v2](https://huggingface.co/albert/albert-base-v2) on an unknown dataset.
+This model is a fine-tuned version of [albert/albert-base-v2](https://huggingface.co/albert/albert-base-v2) on the squad dataset.
 
 ## Model description
 

all_results.json

@@ -1,15 +1,15 @@
 {
-    "epoch": 4.0,
-    "eval_exact_match": 81.36234626300852,
-    "eval_f1": 89.36131376273367,
-    "eval_runtime": 36.9782,
+    "epoch": 6.0,
+    "eval_exact_match": 79.30936613055819,
+    "eval_f1": 88.03839918996134,
+    "eval_runtime": 37.1649,
     "eval_samples": 10808,
-    "eval_samples_per_second": 292.28,
-    "eval_steps_per_second": 36.535,
-    "total_flos": 5872392427106304.0,
-    "train_loss": 0.631021389978457,
-    "train_runtime": 2829.8023,
+    "eval_samples_per_second": 290.812,
+    "eval_steps_per_second": 36.351,
+    "total_flos": 8808588640659456.0,
+    "train_loss": 0.4993116462560624,
+    "train_runtime": 4288.5384,
     "train_samples": 88638,
-    "train_samples_per_second": 125.292,
-    "train_steps_per_second": 3.915
+    "train_samples_per_second": 124.011,
+    "train_steps_per_second": 3.875
 }

eval_nbest_predictions.json

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98cdcd49f13fca647cb743c25812163ac29a522bd487c2442d8ee289edb3c36d
-size 48401088
+oid sha256:503e0e42fd89214862dba2efd2fd5bed48232796cb1ef8fb1f702b3933cf1300
+size 48528698

eval_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 4.0,
-    "eval_exact_match": 81.36234626300852,
-    "eval_f1": 89.36131376273367,
-    "eval_runtime": 36.9782,
+    "epoch": 6.0,
+    "eval_exact_match": 79.30936613055819,
+    "eval_f1": 88.03839918996134,
+    "eval_runtime": 37.1649,
     "eval_samples": 10808,
-    "eval_samples_per_second": 292.28,
-    "eval_steps_per_second": 36.535
+    "eval_samples_per_second": 290.812,
+    "eval_steps_per_second": 36.351
 }

runs/Dec21_03-55-20_xgpi2/events.out.tfevents.1734728898.xgpi2.1075641.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea7682672d453c9dac2d76baea73e89e0277ddf696945655b9045446fcb08259
+size 418

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 4.0,
-    "total_flos": 5872392427106304.0,
-    "train_loss": 0.631021389978457,
-    "train_runtime": 2829.8023,
+    "epoch": 6.0,
+    "total_flos": 8808588640659456.0,
+    "train_loss": 0.4993116462560624,
+    "train_runtime": 4288.5384,
     "train_samples": 88638,
-    "train_samples_per_second": 125.292,
-    "train_steps_per_second": 3.915
+    "train_samples_per_second": 124.011,
+    "train_steps_per_second": 3.875
 }

