End of training

README.md

@@ -17,11 +17,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [aubmindlab/aragpt2-base](https://huggingface.co/aubmindlab/aragpt2-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0556
-- Bleu: 0.0625
-- Rouge1: 0.3737
-- Rouge2: 0.1489
-- Rougel: 0.3670
+- Loss: 0.0534
+- Bleu: 0.0428
+- Rouge1: 0.3139
+- Rouge2: 0.1104
+- Rougel: 0.3097
 
 ## Model description
 

all_results.json

@@ -0,0 +1,19 @@
+{
+    "epoch": 14.0,
+    "eval_bleu": 0.04280893852047394,
+    "eval_loss": 0.05337703973054886,
+    "eval_rouge1": 0.31388693258006284,
+    "eval_rouge2": 0.11036995985896458,
+    "eval_rougeL": 0.3096668301159336,
+    "eval_runtime": 7.0173,
+    "eval_samples": 304,
+    "eval_samples_per_second": 43.322,
+    "eval_steps_per_second": 5.415,
+    "perplexity": 1.0548272819463649,
+    "total_flos": 8918419636224000.0,
+    "train_loss": 0.46412634738241815,
+    "train_runtime": 1276.9352,
+    "train_samples": 1219,
+    "train_samples_per_second": 19.093,
+    "train_steps_per_second": 2.396
+}

egy_training_log.txt

@@ -304,3 +304,5 @@ INFO:root:Epoch 13.0: Train Loss = 0.0331, Eval Loss = 0.05399588495492935
 INFO:absl:Using default tokenizer.
 INFO:root:Epoch 14.0: Train Loss = 0.0308, Eval Loss = 0.05515788868069649
 INFO:absl:Using default tokenizer.
+INFO:__main__:*** Evaluate ***
+INFO:absl:Using default tokenizer.

eval_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 14.0,
+    "eval_bleu": 0.04280893852047394,
+    "eval_loss": 0.05337703973054886,
+    "eval_rouge1": 0.31388693258006284,
+    "eval_rouge2": 0.11036995985896458,
+    "eval_rougeL": 0.3096668301159336,
+    "eval_runtime": 7.0173,
+    "eval_samples": 304,
+    "eval_samples_per_second": 43.322,
+    "eval_steps_per_second": 5.415,
+    "perplexity": 1.0548272819463649
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 14.0,
+    "total_flos": 8918419636224000.0,
+    "train_loss": 0.46412634738241815,
+    "train_runtime": 1276.9352,
+    "train_samples": 1219,
+    "train_samples_per_second": 19.093,
+    "train_steps_per_second": 2.396
+}

