add animations to model card
Browse files
README.md
CHANGED
|
@@ -28,10 +28,14 @@ pipeline_tag: fill-mask
|
|
| 28 |
|
| 29 |
---
|
| 30 |
|
| 31 |
-
<!-- TODO: change widget text -->
|
| 32 |
-
|
| 33 |
# sahajBERT
|
| 34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
Collaboratively pre-trained model on Bengali language using masked language modeling (MLM) and Sentence Order Prediction (SOP) objectives.
|
| 36 |
|
| 37 |
## Model description
|
|
@@ -157,6 +161,12 @@ This model was trained in a collaborative manner by volunteer participants.
|
|
| 157 |
|39|[Rounak](https://huggingface.co/Rounak)|0 days 00:26:10|
|
| 158 |
|40|[kshmax](https://huggingface.co/kshmax)|0 days 00:06:38|
|
| 159 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 160 |
## Eval results
|
| 161 |
|
| 162 |
We evaluate sahajBERT model quality and 2 other model benchmarks ([XLM-R-large](https://huggingface.co/xlm-roberta-large) and [IndicBert](https://huggingface.co/ai4bharat/indic-bert)) by fine-tuning 3 times their pre-trained models on two downstream tasks in Bengali:
|
|
|
|
| 28 |
|
| 29 |
---
|
| 30 |
|
|
|
|
|
|
|
| 31 |
# sahajBERT
|
| 32 |
|
| 33 |
+
|
| 34 |
+
<iframe width="100%" height="1100" frameborder="0"
|
| 35 |
+
src="https://observablehq.com/embed/@huggingface/participants-bubbles-chart?cells=c_noaws%2Ct_noaws%2Cviewof+currentDate"></iframe>
|
| 36 |
+
|
| 37 |
+
|
| 38 |
+
|
| 39 |
Collaboratively pre-trained model on Bengali language using masked language modeling (MLM) and Sentence Order Prediction (SOP) objectives.
|
| 40 |
|
| 41 |
## Model description
|
|
|
|
| 161 |
|39|[Rounak](https://huggingface.co/Rounak)|0 days 00:26:10|
|
| 162 |
|40|[kshmax](https://huggingface.co/kshmax)|0 days 00:06:38|
|
| 163 |
|
| 164 |
+
|
| 165 |
+
### Hardware used
|
| 166 |
+
|
| 167 |
+
<iframe width="100%" height="251" frameborder="0"
|
| 168 |
+
src="https://observablehq.com/embed/@huggingface/sahajbert-hardware?cells=c1_noaws"></iframe>
|
| 169 |
+
|
| 170 |
## Eval results
|
| 171 |
|
| 172 |
We evaluate sahajBERT model quality and 2 other model benchmarks ([XLM-R-large](https://huggingface.co/xlm-roberta-large) and [IndicBert](https://huggingface.co/ai4bharat/indic-bert)) by fine-tuning 3 times their pre-trained models on two downstream tasks in Bengali:
|