rajkumarrrk commited on
Commit
4adbfb5
·
1 Parent(s): e128fa1

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +41 -0
README.md CHANGED
@@ -1,3 +1,44 @@
1
  ---
2
  license: apache-2.0
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: apache-2.0
3
  ---
4
+ T5-base fine-tuned on CNN/DM Summarization dataset.
5
+
6
+
7
+ Training args:
8
+ ```
9
+ {
10
+ "learning_rate": 0.0001,
11
+ "logging_steps": 5000,
12
+ "lr_scheduler_type": "cosine",
13
+ "num_train_epochs": 2,
14
+ "per_device_train_batch_size": 16, # total batch size of 48
15
+ "save_total_limit": 1,
16
+ "weight_decay": 0.1
17
+ }
18
+ ```
19
+
20
+ Generation kwargs:
21
+ ```
22
+ {
23
+ "do_sample": true,
24
+ "max_new_tokens": 100,
25
+ "min_length": 50,
26
+ "temperature": 0.7,
27
+ "top_k": 0
28
+ },
29
+ ````
30
+
31
+ Pre-processing: Append prompt with prefix "Summarize: "
32
+ Post-processing: None
33
+
34
+ Test split metrics:
35
+
36
+ ```
37
+ {"lexical/meteor": 0.30857827917561603,
38
+ "lexical/rouge_rouge1": 0.41099971702474514,
39
+ "lexical/rouge_rouge2": 0.17676173608661166,
40
+ "lexical/rouge_rougeL": 0.2759112075051335,
41
+ "lexical/rouge_rougeLsum": 0.34316108028094616,
42
+ "lexical/bleu": 0.10747816852428271,
43
+ "semantic/bert_score": 0.8760301497472277}
44
+ ```