End of training

README.md

@@ -3,6 +3,9 @@ license: apache-2.0
 base_model: riotu-lab/ArabianGPT-01B
 tags:
 - generated_from_trainer
+metrics:
+- bleu
+- rouge
 model-index:
 - name: test-egy_aragpt
   results: []
@@ -14,6 +17,12 @@ should probably proofread and complete it, then remove this comment. -->
 # test-egy_aragpt
 
 This model is a fine-tuned version of [riotu-lab/ArabianGPT-01B](https://huggingface.co/riotu-lab/ArabianGPT-01B) on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 3.4630
+- Bleu: 0.0984
+- Rouge1: 0.3093
+- Rouge2: 0.0718
+- Rougel: 0.2296
 
 ## Model description
 

all_results.json

@@ -1,19 +1,19 @@
 {
-    "epoch": 2.0,
-    "eval_bleu": 0.11424038411303619,
-    "eval_loss": 3.128293514251709,
-    "eval_rouge1": 0.3297614987151056,
-    "eval_rouge2": 0.08429294540985294,
-    "eval_rougeL": 0.2561476738686219,
-    "eval_runtime": 26.3662,
+    "epoch": 1.0,
+    "eval_bleu": 0.09842217214841122,
+    "eval_loss": 3.4630110263824463,
+    "eval_rouge1": 0.3093411104418377,
+    "eval_rouge2": 0.07177511381538038,
+    "eval_rougeL": 0.22957881226204224,
+    "eval_runtime": 26.8225,
     "eval_samples": 847,
-    "eval_samples_per_second": 32.124,
-    "eval_steps_per_second": 4.02,
-    "perplexity": 22.834978708992665,
-    "total_flos": 2000451796992000.0,
-    "train_loss": 4.049833530916316,
-    "train_runtime": 263.3387,
+    "eval_samples_per_second": 31.578,
+    "eval_steps_per_second": 3.952,
+    "perplexity": 31.912922646039927,
+    "total_flos": 1000225898496000.0,
+    "train_loss": 4.835940561324452,
+    "train_runtime": 118.408,
     "train_samples": 2552,
-    "train_samples_per_second": 19.382,
-    "train_steps_per_second": 2.423
+    "train_samples_per_second": 21.553,
+    "train_steps_per_second": 2.694
 }

egy_training_log.txt

@@ -141,3 +141,5 @@ INFO:datasets.arrow_dataset:Loading cached processed dataset at /home/iais_maren
 INFO:datasets.arrow_dataset:Loading cached processed dataset at /home/iais_marenpielka/.cache/huggingface/datasets/text/default-0637777c38512acf/0.0.0/96636a050ef51804b84abbfd4f4ad440e01153c24b86293eb5c3b300a41f9101/cache-4ddbb6e08bb37d3f.arrow
 WARNING:accelerate.utils.other:Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.
 WARNING:root:Epoch 1.0: No losses recorded yet.
+INFO:__main__:*** Evaluate ***
+INFO:absl:Using default tokenizer.

eval_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 2.0,
-    "eval_bleu": 0.11424038411303619,
-    "eval_loss": 3.128293514251709,
-    "eval_rouge1": 0.3297614987151056,
-    "eval_rouge2": 0.08429294540985294,
-    "eval_rougeL": 0.2561476738686219,
-    "eval_runtime": 26.3662,
+    "epoch": 1.0,
+    "eval_bleu": 0.09842217214841122,
+    "eval_loss": 3.4630110263824463,
+    "eval_rouge1": 0.3093411104418377,
+    "eval_rouge2": 0.07177511381538038,
+    "eval_rougeL": 0.22957881226204224,
+    "eval_runtime": 26.8225,
     "eval_samples": 847,
-    "eval_samples_per_second": 32.124,
-    "eval_steps_per_second": 4.02,
-    "perplexity": 22.834978708992665
+    "eval_samples_per_second": 31.578,
+    "eval_steps_per_second": 3.952,
+    "perplexity": 31.912922646039927
 }

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 2.0,
-    "total_flos": 2000451796992000.0,
-    "train_loss": 4.049833530916316,
-    "train_runtime": 263.3387,
+    "epoch": 1.0,
+    "total_flos": 1000225898496000.0,
+    "train_loss": 4.835940561324452,
+    "train_runtime": 118.408,
     "train_samples": 2552,
-    "train_samples_per_second": 19.382,
-    "train_steps_per_second": 2.423
+    "train_samples_per_second": 21.553,
+    "train_steps_per_second": 2.694
 }

trainer_state.json

@@ -1,46 +1,27 @@
 {
-  "best_metric": 3.128293514251709,
-  "best_model_checkpoint": "/tmp/test-egy_aragpt/checkpoint-500",
-  "epoch": 2.0,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 638,
+  "global_step": 319,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.567398119122257,
-      "grad_norm": 1.3536401987075806,
-      "learning_rate": 5e-05,
-      "loss": 4.3013,
-      "step": 500
-    },
-    {
-      "epoch": 1.567398119122257,
-      "eval_bleu": 0.11424038411303619,
-      "eval_loss": 3.128293514251709,
-      "eval_rouge1": 0.3297614987151056,
-      "eval_rouge2": 0.08429294540985294,
-      "eval_rougeL": 0.2561476738686219,
-      "eval_runtime": 26.7734,
-      "eval_samples_per_second": 31.636,
-      "eval_steps_per_second": 3.959,
-      "step": 500
-    },
-    {
-      "epoch": 2.0,
-      "step": 638,
-      "total_flos": 2000451796992000.0,
-      "train_loss": 4.049833530916316,
-      "train_runtime": 263.3387,
-      "train_samples_per_second": 19.382,
-      "train_steps_per_second": 2.423
+      "epoch": 1.0,
+      "step": 319,
+      "total_flos": 1000225898496000.0,
+      "train_loss": 4.835940561324452,
+      "train_runtime": 118.408,
+      "train_samples_per_second": 21.553,
+      "train_steps_per_second": 2.694
     }
   ],
   "logging_steps": 500,
-  "max_steps": 638,
+  "max_steps": 319,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
+  "num_train_epochs": 1,
   "save_steps": 500,
   "stateful_callbacks": {
     "EarlyStoppingCallback": {
@@ -63,7 +44,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2000451796992000.0,
+  "total_flos": 1000225898496000.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null