End of training

README.md

@@ -3,9 +3,23 @@ license: apache-2.0
 base_model: google-t5/t5-base
 tags:
 - generated_from_trainer
+datasets:
+- fedora-copr/pep-sum
+metrics:
+- rouge
 model-index:
 - name: pep_summarization
-  results: []
+  results:
+  - task:
+      name: Summarization
+      type: summarization
+    dataset:
+      name: fedora-copr/pep-sum
+      type: fedora-copr/pep-sum
+    metrics:
+    - name: Rouge1
+      type: rouge
+      value: 87.9903
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -13,7 +27,14 @@ should probably proofread and complete it, then remove this comment. -->
 
 # pep_summarization
 
-This model is a fine-tuned version of [google-t5/t5-base](https://huggingface.co/google-t5/t5-base) on an unknown dataset.
+This model is a fine-tuned version of [google-t5/t5-base](https://huggingface.co/google-t5/t5-base) on the fedora-copr/pep-sum dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0348
+- Rouge1: 87.9903
+- Rouge2: 87.5298
+- Rougel: 88.0594
+- Rougelsum: 87.9148
+- Gen Len: 67.8551
 
 ## Model description
 

all_results.json

@@ -1,8 +1,18 @@
 {
     "epoch": 10.0,
-    "train_loss": 0.06454240335686796,
-    "train_runtime": 163.8058,
-    "train_samples": 345,
-    "train_samples_per_second": 21.062,
-    "train_steps_per_second": 5.311
+    "eval_gen_len": 67.85507246376811,
+    "eval_loss": 0.03478159010410309,
+    "eval_rouge1": 87.9903,
+    "eval_rouge2": 87.5298,
+    "eval_rougeL": 88.0594,
+    "eval_rougeLsum": 87.9148,
+    "eval_runtime": 11.488,
+    "eval_samples": 69,
+    "eval_samples_per_second": 6.006,
+    "eval_steps_per_second": 1.567,
+    "train_loss": 0.04075576412504998,
+    "train_runtime": 147.9128,
+    "train_samples": 276,
+    "train_samples_per_second": 18.66,
+    "train_steps_per_second": 4.665
 }

eval_results.json

@@ -0,0 +1,13 @@
+{
+    "epoch": 10.0,
+    "eval_gen_len": 67.85507246376811,
+    "eval_loss": 0.03478159010410309,
+    "eval_rouge1": 87.9903,
+    "eval_rouge2": 87.5298,
+    "eval_rougeL": 88.0594,
+    "eval_rougeLsum": 87.9148,
+    "eval_runtime": 11.488,
+    "eval_samples": 69,
+    "eval_samples_per_second": 6.006,
+    "eval_steps_per_second": 1.567
+}

generation_config.json

@@ -1,5 +1,4 @@
 {
-  "_from_model_config": true,
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 10.0,
-    "train_loss": 0.06454240335686796,
-    "train_runtime": 163.8058,
-    "train_samples": 345,
-    "train_samples_per_second": 21.062,
-    "train_steps_per_second": 5.311
+    "train_loss": 0.04075576412504998,
+    "train_runtime": 147.9128,
+    "train_samples": 276,
+    "train_samples_per_second": 18.66,
+    "train_steps_per_second": 4.665
 }

trainer_state.json

@@ -3,33 +3,33 @@
   "best_model_checkpoint": null,
   "epoch": 10.0,
   "eval_steps": 500,
-  "global_step": 870,
+  "global_step": 690,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 5.75,
-      "learning_rate": 2.1264367816091954e-05,
-      "loss": 0.1123,
+      "epoch": 7.25,
+      "learning_rate": 1.3768115942028985e-05,
+      "loss": 0.0529,
       "step": 500
     },
     {
       "epoch": 10.0,
-      "step": 870,
-      "total_flos": 4200948340346880.0,
-      "train_loss": 0.06454240335686796,
-      "train_runtime": 163.8058,
-      "train_samples_per_second": 21.062,
-      "train_steps_per_second": 5.311
+      "step": 690,
+      "total_flos": 3359758649241600.0,
+      "train_loss": 0.04075576412504998,
+      "train_runtime": 147.9128,
+      "train_samples_per_second": 18.66,
+      "train_steps_per_second": 4.665
     }
   ],
   "logging_steps": 500,
-  "max_steps": 870,
+  "max_steps": 690,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 10,
   "save_steps": 500,
-  "total_flos": 4200948340346880.0,
+  "total_flos": 3359758649241600.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null