End of training

README.md

@@ -3,6 +3,8 @@ license: apache-2.0
 base_model: riotu-lab/ArabianGPT-01B
 tags:
 - generated_from_trainer
+metrics:
+- bleu
 model-index:
 - name: test-aragpt
   results: []
@@ -14,6 +16,9 @@ should probably proofread and complete it, then remove this comment. -->
 # test-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: 3.3262
+- Bleu: 0.1060
 
 ## Model description
 

all_results.json

@@ -1,16 +1,16 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.5484330769384699,
-    "eval_loss": 2.3788321018218994,
-    "eval_runtime": 137.4021,
+    "epoch": 1.0,
+    "eval_bleu": 0.10603800831832183,
+    "eval_loss": 3.326219081878662,
+    "eval_runtime": 18.595,
     "eval_samples": 847,
-    "eval_samples_per_second": 6.164,
-    "eval_steps_per_second": 0.771,
-    "perplexity": 10.792291203563014,
-    "total_flos": 1.000225898496e+16,
-    "train_loss": 2.539044189453125,
-    "train_runtime": 1176.1508,
+    "eval_samples_per_second": 45.55,
+    "eval_steps_per_second": 5.7,
+    "perplexity": 27.832908563445223,
+    "total_flos": 1000225898496000.0,
+    "train_loss": 3.8128880228742164,
+    "train_runtime": 119.3474,
     "train_samples": 2552,
-    "train_samples_per_second": 21.698,
-    "train_steps_per_second": 2.712
+    "train_samples_per_second": 21.383,
+    "train_steps_per_second": 2.673
 }

eval_results.json

@@ -1,10 +1,10 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.5484330769384699,
-    "eval_loss": 2.3788321018218994,
-    "eval_runtime": 137.4021,
+    "epoch": 1.0,
+    "eval_bleu": 0.10603800831832183,
+    "eval_loss": 3.326219081878662,
+    "eval_runtime": 18.595,
     "eval_samples": 847,
-    "eval_samples_per_second": 6.164,
-    "eval_steps_per_second": 0.771,
-    "perplexity": 10.792291203563014
+    "eval_samples_per_second": 45.55,
+    "eval_steps_per_second": 5.7,
+    "perplexity": 27.832908563445223
 }

train_results.json

@@ -1,9 +1,9 @@
 {
-    "epoch": 10.0,
-    "total_flos": 1.000225898496e+16,
-    "train_loss": 2.539044189453125,
-    "train_runtime": 1176.1508,
+    "epoch": 1.0,
+    "total_flos": 1000225898496000.0,
+    "train_loss": 3.8128880228742164,
+    "train_runtime": 119.3474,
     "train_samples": 2552,
-    "train_samples_per_second": 21.698,
-    "train_steps_per_second": 2.712
+    "train_samples_per_second": 21.383,
+    "train_steps_per_second": 2.673
 }

trainer_state.json

@@ -1,69 +1,27 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 10.0,
+  "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 3190,
+  "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": 1.1572401523590088,
-      "learning_rate": 4.2163009404388715e-05,
-      "loss": 3.463,
-      "step": 500
-    },
-    {
-      "epoch": 3.134796238244514,
-      "grad_norm": 1.1287379264831543,
-      "learning_rate": 3.4326018808777435e-05,
-      "loss": 2.7518,
-      "step": 1000
-    },
-    {
-      "epoch": 4.702194357366771,
-      "grad_norm": 1.1390870809555054,
-      "learning_rate": 2.6489028213166144e-05,
-      "loss": 2.479,
-      "step": 1500
-    },
-    {
-      "epoch": 6.269592476489028,
-      "grad_norm": 1.1493933200836182,
-      "learning_rate": 1.865203761755486e-05,
-      "loss": 2.3191,
-      "step": 2000
-    },
-    {
-      "epoch": 7.836990595611285,
-      "grad_norm": 1.1453299522399902,
-      "learning_rate": 1.0815047021943574e-05,
-      "loss": 2.2187,
-      "step": 2500
-    },
-    {
-      "epoch": 9.404388714733543,
-      "grad_norm": 1.1752275228500366,
-      "learning_rate": 2.9780564263322885e-06,
-      "loss": 2.1545,
-      "step": 3000
-    },
-    {
-      "epoch": 10.0,
-      "step": 3190,
-      "total_flos": 1.000225898496e+16,
-      "train_loss": 2.539044189453125,
-      "train_runtime": 1176.1508,
-      "train_samples_per_second": 21.698,
-      "train_steps_per_second": 2.712
+      "epoch": 1.0,
+      "step": 319,
+      "total_flos": 1000225898496000.0,
+      "train_loss": 3.8128880228742164,
+      "train_runtime": 119.3474,
+      "train_samples_per_second": 21.383,
+      "train_steps_per_second": 2.673
     }
   ],
   "logging_steps": 500,
-  "max_steps": 3190,
+  "max_steps": 319,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 1,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -77,7 +35,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.000225898496e+16,
+  "total_flos": 1000225898496000.0,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null