trainer_state.json

@@ -0,0 +1,317 @@
+{
+  "best_metric": 0.05337703973054886,
+  "best_model_checkpoint": "/home/iais_marenpielka/Bouthaina/res_nw_yem_aragpt2-base/checkpoint-1377",
+  "epoch": 14.0,
+  "eval_steps": 500,
+  "global_step": 2142,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3274374306201935,
+      "learning_rate": 1.53e-05,
+      "loss": 5.8224,
+      "step": 153
+    },
+    {
+      "epoch": 1.0,
+      "eval_bleu": 0.00392224521343418,
+      "eval_loss": 0.1128806546330452,
+      "eval_rouge1": 0.07124894337258883,
+      "eval_rouge2": 0.0028587092731829577,
+      "eval_rougeL": 0.06948923152022689,
+      "eval_runtime": 7.1373,
+      "eval_samples_per_second": 42.593,
+      "eval_steps_per_second": 5.324,
+      "step": 153
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3334260880947113,
+      "learning_rate": 3.06e-05,
+      "loss": 0.1108,
+      "step": 306
+    },
+    {
+      "epoch": 2.0,
+      "eval_bleu": 0.0,
+      "eval_loss": 0.06912114471197128,
+      "eval_rouge1": 0.09507863230508208,
+      "eval_rouge2": 0.007788772376873063,
+      "eval_rougeL": 0.09317127790126112,
+      "eval_runtime": 6.9813,
+      "eval_samples_per_second": 43.545,
+      "eval_steps_per_second": 5.443,
+      "step": 306
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.2704195976257324,
+      "learning_rate": 4.5900000000000004e-05,
+      "loss": 0.0775,
+      "step": 459
+    },
+    {
+      "epoch": 3.0,
+      "eval_bleu": 0.006737157818299724,
+      "eval_loss": 0.06276746839284897,
+      "eval_rouge1": 0.12907377035145118,
+      "eval_rouge2": 0.015683124441348124,
+      "eval_rougeL": 0.1286203798342947,
+      "eval_runtime": 6.9851,
+      "eval_samples_per_second": 43.521,
+      "eval_steps_per_second": 5.44,
+      "step": 459
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.32800641655921936,
+      "learning_rate": 4.7812500000000003e-05,
+      "loss": 0.0678,
+      "step": 612
+    },
+    {
+      "epoch": 4.0,
+      "eval_bleu": 0.008597461521382226,
+      "eval_loss": 0.059174273163080215,
+      "eval_rouge1": 0.1524066357986381,
+      "eval_rouge2": 0.027263211934264564,
+      "eval_rougeL": 0.14917670820737453,
+      "eval_runtime": 6.9134,
+      "eval_samples_per_second": 43.973,
+      "eval_steps_per_second": 5.497,
+      "step": 612
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.18473878502845764,
+      "learning_rate": 4.482421875e-05,
+      "loss": 0.0603,
+      "step": 765
+    },
+    {
+      "epoch": 5.0,
+      "eval_bleu": 0.016158903735540513,
+      "eval_loss": 0.05656920000910759,
+      "eval_rouge1": 0.19186830417118805,
+      "eval_rouge2": 0.04134465720083692,
+      "eval_rougeL": 0.18829939413002822,
+      "eval_runtime": 6.9842,
+      "eval_samples_per_second": 43.527,
+      "eval_steps_per_second": 5.441,
+      "step": 765
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.2642894685268402,
+      "learning_rate": 4.18359375e-05,
+      "loss": 0.0547,
+      "step": 918
+    },
+    {
+      "epoch": 6.0,
+      "eval_bleu": 0.018741315914778903,
+      "eval_loss": 0.05464606359601021,
+      "eval_rouge1": 0.2238907288379966,
+      "eval_rouge2": 0.05989126658873175,
+      "eval_rougeL": 0.22184683032173025,
+      "eval_runtime": 7.0034,
+      "eval_samples_per_second": 43.408,
+      "eval_steps_per_second": 5.426,
+      "step": 918
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.24372832477092743,
+      "learning_rate": 3.884765625e-05,
+      "loss": 0.0498,
+      "step": 1071
+    },
+    {
+      "epoch": 7.0,
+      "eval_bleu": 0.029468363597052377,
+      "eval_loss": 0.054019927978515625,
+      "eval_rouge1": 0.26839447635173574,
+      "eval_rouge2": 0.07325536872796896,
+      "eval_rougeL": 0.2638108570075386,
+      "eval_runtime": 6.9888,
+      "eval_samples_per_second": 43.498,
+      "eval_steps_per_second": 5.437,
+      "step": 1071
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.2509661316871643,
+      "learning_rate": 3.5859375e-05,
+      "loss": 0.0456,
