Update README.md
Browse files
README.md
CHANGED
|
@@ -2,7 +2,51 @@
|
|
| 2 |
datasets:
|
| 3 |
- csebuetnlp/xlsum
|
| 4 |
language:
|
| 5 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
pipeline_tag: summarization
|
| 7 |
---
|
| 8 |
# Model Card for Model ID
|
|
@@ -70,9 +114,10 @@ It achieves the following results on the evaluation set:
|
|
| 70 |
|
| 71 |
## Training hyperparameters
|
| 72 |
|
| 73 |
-
The model
|
| 74 |
-
- model:
|
| 75 |
-
- batch size:
|
| 76 |
-
- learning rate:
|
| 77 |
-
- number of epochs:
|
| 78 |
-
- warmup steps:
|
|
|
|
|
|
| 2 |
datasets:
|
| 3 |
- csebuetnlp/xlsum
|
| 4 |
language:
|
| 5 |
+
- am
|
| 6 |
+
- ar
|
| 7 |
+
- az
|
| 8 |
+
- bn
|
| 9 |
+
- my
|
| 10 |
+
- zh
|
| 11 |
+
- en
|
| 12 |
+
- fr
|
| 13 |
+
- gu
|
| 14 |
+
- ha
|
| 15 |
+
- hi
|
| 16 |
+
- ig
|
| 17 |
+
- id
|
| 18 |
+
- ja
|
| 19 |
+
- rn
|
| 20 |
+
- ko
|
| 21 |
+
- ky
|
| 22 |
+
- mr
|
| 23 |
+
- ne
|
| 24 |
+
- om
|
| 25 |
+
- ps
|
| 26 |
+
- fa
|
| 27 |
+
- pcm
|
| 28 |
+
- pt
|
| 29 |
+
- pa
|
| 30 |
+
- ru
|
| 31 |
+
- gd
|
| 32 |
+
- sr
|
| 33 |
+
- si
|
| 34 |
+
- so
|
| 35 |
+
- es
|
| 36 |
+
- sw
|
| 37 |
+
- ta
|
| 38 |
+
- te
|
| 39 |
+
- th
|
| 40 |
+
- ti
|
| 41 |
+
- tr
|
| 42 |
+
- uk
|
| 43 |
+
- ur
|
| 44 |
+
- uz
|
| 45 |
+
- vi
|
| 46 |
+
- cy
|
| 47 |
+
- yo
|
| 48 |
+
multilinguality:
|
| 49 |
+
- multilingual
|
| 50 |
pipeline_tag: summarization
|
| 51 |
---
|
| 52 |
# Model Card for Model ID
|
|
|
|
| 114 |
|
| 115 |
## Training hyperparameters
|
| 116 |
|
| 117 |
+
The model trained with a p4d.24xlarge instance on aws sagemaker, with the following config:
|
| 118 |
+
- model: deltalm base
|
| 119 |
+
- batch size: 8
|
| 120 |
+
- learning rate: 1e-5
|
| 121 |
+
- number of epochs: 3
|
| 122 |
+
- warmup steps: 500
|
| 123 |
+
- weight decay: 0.01
|