trainer_state.json

@@ -1,181 +1,258 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.0,
+  "epoch": 6.0,
   "eval_steps": 500,
-  "global_step": 11080,
+  "global_step": 16620,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.18050541516245489,
-      "grad_norm": 32.10551452636719,
-      "learning_rate": 2.864620938628159e-05,
-      "loss": 1.3035,
+      "grad_norm": 31.118467330932617,
+      "learning_rate": 2.909747292418773e-05,
+      "loss": 1.3055,
       "step": 500
     },
     {
       "epoch": 0.36101083032490977,
-      "grad_norm": 16.806726455688477,
-      "learning_rate": 2.729241877256318e-05,
-      "loss": 1.016,
+      "grad_norm": 16.015565872192383,
+      "learning_rate": 2.819494584837545e-05,
+      "loss": 1.0185,
       "step": 1000
     },
     {
       "epoch": 0.5415162454873647,
-      "grad_norm": 17.086566925048828,
-      "learning_rate": 2.5938628158844765e-05,
-      "loss": 0.9497,
+      "grad_norm": 16.465599060058594,
+      "learning_rate": 2.729241877256318e-05,
+      "loss": 0.9546,
       "step": 1500
     },
     {
       "epoch": 0.7220216606498195,
-      "grad_norm": 11.685013771057129,
-      "learning_rate": 2.4584837545126353e-05,
-      "loss": 0.925,
+      "grad_norm": 11.709087371826172,
+      "learning_rate": 2.6389891696750903e-05,
+      "loss": 0.9259,
       "step": 2000
     },
     {
       "epoch": 0.9025270758122743,
-      "grad_norm": 21.977975845336914,
-      "learning_rate": 2.3231046931407943e-05,
-      "loss": 0.8909,
+      "grad_norm": 22.127365112304688,
+      "learning_rate": 2.548736462093863e-05,
+      "loss": 0.8966,
       "step": 2500
     },
     {
       "epoch": 1.0830324909747293,
-      "grad_norm": 22.408674240112305,
-      "learning_rate": 2.1877256317689534e-05,
-      "loss": 0.7819,
+      "grad_norm": 14.972929954528809,
+      "learning_rate": 2.4584837545126353e-05,
+      "loss": 0.7928,
       "step": 3000
     },
     {
       "epoch": 1.263537906137184,
-      "grad_norm": 17.96110725402832,
-      "learning_rate": 2.0523465703971117e-05,
-      "loss": 0.6782,
+      "grad_norm": 16.17904281616211,
+      "learning_rate": 2.368231046931408e-05,
+      "loss": 0.6899,
       "step": 3500
     },
     {
       "epoch": 1.444043321299639,
-      "grad_norm": 69.72351837158203,
-      "learning_rate": 1.9169675090252708e-05,
-      "loss": 0.6963,
+      "grad_norm": 48.85356521606445,
+      "learning_rate": 2.2779783393501805e-05,
+      "loss": 0.7077,
       "step": 4000
     },
     {
       "epoch": 1.6245487364620939,
-      "grad_norm": 18.7786865234375,
-      "learning_rate": 1.7815884476534298e-05,
-      "loss": 0.6797,
+      "grad_norm": 26.525949478149414,
+      "learning_rate": 2.1877256317689534e-05,
+      "loss": 0.6929,
       "step": 4500
     },
     {
       "epoch": 1.8050541516245486,
-      "grad_norm": 15.781920433044434,
-      "learning_rate": 1.6462093862815885e-05,
-      "loss": 0.6732,
+      "grad_norm": 17.478593826293945,
+      "learning_rate": 2.0974729241877255e-05,
+      "loss": 0.6838,
       "step": 5000
     },
     {
       "epoch": 1.9855595667870036,
-      "grad_norm": 21.812726974487305,
-      "learning_rate": 1.5108303249097474e-05,
-      "loss": 0.689,
+      "grad_norm": 28.06412124633789,
+      "learning_rate": 2.0072202166064983e-05,
+      "loss": 0.7085,
       "step": 5500
     },
     {
       "epoch": 2.1660649819494586,
-      "grad_norm": 19.687843322753906,
-      "learning_rate": 1.3754512635379063e-05,
-      "loss": 0.5086,
+      "grad_norm": 13.594704627990723,
+      "learning_rate": 1.9169675090252708e-05,
+      "loss": 0.5315,
       "step": 6000
     },
     {
       "epoch": 2.3465703971119134,
-      "grad_norm": 29.61492919921875,
-      "learning_rate": 1.240072202166065e-05,
-      "loss": 0.4964,
+      "grad_norm": 18.916522979736328,
+      "learning_rate": 1.8267148014440436e-05,
+      "loss": 0.5234,
       "step": 6500
     },
     {
       "epoch": 2.527075812274368,
-      "grad_norm": 22.271451950073242,
-      "learning_rate": 1.1046931407942239e-05,
-      "loss": 0.5023,
+      "grad_norm": 27.207935333251953,
+      "learning_rate": 1.7364620938628157e-05,
+      "loss": 0.5258,
       "step": 7000
     },
     {
       "epoch": 2.707581227436823,
-      "grad_norm": 12.808157920837402,
-      "learning_rate": 9.693140794223826e-06,
-      "loss": 0.5029,
+      "grad_norm": 14.564919471740723,
+      "learning_rate": 1.6462093862815885e-05,
+      "loss": 0.5304,
       "step": 7500
     },
     {
       "epoch": 2.888086642599278,
-      "grad_norm": 18.084169387817383,
-      "learning_rate": 8.339350180505416e-06,
-      "loss": 0.4949,