+      "step": 1224
+    },
+    {
+      "epoch": 8.0,
+      "eval_bleu": 0.029195772701040013,
+      "eval_loss": 0.053611643612384796,
+      "eval_rouge1": 0.2883971192764231,
+      "eval_rouge2": 0.08184845936724905,
+      "eval_rougeL": 0.28406572737215363,
+      "eval_runtime": 6.9871,
+      "eval_samples_per_second": 43.509,
+      "eval_steps_per_second": 5.439,
+      "step": 1224
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.23149649798870087,
+      "learning_rate": 3.287109375e-05,
+      "loss": 0.0419,
+      "step": 1377
+    },
+    {
+      "epoch": 9.0,
+      "eval_bleu": 0.04280893852047394,
+      "eval_loss": 0.05337703973054886,
+      "eval_rouge1": 0.31388693258006284,
+      "eval_rouge2": 0.11036995985896458,
+      "eval_rougeL": 0.3096668301159336,
+      "eval_runtime": 6.9883,
+      "eval_samples_per_second": 43.501,
+      "eval_steps_per_second": 5.438,
+      "step": 1377
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.2521679401397705,
+      "learning_rate": 2.9882812500000002e-05,
+      "loss": 0.0385,
+      "step": 1530
+    },
+    {
+      "epoch": 10.0,
+      "eval_bleu": 0.046113511822263795,
+      "eval_loss": 0.05344715714454651,
+      "eval_rouge1": 0.32550355167793193,
+      "eval_rouge2": 0.11175887195986847,
+      "eval_rougeL": 0.3185482802099305,
+      "eval_runtime": 6.9943,
+      "eval_samples_per_second": 43.464,
+      "eval_steps_per_second": 5.433,
+      "step": 1530
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.23900777101516724,
+      "learning_rate": 2.689453125e-05,
+      "loss": 0.0354,
+      "step": 1683
+    },
+    {
+      "epoch": 11.0,
+      "eval_bleu": 0.047259792068284935,
+      "eval_loss": 0.05403715744614601,
+      "eval_rouge1": 0.3357586244098568,
+      "eval_rouge2": 0.12191279657273882,
+      "eval_rougeL": 0.3287623290777512,
+      "eval_runtime": 7.0002,
+      "eval_samples_per_second": 43.427,
+      "eval_steps_per_second": 5.428,
+      "step": 1683
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.24781078100204468,
+      "learning_rate": 2.3906250000000002e-05,
+      "loss": 0.0331,
+      "step": 1836
+    },
+    {
+      "epoch": 12.0,
+      "eval_bleu": 0.0475663160301358,
+      "eval_loss": 0.05399588495492935,
+      "eval_rouge1": 0.3483203778520053,
+      "eval_rouge2": 0.13122980561384404,
+      "eval_rougeL": 0.3442445629111672,
+      "eval_runtime": 6.9958,
+      "eval_samples_per_second": 43.455,
+      "eval_steps_per_second": 5.432,
+      "step": 1836
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.3433144986629486,
+      "learning_rate": 2.091796875e-05,
+      "loss": 0.0308,
+      "step": 1989
+    },
+    {
+      "epoch": 13.0,
+      "eval_bleu": 0.05895730441080768,
+      "eval_loss": 0.05515788868069649,
+      "eval_rouge1": 0.3599156447280303,
+      "eval_rouge2": 0.14385576959590118,
+      "eval_rougeL": 0.3539109033921022,
+      "eval_runtime": 6.9936,
+      "eval_samples_per_second": 43.469,
+      "eval_steps_per_second": 5.434,
+      "step": 1989
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.3219660520553589,
+      "learning_rate": 1.79296875e-05,
+      "loss": 0.0291,
+      "step": 2142
+    },
+    {
+      "epoch": 14.0,
+      "eval_bleu": 0.06246557033315383,
+      "eval_loss": 0.05557234585285187,
+      "eval_rouge1": 0.37373672906337496,
+      "eval_rouge2": 0.14894175508943397,
+      "eval_rougeL": 0.3669928154707829,
+      "eval_runtime": 6.9886,
+      "eval_samples_per_second": 43.499,
+      "eval_steps_per_second": 5.437,
+      "step": 2142
+    },
+    {
+      "epoch": 14.0,
+      "step": 2142,
+      "total_flos": 8918419636224000.0,
+      "train_loss": 0.46412634738241815,
+      "train_runtime": 1276.9352,
+      "train_samples_per_second": 19.093,
+      "train_steps_per_second": 2.396
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 3060,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 20,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8918419636224000.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}