Hugging Face's logo

SEBIS
/
code_trans_t5_base_code_documentation_generation_php_multitask_finetune

Summarization
Transformers
PyTorch
JAX
t5
feature-extraction
text-generation-inference
Model card Files
xet
 Community
2
wei commited on
Commit
13d2f17
·
1 Parent(s): 742aa79

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -7
README.md CHANGED
@@ -53,10 +53,6 @@ The optimizer used is AdaFactor with inverse square root learning rate schedule
53
  This model was then fine-tuned on a single TPU Pod V2-8 for 2000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing php code.
54
 
55
 
56
- ## Evaluation results
57
-
58
- For the code documentation tasks, different models achieves the following results on different programming languages (in BLEU score):
59
-
60
  Test results :
61
 
62
  | Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
@@ -69,9 +65,9 @@ Test results :
69
  | MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
70
  | MT-Base | **20.39** | **21.22** | 19.43 | **26.23** | **15.26** | 16.11 |
71
  | MT-Large | XX | XX | XX | XX | XX | XX |
72
- | MT-TF-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
73
- | MT-TF-Base | 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | **18.62** |
74
- | MT-TF-Large | XX | XX | XX | XX | XX | XX |
75
  | State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
76
 
77
 
 
53
  This model was then fine-tuned on a single TPU Pod V2-8 for 2000 steps in total, using sequence length 512 (batch size 256), using only the dataset only containing php code.
54
 
55
 
 
 
 
 
56
  Test results :
57
 
58
  | Language / Model | Python | Java | Go | Php | Ruby | JavaScript |
 
65
  | MT-Small | 19.64 | 19.00 | 19.15 | 24.68 | 14.91 | 15.26 |
66
  | MT-Base | **20.39** | **21.22** | 19.43 | **26.23** | **15.26** | 16.11 |
67
  | MT-Large | XX | XX | XX | XX | XX | XX |
68
+ | MT-FT-Small | 19.77 | 20.04 | 19.36 | 25.55 | 13.70 | 17.24 |
69
+ | **MT-FT-Base**| 19.77 | 21.12 | 18.86 | 25.79 | 14.24 | **18.62** |
70
+ | MT-FT-Large | XX | XX | XX | XX | XX | XX |
71
  | State of the art | 19.06 | 17.65 | 18.07 | 25.16 | 12.16 | 14.90 |
72
 
73