+      "grad_norm": 26.674354553222656,
+      "learning_rate": 1.555956678700361e-05,
+      "loss": 0.5306,
       "step": 8000
     },
     {
       "epoch": 3.068592057761733,
-      "grad_norm": 16.37100601196289,
-      "learning_rate": 6.985559566787004e-06,
-      "loss": 0.4302,
+      "grad_norm": 17.092002868652344,
+      "learning_rate": 1.4657039711191336e-05,
+      "loss": 0.4566,
       "step": 8500
     },
     {
       "epoch": 3.2490974729241877,
-      "grad_norm": 21.810672760009766,
-      "learning_rate": 5.631768953068592e-06,
-      "loss": 0.3405,
+      "grad_norm": 26.201404571533203,
+      "learning_rate": 1.3754512635379063e-05,
+      "loss": 0.3703,
       "step": 9000
     },
     {
       "epoch": 3.4296028880866425,
-      "grad_norm": 9.269281387329102,
-      "learning_rate": 4.277978339350181e-06,
-      "loss": 0.3505,
+      "grad_norm": 7.88193416595459,
+      "learning_rate": 1.2851985559566788e-05,
+      "loss": 0.3834,
       "step": 9500
     },
     {
       "epoch": 3.6101083032490973,
-      "grad_norm": 25.89681053161621,
-      "learning_rate": 2.924187725631769e-06,
-      "loss": 0.3474,
+      "grad_norm": 17.27417755126953,
+      "learning_rate": 1.1949458483754514e-05,
+      "loss": 0.3878,
       "step": 10000
     },
     {
       "epoch": 3.7906137184115525,
-      "grad_norm": 19.89838409423828,
-      "learning_rate": 1.5703971119133576e-06,
-      "loss": 0.3391,
+      "grad_norm": 19.228748321533203,
+      "learning_rate": 1.1046931407942239e-05,
+      "loss": 0.3887,
       "step": 10500
     },
     {
       "epoch": 3.9711191335740073,
-      "grad_norm": 28.91718864440918,
-      "learning_rate": 2.1660649819494586e-07,
-      "loss": 0.334,
+      "grad_norm": 25.48954963684082,
+      "learning_rate": 1.0144404332129965e-05,
+      "loss": 0.3818,
       "step": 11000
     },
     {
-      "epoch": 4.0,
-      "step": 11080,
-      "total_flos": 5872392427106304.0,
-      "train_loss": 0.631021389978457,
-      "train_runtime": 2829.8023,
-      "train_samples_per_second": 125.292,
-      "train_steps_per_second": 3.915
+      "epoch": 4.1516245487364625,
+      "grad_norm": 27.043731689453125,
+      "learning_rate": 9.24187725631769e-06,
+      "loss": 0.2671,
+      "step": 11500
+    },
+    {
+      "epoch": 4.332129963898917,
+      "grad_norm": 19.578210830688477,
+      "learning_rate": 8.339350180505416e-06,
+      "loss": 0.2483,
+      "step": 12000
+    },
+    {
+      "epoch": 4.512635379061372,
+      "grad_norm": 21.307836532592773,
+      "learning_rate": 7.436823104693141e-06,
+      "loss": 0.2513,
+      "step": 12500
+    },
+    {
+      "epoch": 4.693140794223827,
+      "grad_norm": 21.601924896240234,
+      "learning_rate": 6.534296028880867e-06,
+      "loss": 0.2517,
+      "step": 13000
+    },
+    {
+      "epoch": 4.873646209386282,
+      "grad_norm": 20.635147094726562,
+      "learning_rate": 5.631768953068592e-06,
+      "loss": 0.2527,
+      "step": 13500
+    },
+    {
+      "epoch": 5.054151624548736,
+      "grad_norm": 21.017452239990234,
+      "learning_rate": 4.729241877256318e-06,
+      "loss": 0.2183,
+      "step": 14000
+    },
+    {
+      "epoch": 5.234657039711191,
+      "grad_norm": 18.656475067138672,
+      "learning_rate": 3.826714801444043e-06,
+      "loss": 0.1441,
+      "step": 14500
+    },
+    {
+      "epoch": 5.415162454873646,
+      "grad_norm": 16.917116165161133,
+      "learning_rate": 2.924187725631769e-06,
+      "loss": 0.1417,
+      "step": 15000
+    },
+    {
+      "epoch": 5.595667870036101,
+      "grad_norm": 25.11797332763672,
+      "learning_rate": 2.0216606498194946e-06,
+      "loss": 0.1363,
+      "step": 15500
+    },
+    {
+      "epoch": 5.776173285198556,
+      "grad_norm": 25.358491897583008,
+      "learning_rate": 1.1191335740072204e-06,
+      "loss": 0.1341,
+      "step": 16000
+    },
+    {
+      "epoch": 5.956678700361011,
+      "grad_norm": 24.253881454467773,
+      "learning_rate": 2.1660649819494586e-07,
+      "loss": 0.1334,
+      "step": 16500
+    },
+    {
+      "epoch": 6.0,
+      "step": 16620,
+      "total_flos": 8808588640659456.0,
+      "train_loss": 0.4993116462560624,
+      "train_runtime": 4288.5384,
+      "train_samples_per_second": 124.011,
+      "train_steps_per_second": 3.875
     }
   ],
   "logging_steps": 500,
-  "max_steps": 11080,
+  "max_steps": 16620,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 6,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -189,7 +266,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5872392427106304.0,
+  "total_flos": 8808588640659456.0,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null